DATC

DIPLOMACY ADJUDICATOR TEST CASES

Copyright: Lucas B. Kruijswijk

HISTORY OF CHANGES

Reviewed by:

• Flooey X. McBob
• Brian Roberts
• Andrew Rose
• Rick Desper
• Simon Szykman
• Millis L. Miller
• David McCooey
• Christian Hagenah
• Estaban U.C. Castro
• Ronnie van 't Westeinde
• David E. Cohen
• Ted Driggs
• Matt Pavord

You are allowed to call an adjudicator DATC compliant if all test cases of chapter 6 pass or deviations are done consciously.

License note: You are free to copy or use this document. When changing this document, this license must be included and the change log must be maintained. The test cases of chapter 6 are allowed to be copied separately without this restriction.

The main location where the latest version of this document is published is on BoardGameGeek, Diplomacy, Files.

Diplomacy is the Avalon Hill Game Company's trademark for its game of international intrigue, which game is copyright 1976 by Avalon Hill. Avalon Hill belongs to Hasbro.

TABLE OF CONTENTS

1.INTRODUCTION
2.HISTORY OF RULES
3.SUGGESTIONS FOR FUTURE RULEBOOKS
4.DISPUTABLE ISSUES
A.CONVOY ISSUES
B.MUTLI-COAST ISSUES
C.UNIT DESIGNATION AND NATIONALITY ISSUES
D.TOO MANY AND TOO FEW ORDERS
E.MISCELLANEOUS ISSUES
5.THE PROCESS OF ADJUDICATION
A.OVERVIEW OF ADJUDICATION DECISIONS
B.PRECISE DESCRIPTION OF MAKING DECISIONS
C.FROM CONDITIONS TO ALGORITHM
D.THE TROUBLE WITH PANDIN'S PARADOX
E.THE PARTIAL INFORMATION ALGORITHM
F.THE GUESS ALGORITHM
6.TEST CASES
A.TEST CASES, BASIC CHECKS
B.TEST CASES, MULTI-COAST ISSUES
C.TEST CASES, CIRCULAR MOVEMENT
D.TEST CASES, SUPPORTS AND DISLODGES
E.TEST CASES, HEAD-TO-HEAD BATTLES AND BELEAGUERED GARRISON
F.TEST CASES, CONVOYS
G.TEST CASES, CONVOYING TO ADJACENT PROVINCES
H.TEST CASES, RETREATING
I.TEST CASES, BUILDING
J.TEST CASES, CIVIL DISORDER AND DISBANDS

1. INTRODUCTION

Writing an adjudicator computer program for the game Diplomacy is not an easy job. Prior to this document many adjudicator programs contained several bugs on their first release and even after many years, when the most severe bugs were removed, adjudication errors were still found for more complex situations. To achieve a high quality adjudicator, the programmer has to overcome the following difficulties:

• Different rulebooks.
• Ambiguities in the rulebooks.
• An algorithm that is sophisticated enough to handle complex situations.
• Systematic testing of the adjudicator.

This document is a guide to handling these problems and enabling the programmer to write an adjudicator that is correct on the first release. The principle of this document is to give the reader information and not to tell the reader how to do things. Therefore, all information is presented as neutrally as possible and given with arguments or with reference to the source.

The sources of this document are the official English rulebooks, material from Allan B. Calhamer (the creator of the game), the Diplomatic Pouch the Zine, Diplomacy World, the 1998 DPTG (Diplomacy Players Technical Guide), the Model House Rules by David E. Cohen (see the Diplomatic Pouch the Zine, Spring 2001 Movement), posts on different online forums, house rules and zines. Some of the material was provided by Stephen Agar who goes to great lengths to preserve all things related to Diplomacy. All this makes the DATC the most elaborate and complete source on the rules of Diplomacy.

An overview of the official English rulebooks is given in chapter 2. The older rulebooks may differ on small issues with the most recent rulebook or are ambiguous on certain issues. These issues are listed in chapter 4 with the alternatives to handle them. An analysis of the process of adjudication is in chapter 5. With this analysis, writing an adjudicator program is a straight forward job. Finally, an adjudicator program needs to be tested in a systematic way and chapter 6 contains an extensive list of test cases with the expected resolution.

In case the rules are not clear on a certain situation (most of the issues are fixed in the 2023 rules), then the alternatives are in the list of issues and in the test cases. To prevent the reader from becoming lost in lots of choices I (Lucas B. Kruijswijk) commented every rule issue. Those comments contain my preference as to how the rule should be interpreted. In this way the reader can start with these preferences and deviate where he or she wishes. To make a clear distinction between the text that is not disputed and my comments with my preferences, my comments are written in italics.

With the annotated information about the rules, this document can also be used by people who judge a game manually, using house rules with the preferences of the player group and the DATC recommendations for everything that isn't covered.

The version of this document consists of two numbers. The first number will only change when one or more preferences changes. The second one is a sequence number.

2. HISTORY OF RULES

Allan Calhamer started the development of Diplomacy in 1953. The first commercial sets were assembled by Calhamer and sold from his apartment in 1959. So, the first official rulebook dates 1959. He describes these initial efforts in a long letter published in Diplomania No. 12, 1966.

Games Research bought the rights of Diplomacy in 1960 and their first edition was in 1961. Except from some restyling and the copyright notice of Games Research, the rules of 1961 are not different from the 1959 rules. These rulebooks have many ambiguities and are rather useless now. Dealing with the issues of these rulebooks is out of the scope of this document.

Most of the issues of the 1959 and 1961 rulebooks were identified and addressed in the 1971 rulebook. The 1971 rulebook is the first mature rulebook. Background on how these rules were drafted can be found in the article by Rod Walker in Diplomacy World No. 28 page 16, 1981. Since the 1971 rulebook has been popular for a long time and maybe still played by, the differences with newer rulebooks are all explained in this document.

In 1976 Avalon Hill bought the rights from Games Research. In their first edition in 1976, they kept the rules the same and they only put their own name on the rulebook. In this document, only the first appearance of a set of rules is referred. That means that the 1971/1976 rulebooks are referred as the 1971 rulebook in this document.

In 1982 Avalon Hill made a new edition of the game and made some small, but significant changes to the rules. These changes include the multi-route convoy disruption rule (see issue 4.A.1), convoy paradox rule (see issue 4.A.2) and some clarifications on convoying to adjacent provinces (see issue 4.A.3).

In 1992 Avalon Hill they introduced the 'Diplomacy de Luxe' edition. The rulebook was restyled, but the rules are identical to the 1982 rules. The rulebook does also contain a list of abbreviations for the provinces. Since, the rules did not change, the 1982/1992 rulebooks are referred as the 1982 rulebook in this document.

In 1994 Colonial Diplomacy was released. A variant with a Eurasian map. Strangely enough, the rules were based on the 1971 rulebook, rather than on the newer 1982 one. It contains a special rule for the Suez-Canal and the Trans-Siberian-Railroad. These rules were specific to this variant and with new issues and complications, while at the same time not adding much to the game.

In 1998 Avalon Hill was bought by Hasbro and in 2000 the first Hasbro edition emerged. Note that you can see the year 2000 on the front page of the rulebook, however the copyright notice on the last page says 1999. Some people refer to these rules as the 1999 rules. In this document it is assumed that the front page is read more than the last page and therefore these rules are referred to as the 2000 rules. The rules were completely rewritten with most notable change the illustration of examples with diagrams. In previous editions they were just text. Edi Birsan was the main contributor to this rulebook. Aside from restyling, there are also some real changes to the rules. The unpopular 1982 convoy paradox rule (see issue 4.A.2) was changed again. The issue on convoys to adjacent provinces was further clarified (see issue 4.A.3). Declining builds is explicitly allowed (see issue 4.D.7). Also, a significant change in the abbreviations of the provinces (note that the abbreviation of 'Denmark' is missing).

Unfortunately, there are a few changes in the 2000 rulebook that made things worse than the 1982 rules. The first issue has to do with the removal of units in civil disorder (see 4.D.8). In the 1971/1982 rulebooks the distance had to be calculated to the home supply centers. The exact way was not specified. Instead of fixing this issue, the situation became worse. The distance (precise method still unspecified) had to be calculated to "home country". The home country is more than the home supply centers, but this information is often not available in map data. This change was ignored entirely by the community.

A second issue was that it forgot to say that a disrupted convoy has no influence on the destination province. Page 12 of the 2000 rulebook:

Dislodgment of a fleet in a convoy causes the convoy to fail. If a Fleet ordered to convoy is dislodged during the turn, the Army to be convoyed remains in its original province.

While rule XII.3 of the 1971 and 1982 rules say:

DISRUPTING A CONVOY. If a fleet ordered to convoy is dislodged during the turn, the army to be convoyed remains in its original province and has no effect on the province to which it was ordered.

There are several reasons to assume that the 2000 rulebook was not intended like this and that this must be considered an error in the rules. First of all, Edi Birsan the main contributor to the 2000 rulebook, confirmed that this was not intended. Second, there is evidence in the rulebook itself that the rule should be read as the 1971/1982 rule. The whole page 16 would become unnecessary if convoying armies can still cut support when the convoy is dislodged (there are no paradoxes anymore). But especially the following phrase on that page is clear evidence:

Italy could argue that dislodgment of the Fleet disrupted the convoy so that the Army could not arrive in Naples to cut the support. (Italy could state the rule, "Dislodgment of a fleet in a convoy causes the convoy to fail.")

In 2000 Hasbro/Avalon Hill/Microprose also released a CD-ROM version with a tutorial of the rules. In the research of this document no additional information could be found about this version.

In 2008, a new version of the game was published by Wizards of the Coast (also bought by Hasbro), although the box still has the Avalon Hill brand. The rules were not changed except that the Denmark abbreviation was added. In this document they are still referred as 2000 rules.

In 2023 Renegade Games Studios took a license on the game and updated the rulebook fixing most ambiguities. The most notable change is disbanding units in civil disorder. The distance calculation is now unambiguous specified, but calculated from "owned" supply center instead of "home" supply center (see issue 4.D.8). Coast specification in support order is possible and must match (issue 4.B.4). Impossible move orders do not prevent support in hold (issue 4.E.1). Retreating to the area of the attacker is possible if it was via convoy (issue 4.A.5). Unfortunately, the issue of disrupted convoys, that slipped into the 2000 rules, was not fixed.

In 2025 Renegade Games Studios rebooted Colonial Diplomacy under the new name Era of Empire. Although the map looks the same on first glance it has many significant changes. The Suez-Canal and Trans-Siberian-Railroad special rules were dropped, but it allows to build in any owned supply center and it also allows to convert a unit in a supply center (of course, this only applies for this variant). These deviations were already present in other variants and are not difficult to implement. The rulebook is an evolution of the 2023 edition. Many textual changes were made, something that was not done to the 2023 edition. The diagrams are better placed within the text, dropping the numbers and references. The disrupted convoy issue of the 2000 rules is fixed. The ambiguity in the rule for convoying to an adjacent province (issue 4.A.3) was removed. New in these rules are the clarifications about poorly written orders. The issues of double orders (4.D.1), missing unit designation (4.C.1), too many build orders (4.D.4), too many disband orders (4.D.6) are all addressed. Basically, the rules imply (but do not explicit state) that orders should be processed from top to bottom. Earlier orders take precedence over subsequent orders.

Since it is expected that the changes will also end up in the next edition of the standard rules, these rules are considered canon and are referred as 2025EE (Era of Empire) rules.

Note, I got the honor to participate in the review of this edition.

Finally, there are several translated rulebooks made for non-English versions of Diplomacy. These are not treated in this document.

3. SUGGESTIONS FOR FUTURE RULEBOOKS

The list of differences as listed in the previous chapter may look a lot, but in actual play they are of minor significance. If a modern adjudicator would follow the 1971 rules unanounced, then it would hardly be noticed. The rules of Diplomacy are sacred and changing them should only be done to improve readability or to close one of the few remaining gaps. Taking this into account I suggest the following for the person that gets the honor to be editor of a new edition.

There are three substantive changes which I would like to be considered. Substantive, in the sense that the changes are more than textual although still tiny and only of significance in very rare situations:

1. Allow support orders without coastal specification be valid for both coasts, see discussion by issue 4.B.4.
2. Simplify the order writing for a convoy to adjacent province, by requiring the 'via convoy' tag in all situations. See discussion by issue 4.A.3.
3. Adopt the Szykman rule for convoy paradoxes. The community has settled on this. See issue 4.A.2.

Changes that only improve readability and understandability are more numerous.

The 2025EE rules clarify a number of issues when orders are not properly written, such as too many builds or disbands. However, this information is scattered over the rulebook and sometimes repeated. It is three times mentioned that a unit without an order will hold. I think it will improve readability if this is bundled in a separate section. A player can skip this information when reading the rulebook for the first time and when a situation occurs that it is relevant, a separate section is easier to find back. We may call this section 'poorly written orders', but if we want to reserve that term for an order that is actually acceptable, then we can also call the section 'improper orders'. Such section could look like:

POORLY WRITTEN ORDERS

When the orders are revealed, it may happen once in a while, that an order is not clear or not according to the rules. It is up to the Game Master what to accept, but if played without, the following guidelines can be used:

• In all phases, including the movement phase, retreat phase and adjustment phase, orders are executed from top to bottom. Earlier orders take precedence over subsequent orders. A unit given multiple orders, will execute the first order given. In the case of too many builds or disbands, the first ones are executed.
• A poorly written order that has only one meaning must be followed. For example, "North Atlantic Ocean - Mid-Atlantic Ocean". Although the fleet designation is missing, the order has still one meaning.
• Writings that are ambiguous, not an order at all, or an order that is impossible are ignored entirely. For example, "A Burgundy—Moon" or "A Bohemia—Edinburgh." As consequence, such impossible move cannot be used to refuse a hold support. Orders that do not match or move orders requiring a convoy for which the fleets are there but without the right convoy orders, are legal, but fail directly.
• If a unit has no order at the end of processing the order set, it will hold and can receive support.
• In the case of insufficient disbands in the adjustment phase, the remaining units are automatically disbanded by the procedure described by civil disorder.

The automatic disband refers to the civil disorder. It would make sense to do it the other way around. Civil disorder is in principle the same as submitting an empty order set. However, traditionally the removal procedure has been part of civil disorder.

The DATC examines many more issues regarding poorly written orders. However, codifying this in the rulebook has also a disadvantage. Players with good knowledge of the ruling may use it to argue with Game Master or deliberately write poorly written order, knowing how they will eventually be interpreted.

With this section, some language in the start of the rulebook can be cleaned up. The current text is:

Each player secretly writes "orders" for each of his or her units on a slip of paper. All players then reveal orders at the same time. Each player reads his or her orders while others make sure that what they hear is what is written. Alternatively, the gamemaster can collect all orders and read them for the players. A legal order must be followed. An order written by mistake, if legal, must be followed. An "illegal" order or an order that is judged to be unsuccessful isn’t followed. A unit that is given an illegal order (or given no order) must stand in place (the unit holds). A poorly written order that has only one meaning must be followed.

These technical details set forth immediately at the beginning of the rulebook, while writing orders is not explained yet, is not a good idea in my opinion. With the separate section we can change it to something more conceptual. Something that refers to the fact that revealing the orders is an irreversible act:

Each player secretly writes "orders" for each of his or her units on a slip of paper. All players then reveal orders at the same time. Each player reads his or her orders while others make sure that what they hear is what is written. Alternatively, the gamemaster can collect all orders and read them for the players. Now, the intent of the players has become clear. Orders should be executed as written, without clarification of the submitter. It may happen that an order is not that clear. Guidelines on how to deal with such situation are given later in the rulebook.

Similarly to poorly written order, one might also consider to put all the rules related to multi-coast provinces in one section. This includes movement, support, building and order writing. In separate section it is probably easier for a player to find back, if an question arises during a game.

The 2025EE rules do have the examples better placed within the text and dropped the numbering and referring of the diagrams. This makes it more readable, but it could be done more thoroughly. There are still places where example orders are given in the text and then repeated below the diagram.

The rule that a support is cut when dislodged is mentioned on page 9 of the 2025EE rules, but without any example. Earlier rulebooks do give an example.

Since the 2000 rules the term beleaguered garrison is not used anymore. This is the situation where a unit survives multiple attacks by superior forces, if these attacks prevent each other from dislodging the unit. This rule is still explicitly given (on page 6 of the 2000 rules and with similar text in later rules), but without any example. Also, it is mentioned on a place when dislodgment has not yet been explained. In the 1982 rules there was an example (see page 5). I would like to see the term beleaguered garrison restored with an example in the section of "SPECIAL CIRCUMSTANCES".

On the other hand, the 2000/2023/2025EE rules have at the end two diagrams related to convoy paradoxes. The situations depicted do not introduce a new rule and a simple note is sufficient in my opinion. It is not necessary to illustrate all the theoretical cases with an example as long as the text is unambiguous and the player can find the rule in a logical place.

The 2023/2025EE rules have the following text in the civil disorder rule:"(a) No unit on a supply center is to be removed unless there are no viable options." This can be entirely removed, because a unit on a supply center in winter will make the supply centers always owned and will have a distance of zero.

A player is not obliged to build a unit. The 2000 rules were the first rules to explicitly state this with "declining" a build. I get the impression that "waiving" a build is (or was) the more common terminology used by the community. A meaning found on the internet for "waive" is: "refrain from insisting on or using a right or claim." While for "decline" the following meaning can be found: "politely refuse an invitation or offer." Given these meanings, I think the term "waive" is more appropriate.

On the 22 summary rules, I would suggest:

• "All units have the same strength." I would prefer: "The strength of a unit is one plus the number of successful supports."
• "Units of equal strength trying to occupy the same province cause all those units to remain in their original provinces." I would prefer: "For a move to succeed the strength of a unit must be superior to any opposing or competing units."
• "A dislodgment of a Fleet necessary to a convoy causes that convoy to fail." I would prefer to split this in two rules (more in line with the main rules): "A dislodged Fleet cannot participate in a convoy. A fleet that is merely attacked can participate in a convoy." And a second part: "A convoy fails if all possible routes are disrupted."
• "Two units can exchange places if either or both are convoyed. (This is the exception to Rule 6.)" I do not consider this text necessary, because rule 6 already makes it clear. However, one could add: "To convoy to an adjacent province, for the purpose of a swap, the order must be tagged with 'via convoy'"
• "An Army with at least one successful convoy route will cut the support given by a unit in the destination province that is supporting an attack on a Fleet in an alternate route in that convoy. (This supersedes Rule 21.)" I consider this text unnecessary. The convoy paradox rule already requires that the fleet is "necessary". If the Szykman rule is adopted, then this is also not necessary.

The 2025EE rules have on page 11 a summary of the resolution phase. I think one should take care that things are in general not repeated. There is already the 22 rules summary at the end and there is a Quick Start Rules. The drawback is that you end up with more text, making it less accessible. Diplomacy rules are not the easiest ones, but at the other hand there are grand board games with far more voluminous and complex rules.

4. DISPUTABLE ISSUES

In this chapter the most known rule issues and issues of order interpretation are discussed. Note, with the 2023/2025EE rules most adjudication issues are settled.

The geography of the map is not disputed and therefore not discussed. To the surprise of some new players, Norway is connected to St Petersburg both for armies and for fleets, but this is not disputed. Although also not disputed, on some maps it is not completely clear, that Liverpool is connected to North Atlantic Ocean and Clyde is not connected to Irish Sea.

The issues in this chapter are not a FAQ (Frequently Asked Question) list, but most frequently asked questions can be found in the test cases. For instance, the coastal crawl is in test case 6.B.13.

4.A. CONVOY ISSUES

4.A.1. MULTI-ROUTE CONVOY DISRUPTION

When a convoy has multiple routes, the question arises as to when the convoy is disrupted.

The following interpretations are possible:

1. The convoy is disrupted when one of the routes is disrupted.
   This is according to the 1971 rulebook.
2. The convoy is disrupted when all of the possible routes are disrupted.
   This is according to all subsequent rulebooks.

See test case 6.F.10. In 1974 Eric Verheijden proposed and promoted the change to the rulebook in Diplomacy World No. 5 page 22:

Unfortunately, this manner of handling the problem can lead to some truly singular rulings. To take only a few illustrative examples:

• England: A Lon-Bel, F Wal-Eng, F Iri S F Wal-Eng, A Hol S A Lon-Bel
• Germany: F Eng C ENGLISH A Lon-Bel /d/
• Russia: F Nth C ENGLISH A Lon-Bel

Here, it is fair to say, English intentions are reasonably clear. You do not dislodge the fleet you are intending to have convoy your way.
...
Why not simply state that if, after all dislodgments have been taken into account, a continuous route of convoying fleets exists from an army's point of departure to its destination, the convoy proceeds, barring of course prohibition by some other rule?
...
In my opinion, any increase in the tactical force of the convoy order--or indeed any change in the way the game is now played--would be negligible. There would indeed be some slight changes in certain rare situations. But these situations would be very rare indeed. In almost all convoy situations, alternate convoy routes are impossible. In the few situations where they are possible, in most cases they would be unprofitable; instead of ordering two fleets to convoy, for example, a player could order one to convoy and the other to support the convoying fleet. It would only be in the relative handful of cases remaining that there would be any change and even these would be amply compensated for by the elimination of spurious convoy orders in a comparable number of cases.
...
In closing, it might be noted, in the interest of giving credit where it is due, that the idea of alternate convoy routes themselves is not original with me, although their use in the present context is, to the best of my knowledge. The term which has been used previously is "Shagrin's Alternate Convoy," although the details are not known to me.

Verheijden was right and there is no reason to go back to the 1971 ruling. I prefer choice b.

4.A.2. CONVOY PARADOXES

A convoy paradox is a situation with a possible disrupted convoy and for which the rules give no resolution or more than one resolution.

Some people argue that some situations are not convoy paradoxes, since the rules give a resolution for those situations. Hence, for a proper discussion on paradoxes, rule XII.5 of the 1971, 1976, 1982 and 1992 rulebooks, the rule on the top of page 16 of the 2000 rulebook, rule on page 18 of the 2023 rulebook and rule on page 16 of 2025EE rulebook should not be taken into account in the above definition of a paradox.

Different rulebooks rule differently:

1. The 1971 rule about this issue is:
   "If a convoyed army attacks a fleet which is supporting a fleet which is attacking one of the convoy fleets, that support is not cut."
   This rule has a few drawbacks. The rule does not have a resolution for second order paradoxes (see test cases 6.F.22, 6.F.23, and 6.F.24) and the betrayal paradox (see test case 6.F.18).
2. The 1982 rule about this issue is:
   "If a convoyed army attacks a fleet which is supporting an action in a body of water; and that body of water contains a convoying fleet, that support is not cut."
   The advantage of this rule is that it is a simple and effective rule that eliminates all paradoxes. The disadvantage is that it effects not only the paradoxical situations, but also many other cases. This has as consequence that players can do tricks with this rule to take advantage (see test cases 6.F.20 and 6.F.21). It is also not clear what a "convoying fleet" means. Does it mean a fleet that is ordered to convoy? Or must there be at least an army that tries to convoy? Or must there be a possible route?
3. The 2000/2023/2025EE rule about this issue is:
   "A convoyed Army does not cut the support of a unit supporting an attack against one of the fleets necessary for the army to convoy."
   This rule is just a reversal of the 1982 rule back to the 1971 rule, but adapted for the changed multi-route convoy disruption rule (see issue 4.A.1) which was introduced in the 1982 rulebook. This has as advantage that it leads to more logical adjudication in the case of multi-route convoys (see test case 6.F.19). It has also all disadvantages of the 1971 rule. So, the rule does not have a resolution for second order paradoxes (see test cases 6.F.22, 6.F.23 and 6.F.24) and the betrayal paradox (see test case 6.F.18).

All rulebooks allow a situation where a convoying army dislodges a unit that gives successful support (see test case 6.F.17). This violates the principle that the support of a dislodged unit is cut and this aberration pops up in almost every discussion about convoy paradoxes. In 1972 a letter of Calhamer was published in Erehwon No. 65:

3. ENGLAND: A Edi-Hol, F Nth C A Edi-Hol, A Bel S A Edi-Hol, F Eng S F Nth.
   GERMANY: F Den-Nth, F Hol S F Den-Nth /r/.

I think the support of F Hol is not cut, due to Rule XII.5, but the result is the same either way. To this extend, it appears that XII.5 is an exception to X.
...
So far as I can see, the only real problem is 3, because it is necessary to treat XII.5 as overriding X. I think this construction would be understood because XII.5 is more specific than X and comes afterward, but I agree that I would like it better if XII.5 were specified in the rules as overriding X.

Where X is the dislodge rule and XII.5 the convoy paradox rule (these are the numbers used in the 1971 and 1982 rulebooks).

Due to the incompleteness of the paradox rules, several alternatives or additions has been proposed over the years:

4. The 'All Hold' backup rule:
   "If a situation arises in which an army's convoy order results in a paradoxical adjudication, all the moves part of the paradoxical situation fail."
   This rule was played in the early days. Rod Walker writes in Erehwon No. 47, 1971: "John Boardman writes that he has reached much the same conclusions that I have, but his ruling at present would be that no units move. That was the ruling I first proposed to Tony Pandin and Eric Just, when we first discussed this problem. (see THE DIPLOMAT I.47) This ruling recognizes, but does not resolve the paradox."
   The rule cannot replace the existing rules without breaking tradition, because it gives a different adjudication for the simplest paradox (see test case 6.F.14), while that paradox might happen in actual play. Having the rule as backup rule means that we have two paradox rules stacked, instead of a single rule. Furthermore, it is not entirely trivial to determine which moves belong to the paradox. With new adjudicators opting for the Szykman rule, the 'All Hold' rule lost its popularity and its resolution will not be given by the test cases.
5. Some ideas looked at whether a convoy has "affect" or "cause" on itself. For instance, in Diplomacy World No. 14 page 9, 1976, Eric Verheijden defines his rule: "A convoyed attack may not affect the convoying fleets. If a convoyed army's attack would affect the outcome of an attack on any of its convoying fleets, however directly or indirectly, then the convoyed attack may not take place and the army to be convoyed must hold in its original position." He intends the same as the Szykman rule, making the convoy the weaker operation while adjudicating the other orders as normal. However, the use of the word "affect" doesn't do more than identifying an order within a paradox. Since convoy paradoxes are a known phenomenon, this does not add much. When applying a backup rule on the paradox this must be done on all orders of the paradox simultaneous. This requires that the orders of a paradox should be identified as a set and not just as single orders. Something that Verheijden doesn't do.
6. Simon Szykman alternative:
   "If a situation arises in which an army's convoy order results in a paradoxical adjudication, the turn is adjudicated as if the convoying army had been ordered to hold."
   This rule was proposed by Simon Szykman in a discussion with Manus Hand in the Diplomatic Pouch the Zine, Fall 1999 Retreat. Simon defended his rule, while Manus Hand defended the 1982 rule. The Szykman rule is mostly compatible with the 2000/2023 rulebook and has a resolution for all paradoxes. Only in a case where the convoying army dislodges a unit (see test case 6.F.17) the result is different. In the many letters that were send as reaction on this article, it was noted that adjudicating the order as a hold order means that the unit can receive support in holding. This was not intended and in the Diplomatic Pouch the Zine, Winter 1999 Adjustment, Szykman together with the editor proposed: "If a situation arises in which an army's convoy order results in a paradoxical adjudication, the turn is adjudicated as if the convoying fleet had not been ordered to convoy but holds instead." Targeting the convoying fleet instead of the convoying army has as advantage that there is less room for error when manually adjudicated, since there is often only one convoying fleet. So, unlike the 'All Hold' rule not much can go wrong. Implementing it in a program (see chapter 5) will likely target the fleet also. Although the convoying army is an essential ingredient of the convoy paradox, whether the convoying army succeeds or not, is not relevant and therefore not part of the cyclic dependent orders.

Several posts and articles (the DPTG and the further explanations by Verheijden) attempt to clarify things with an algorithm. It should be noted that the official rulebooks do not describe an algorithm, but merely describe conditions to which the adjudication should comply (see chapter 5). A rule with an algorithmic approach is unlikely to be adopted in a future rulebook.

Of all the proposals the Szykman rule has the weakest convoy (with the 'All Hold' rule the convoying fleet will never be dislodged, while with the Szykman rule this can happen). The weak convoy would probably be appreciated by Calhamer. In his letter published in Erehwon No. 65, 1972 he writes:

Also I do not desire to give "great flexibility and tactical force to the convoy order". I think World War I represented just about the nadir of amphibious warfare. World War II represented something of a peak. Furthermore, we had some experience in our original group with more highly tactical versions of Diplomacy and did not like them, because players paid too much attention to tactics, and tried to achieve everything tactically, and didn't pay much attention to the diplomacy.

All adjudicators written after the inception of the DATC opted for the Szykman rule.

I prefer the Szykman rule, because of the given advantages. If the rule is adopted in an official rulebook, then one probably wants to reformulate it. I recommend to not try to define a convoy paradox. They are called convoy paradoxes in the literature starting from the early days. The attempt of Eric Verheijden made it more confusing, while not adding anything. One or two examples should be sufficient to explain. As said earlier, the rule should target the convoying fleet, rather than the convoying army. The paradoxical situation can have more than one fleet (higher order paradoxes) and the orders of those fleets should fail (or changed to hold) at once. The formulation should be such that situations with multiple fleets are covered.

4.A.3. CONVOYING TO ADJACENT PROVINCE

The 1971/1982 rulebooks contain the text: "Two pieces may exchange places if either or both are convoyed." The formulation in later rulebooks is slightly differently: "Two units can’t trade places without the use of a convoy."

From the text it is clear that two units can trade places with just one convoy and in such case the provinces must be adjacent. This further means that a convoy to an adjacent province is possible. Issues arise when such convoy is ordered by a foreign power. This can happen in agreement with the player of the convoying army, but can also be "unwanted". This to try to swap two units or to frustrate the move of the army when the fleet is likely to be dislodged. This is called "kidnapping" an army. In 2021 Diplostrats made the YouTube movie "DiploShorts - Convoy Kidnapping"

Relevant test cases are in section 6.G.

1. The 1971 rulebook does not specify the route in case a convoy to an adjacent province is possible.

Rod Walker doesn't consider this as a problem and writes in Diplomacy World No. 28 page 20, 1981:

6A. The Unwanted Convoy
ENGLAND: F Eng C FRENCH A Pic-Bel. (dislodged)
FRANCE: A Pic-Bel S by A Bur, F Bre-Eng S by F Mid.
This peculiar special case crops up a lot in discussions of convoy problems. It is nothing more than a red herring. An army does not need a convoy to move to an adjacent province; therefore (Rule VII.1) A Pic- Bel succeeds regardless of what happens to the convoy. Whether the convoy was wanted or not is also not relevant. Ruling otherwise would allow a player to take advantage of a "technicality" which is in fact not even applicable.

A good way of dealing with this is to only consider the convoy route when there is a non-disrupted convoy. This allows kidnapping the army for unwanted swap, but doesn't allow kidnapping the army to frustrate the move (as described by Rod Walker). A further recommendation is to require that the unit on the destination moves in the opposite direction. If played that way, in many cases the convoy doesn't need to be considered at all and it also eliminates issue 4.A.4 point 5 and 4.A.5.

In 1973 Edi Birsan played a tournament in Boston. As Italy he ordered an unwanted convoy for the French army in Marseilles to Piedmont. His unit in Piedmont moved to Marseilles securing the 18^th supply center needed for the solo win. Yet, he considered such unexpected move a loophole, harmful for the enjoyment of the game for new players and pursued a rule change after the event.

2. So, the 1982 rulebook was changed, but unfortunately the ambiguity was just made more explicit. "If an army could arrive at its destination either overland or by convoy, one route must be considered and the other disregarded, depending on intent as shown by the totality of the orders written by the player governing the army."

To fil in the gap, I would say that there are two situations where the convoy route must be chosen, (a) a fleet of the same power ordered the convoy (b) the unit at the destination of the convoy is of the same power and moves in opposite direction. Disadvantage of this rule is that swapping two units with a foreign fleet is in many cases not possible.

3. The 2000/2023 rulebooks introduce the tag 'via convoy' and read: "In some rare cases, orders are written so that an Army could arrive at its destination either by land or convoy. When this happens, the following qualifiers apply: • If at least one of the convoying Fleets belongs to the player who controls the Army, then the convoy is used. The land route is disregarded. • If none of the convoying Fleets belongs to the player who controls the Army, then the land route is used. However, the player controlling the Army can use the convoy route if he or she indicated 'via convoy' on the Army move order in question. This prevents foreign powers from kidnapping an Army and convoying it against its will."

The problem of the 2000/2023 text is that it is still ambiguous. What does the phrase "could arrive" mean? It can be one of three things: (a) there are fleets on the board that could convoy, (b) a convoy path is ordered or (c) a convoy path is ordered and it is not disrupted. Or in other words, if ordered with 'via convoy' but the convoy is not there, does it fall back to the land route?

4. The 2025EE rulebook fixes the ambiguity. There is no fallback to the land route: "An Army can swap places with a unit of an adjacent province by using a convoy. A convoy route by an Army to an adjacent province is not considered unless explicitly identified by that same Power in their Fleet order. A convoy route for the Army can be specified by identifying the Army order with 'via convoy'. In such a case the land route is disregarded.

Given the latest 2025EE rules, I prefer that there is no fallback to the land route.

The DPTG requires that convoying to an adjacent province is always tagged with 'by convoy' (it uses 'by' instead of 'via') even if own fleets are used. Note, compared with the 2025EE rules this only a difference in order writing and not in game mechanics.

For a future rulebook, I prefer the DPTG way. This the only place where an order is interpreted by looking at other orders (see also issue 4.E.3). By eliminating this, the text is simplified. The current rule optimizes two words away in order writing for a once in a life time experience, while adding the burden to remember something complicated for every game. The tactic is very specific and therefore the player won't forget the 'via convoy' tag. Finally, for an user interface that lists all the possible orders for a unit, it is rather strange that in certain cases a move with and without 'via convoy' is offered, while the move without 'via convoy' may still go via convoy.

Further note that in the research for this document very little information about this subject has been found (except for the 1973 event of Edi Birsan). Adjudicators written prior to 2000 likely do not support the 'via convoy' tag or something similar and even adjudicators written after, such as Backstabbr (as of 2024), allow an unwanted swap. Yet, no post or article report about a player that felt fooled by this trick. So, one can question whether the 'via convoy' cluttering in the order writing is worth it.

4.A.4. THE ORIGIN OF THE CONVOYED ATTACK

For some rules it is significant how the origin of the convoyed attack is interpreted. Does the attack come "from" the starting position of the army or the last fleet of the convoy? The house rules of War Bulletin in July 1971 state:

23. Brannan's rule. A convoyed attack comes from the direction of the last convoying fleet.

This interpretation has effect on five situations, but note that each can have its specific ruling:

1. Under Brannan's rule a convoyed attack on a fleet that is targeting the last convoying fleet, is a kind of head-to-head battle. Brannan's rule is to the advantage of the convoy (see test case 6.F.30).
2. With Brannan's rule some of the convoy paradoxes (but not all) are not paradoxes, because a support is not cut if the support is against the attacker (see issue 4.A.2).

The last three situations are related to convoying to an adjacent province:

3. If a unit is dislodged by a convoyed army from an adjacent province, can it retreat to that province (see issue 4.A.5)?
4. If a unit is dislodged by a convoyed army from an adjacent province, can it bounce a unit moving to that province (see issue 4.A.7)?
5. If the convoy is to an adjacent province and attacks a unit supporting an attack on the starting position of the army, then the question arises whether the support is cut (see test case 6.G.13).

Calhamer opposed Brannan's rule. Rod Walker writes in Diplomacy World No. 28 page 17, 1981:

... although the language of Brannan's Rule is deliberately omitted. It was Allan's intent that Brannan's Rule should not be used in Diplomacy adjudications, but no language forbidding it was ever inserted.

On page 19 it is clear that Rod Walker continued using Brannan's rule, while it was ten years after the 1971 rulebook was published:

3B. Behold Brannan's Rule
FRANCE: A Spa-Nap C by F Lyo & F Tyn and S by A Apu.
ITALY: F Nap-Tyn S by F Ion.
This is 3A without F Rom and with the Italian orders reversed. Here the attacks come out of and go into Tyn. Because under Brannan's Rule the attack of A Spa is coming from the direction of Tyn, the two attacks are stand-offs. This is a logical extension of Brannan's Rule and of the ruling made in Example 13. However, Allan Calhamer specifically disagrees with this ruling and would allow F Nap-Tyn to succeed, disrupting the convoy.

In his later years, this issue popped up again. On the 9^th of November 2001 a long thread was started by me on Usenet group rec.games.diplomacy about point 5:

I thought I had documented all difficult adjudication problems and then I see this on RedScape:

Austria:
A Trieste - Venice
F Adriatic Sea Convoys A Trieste - Venice

Italy:
A Venice Supports F Albania - Trieste
F Albania - Trieste

Is the support of Venice cut? Is the attack coming from Trieste (support is cut) or from the Adriatic Sea (support is not cut)?

The adjudicators:
Realpolitik: Support is not cut
Diplo: Support is cut
DipAdvisor: Support is cut
Warroom: Support is not cut

The conclusions between parenthesis are swapped (if attack comes from Trieste, the support is not cut). This was noted in a follow-up post.

On the 21^st of November Randy Hudson and Mike Lease made a post to the thread and gave the following arguments why "from" should be interpreted as the starting position of the army:

It's not phrased as a clarification of the "cutting support" rule (Rule X) in the 1976 rules. I've now printed out a copy of the 2000 rules from the Hasbro web site, and in the explanation, it offers the example:

France: A Tun-Nap, F Tyn C A Tun-Nap
Italy: F Ion-Tyn, F Nap S F Ion-Tyn

The explanatory text goes on to say that France could argue that support is cut (thus preventing the convoy from being disrupted), citing the rule, "Support is cut if the unit giving support is attacked from any province but the one where support is being given." IOW, since the army is coming from Tunis, it would normally be entitled to cut the support for F Ion-Tyn given by Naples, thereby preventing the convoy from being disrupted. This "new rule" (author's words) gives an exception to that rule, overriding the usual rule and eliminating the paradox. This rule WOULD NOT BE NECESSARY if the army were deemed to be coming from Tyn (the space being attacked). But the author says it IS NECESSARY ("...this rule takes precedence" [over the usual rule that would apply]), therefore, the army MUST be deemed to come from Tun, and this rule exists to provide an exception to avoid a paradox. But it only applies to a situation in which a convoy would be disrupted if the support is effective, but not disrupted if the support is ineffective. Therefore, for all other convoys, the ordinary interpretation should be that the army is deemed to come from the province in which it began the turn, and thus an army which can reach a province either via land or via convoy cannot cut support for an action against the province in which it started the turn, whether or not it is convoyed to the supporting unit's province. QED.

Stephen Agar sent this problem to Allan Calhamer. Before he got a response he followed up with the arguments of Randy Hudson and Mike Lease. On the 24^th of November, Allan Calhamer responded:

Good argument. I had always thought of the Army as coming from its province, not from the body of water, but your correspondent appears to show actual rule support for the proposition.
Best regards, Allan

For point 1, the case that a fleet at the destination of the convoy is attacking the last convoying fleet (test case 6.F.30) the consensus is (in agreement with Calhamer) that Brannan's rule should not be followed. None of the adjudicators have implemented Brannan's rule.

For point 2 Brannan's rule is consistent with the paradox rule of the 1971/2000/2023/2025EE rulebooks and Szykman's rule (issue 4.A.5), but covers fewer cases. In case the convoying army in a convoy paradox dislodges the supporting unit, it can be assumed that the paradox rule overrides the dislodge rule. Since Brannan's rule is more generic, the same assumption cannot be made. With a multi-route convoy multiple last fleets are possible. Since the attacked supporting fleet will only resist a cut from one of the convoying last fleets, it must be assumed that the support is cut by the other convoying last fleet(s). This outcome is consistent with the 2000/2023/2025EE paradox rule and Szykman's rule. Although Brannan's rule is consistent regarding convoy paradoxes the argument of Randy Hudson and Mike Lease is that wording of the rulebooks imply that Brannan's rule in general should not be followed.

For point 3 and 4 (issue 4.A.5 and issue 4.A.7) the more intuitive adjudication is consistent with the attack coming from the last convoying fleet. In the 2023/2025EE rulebooks point 3 has a specific rule that gives that outcome.

So, point 5 remains. Unlike point 3 and 4, there is no clear intuitive choice, it can go either way.

Given the dislike of Brannan's rule by Calhamer and the consensus on point 1, I think that in general Brannan's rule should not be followed and the attack comes "from" the starting position of the convoying army. If there is a desire to have a different outcome for a specific case, this should be covered by a separate rule (as is the case for point 2 and 3, and should be for point 4). For the case a convoy is to an adjacent province on a support that is on attack of the starting position of the army (point 5), I see no compelling reason to introduce or assume a separate rule. Therefore, I prefer that the support is not cut.

4.A.5. RETREAT WHEN DISLODGED BY CONVOY

In a very rare situation (see test case 6.H.11 and 6.H.12) a unit can be dislodged by a convoy from an adjacent province. Then the question arises whether the dislodged unit may retreat to the starting position of the convoyed army.

The following two interpretations are possible:

1. The unit may not retreat to the starting position of the attacker.
2. The unit may retreat to the starting position of the attacker when the attack was with convoy.

The 1971/1982/2000 rulebooks are not explicit on this. If the text is taken literally and the attack is considered coming "from" the starting position of the army (see issue 4.A.4), then such retreat is not possible.

The 2023/2025EE rules allows this with the text: "A dislodged unit can retreat to the position of the attacker when there is a convoy to an adjacent province."

It should be noted, that if played according to the 1971 rules with the recommended handling of issue 4.A.3 this situation cannot occur. The move would not take the convoy route or the departure province was contested by other units.

There is no reason to deviate from the 2023/2025EE rules. So, I prefer choice b.

4.A.6. CONVOY PATH SPECIFICATION

To convoy an army, only the departure and destination province need to be given. An alternative would be to specify the full route that the army should follow. Example:

The only reason to require such complete path is to prevent interference by an opponent with an "unwanted convoy". There are four types of them:

• An unwanted multi-route convoy to prevent convoying.
  This is only possible with the 1971 rules where a multi-route convoy is disrupted when one of the routes is disrupted (see issue 4.A.1). If a fleet is almost certainly dislodged, the player can anticipate a convoy of the enemy and give the fleet the order to convoy the army of the enemy to frustrate the convoy. See also test case 6.F.11.
• An unwanted multi-route convoy to take advantage of the paradox rule.
  This is only possible with the 1982 paradox rule (see issue 4.A.2). If a support of a fleet on another fleet is almost certainly cut by a convoy, the player can give the other fleet a convoy order to prevent dislodgment. See also test case 6.F.20.
• To kidnap an army.
  A player ordering a convoy of a foreign army to an adjacent province, to make it swap instead of a standoff. This to get the unit moving in the opposite direction (probably a fleet that is unable to convoy) behind enemy lines or just to make a joke (see test case 6.G.2). Another reason is to let the move fail by convoying with a fleet that will almost certainly be dislodged. Whether those tactics are possible depends on how issue 4.A.3 is handled.
• To let a convoy succeed that would otherwise fail.
  A situation is possible where a player is betrayed when a convoying fleet is dislodged. In such case the player might want that the convoy fails, because of a better defense position. The enemy can prevent this by ordering an alternative convoy route. See also test case 6.F.13.

The publisher never opted for the convoy specification in an update of the rulebook. Instead, it tackled the issue in other ways making the first three unwanted convoys not possible with the 2000/2023/2025EE rulebooks. When a computer program requires path specification then the players have no choice other than to comply. The judge, an email based automatic adjudication program written in the eighties (see interview with Ken Lowe in the Diplomatic Pouch the Zine, Spring 1995 Movement) was such program. However, for a mode with manually processed orders the situation is more nuanced. Even if the path is mandated, a move order without path can be viewed as a poorly written order that should still be followed. For instance, the order "A Yorkshire - Holland" can only happen via the North Sea, so the order is not ambiguous. These kinds of problems were already recognized in the early days. Eric Verheijden writes in Diplomacy World No. 5 page 22, 1974:

So much for a description of the problem. A number of solutions have been--or soon will be--proposed. All of them, to the best of my knowledge, involve some clarification of the convoy routing. Further, all of them involve either increased notation for the player to deal with, increased complications in the rules for the gamesmaster to deal with, or both. None of them are very palatable from my point of view--even my own. Few of them solve the problem completely and those that do, accomplish that feat by essentially insisting on route specification by the player in all cases in which it matters, in advance. Players being unfortunately lacking in precognition (although they may prefer it otherwise), the net effect would most likely be to disallow perfectly good convoyed attack, perhaps changing the course of the game, on the basis of clever but essentially technical plays used by opponents.

Verheijden continues with promoting the change that a multi-route convoy should only be disrupted when all routes are disrupted (which was adopted in the 1982 rulebook). See for this discussion issue 4.A.1.

In the Diplomatic Pouch the Zine, Fall 1999 Retreat, Manus Hand asked Allan Calhamer for an opinion. The discussion doesn't address the matter of order interpretation as mentioned by Verheijden, but is limited to unwanted convoys and the tactical implications of disrupting a multi-route convoy.

This issue can be handled in the following ways:

1. Path specifications are not allowed and if they are ordered then they are ignored.
2. Path specifications are allowed but not required.
3. Path specifications are required. The convoy fails if the order does not contain the path.

Requiring path specification is a burden on order writing for all convoys, while only making a difference for a handful of cases. With multi-route convoys only be disrupted when all routes are disrupted and convoying to adjacent province requiring 'via convoy', there is not much of rational left to require a path or to allow it. So, my preference is choice a.

4.A.7. AVOIDING A HEAD-TO-HEAD BATTLE TO BOUNCE A UNIT

The rulebooks say that if a unit is dislodged, then it has no influence on the province where the attacker came from. Of course, this is only significant when the units move in opposite direction. However, when two units move in opposite direction, they are not necessary engaged in a head-to-head battle. It is possible that one of the units convoys. The question arises, whether the dislodged unit can still bounce a third unit. See also test case 6.G.10, 6.G.14 and 6.G.15.

The following interpretations are possible:

1. A dislodged unit has never effect on the province where the attacker departed from.
   This is the literal interpretation of the rulebook with "from" considered the starting position (see also 4.A.4). It can be properly adjudicated, but the adjudication is confusing (see also 5.B.6).
2. A dislodged unit only has no effect on the province where the attacker departed from, when it was engaged in a head-to-head battle.
   From a mechanical point of view this makes more sense, because the unit that is being dislodged was engaged with the third unit, prior to the dislodgment. So, not bouncing the third unit is wiping out this battle from history.

Although choice a is more according to the rulebook, the idea that you battle a unit with equal strength and then not bounce it, is weird. Therefore, I prefer choice b. This choice is also consistent with choice b of issue 4.A.5, where we really look what is passing the border, instead of looking where the unit comes from.

4.B. MULTI-COAST ISSUES

4.B.1. OMITTED COAST SPECIFICATION IN MOVE ORDER WHEN TWO COASTS ARE POSSIBLE

If a move order of a fleet to a multi-coast province does not contain a coast where the fleet can move to both coasts, then the move should fail. See test case 6.B.1

Email judges in the early days of internet Diplomacy used a default coast. However, this principle was never used in face to face Diplomacy. Which coast should be taken?

4.B.2. OMITTED COAST SPECIFICATION IN MOVE ORDER WHEN ONE COAST IS POSSIBLE

If a move order of a fleet to a multi-coast province does not contain a coast where the fleet can only move to one coast, the following interpretations are possible:

1. A move is attempted to the only possible coast.
2. The move fails.

See test case 6.B.2

Omitting the coast can be regarded as a poorly written order which needs to be followed. So, choice a.

4.B.3. MOVE ORDER TO IMPOSSIBLE COAST

If a move order of a fleet to a multi-coast province does contain a coast but the specified coast is not possible. See test case 6.B.3

The following interpretations are possible:

1. The coast is ignored and a move is attempted to the possible coast.
2. The order is considered illegal.

I do not like the idea of changing an unambiguous order. Therefore, I prefer that the order is declared illegal (choice b).

4.B.4. COAST SPECIFICATION IN SUPPORT ORDER

Even the rulebook of 1959 allows that a fleet can support another fleet to a coast which it cannot reach (a fleet in Greece can support a fleet from the Black Sea to the east coast of Bulgaria). However, for a long time it was disputed whether the coast must be specified in the support order. The 1971/1982/2000 rulebooks do not clarify this.

Allan Calhamer comments on this issue in a post on Usenet rec.games.diplomacy (August 29^th 2002):

Stephen Agar wrote:

Allan

Could you spare the time to give your view on how to adjudicate the following situations - as you can see they all depend on how you treat the coast designations in each situation.

Turkey: F Con-Bul(ec)
Russia: A Rum S Turkish F Con-Bul(sc)
Austria: A Bul Holds

Turkey: F Con-Bul(ec)
Russia: A Rum S Turkish F Con-Bul
Austria: A Bul Holds

Turkey: F Con-Bul(ec); A Rum S F Con-Bul
Austria: A Bul Holds

Many thanks.
PS. The next issue of Armistice Day is in the post - hope you like the cover!

Regards

Stephen Agar

--
Dear Stephen: I would consider the supports in the first two cases no good, since I think allies should be in evident agreement on the moves. The support in the third case I would consider good.

Best regards, Allan

See test cases 6.B.7, 6.B.8, 6.B.9 and 6.B.15.

Since this comment it has slowly become the consensus that coastal specification in a support order must be possible. This view has also been adopted by the 2023 rulebook with the text: "The specification of Support to a specific coast in a split province such as Spain must match the move of the piece ordered to that province."

Still the text is not very clear what happens if the support order lacks the coastal specification. If the move is only possible to one coast, then the same ruling as issue 4.B.2 would make sense. If both coasts are possible, then the following could be ruled:

1. The support is illegal.
2. The support is only legal if the moving unit is of the same power (the choice of Calhamer).
3. The support is legal to both coasts.

For a program that offers the user the possible orders, I don't think it makes sense that a support order is offered without coast in addition to the orders with coast. For other modes of playing, allowing the coast to be omitted only when the unit is of the same power (Calhamer's position) is a kind of implicit order (see issue 4.E.3, because an order is interpreted by looking at another order in the same order set. For support and convoy orders, implicit orders are in general not allowed, so, it is inconsistent to allow it here. It is also rather harsh to invalidate an order when the coast is missing. Allowing the support order to proceed, introduces the possibility that a cunning player will exploit it, but this idea is rather farfetched. Therefore, I prefer that the order is legal for both coasts (choice c).

To express this in a future rulebook is a little bit tricky. One can think of a text like: "A support order for a fleet that can move to different coasts, must specify the coast and must match to be valid. However, if the coast specification is forgotten, then this is forgiven and the support is valid for both coasts." By expressing that the order must specify coast, programs should not offer the possibility to support without coastal specification.

4.B.5. WRONG COAST WHEN IT IS NOT REQUIRED

There are a number of situations where a coast specification is not necessary when a fleet is on a multi-coast province.

• The unit being ordered, see test case 6.B.10.
• The departure province in a move support order, see test case 6.B.17.
• The supported province in a hold support, see test case 6.B.16.

What to do if in such case the province is specified but incorrect? The following interpretations are possible:

1. The order is illegal.
2. Such coast specification is just ignored and the order is legal.

Mark Berch writes in Diplomacy World No. 20 page 10, 1978:

This situation is more complex, as a portion of the order is incorrect. However, disallowing the move seems an extremely serious penalty for giving some wrong (coastal) information that wasn't required, but is provided as a favor to the gamesmaster, the other players, and the general quality of the game.

I prefer that such garbage in the orders is just ignored (choice b).

4.B.6. UNKNOWN COASTS OR IRRELEVANT COASTS

How should an order be executed that contains an unknown coast, such as Spain west coast or Brest east coast. And how should an order be executed when the coast is just irrelevant, such as an army movement to the north coast of Spain?

The following interpretations are possible:

1. The order is illegal.
2. Such coast specification is just ignored and the order is legal.

I prefer that such garbage in the orders is just ignored (choice b).

4.B.7. COAST SPECIFICATION IN BUILD ORDER

When a fleet is built in multi-coast province (St Petersburg), the coast must be specified. The 1971 rulebook reads: "if Russia builds a fleet in St. Petersburg he must specify the coast on which it is to appear, or the build is invalid". Newer rulebooks have similar text.

4.C. UNIT DESIGNATION AND NATIONALITY ISSUES

4.C.1. MISSING UNIT DESIGNATION

When the designation of the type of a unit (A or F) is omitted, the following could be decided:

1. The order is illegal.
2. The order is still legal.

The 2025EE rules are explicit: "If you leave out the unit designation in an order, the order doesn’t fail since there can only be one possible unit in a province."

The 2025EE rules make choice b the preferred resolution.

4.C.2. WRONG UNIT DESIGNATION

When the designation of the type of a unit (A or F) does not match with the actual type of the unit, the following could be decided:

1. The order is illegal.
2. The order is still legal.

Mark Berch writes in Diplomacy World No. 20 page 10, 1978:

This is probably the most common error in writing orders; e.g. A Tri-Alb. The entire situation closely resembles #3, in that the labelling of the unit F or A is not required. Tri-Alb complies with the Rulebook, which requires only that you list the space each unit is in. Further, there is nothing sacred about those particular abbreviations. You could have A for 'Armada' and F for 'Footsoldier'. Or you could use just U for all units. True, these are not the Rulebook abbreviations. But then again, the Rulebook's sample game has 'Norw.' for Norway and 'Norw. Sea' for the Norwegian Sea. When was the last time you saw those in a Diplomacy magazine?

I think a best effort should be made to interpret the order, therefore I prefer that the order is still legal (choice b). This means that the unit type designation is in fact ignored. You can argue whether the unit type designation has any use. However, if other parts of the order are unclear, the unit type designation might help.

4.C.3. MISSING UNIT DESIGNATION IN BUILD ORDER

A player might forget to specify whether it wants to build an army or a fleet.

The rulebook is explicit that the unit only needs to be specified in coastal areas. The 2000 and later rulebooks reads: "When building a unit on a coastal province supply center, a Fleet or Army must be specified in the written build order." Earlier rulebooks have similar wording.

4.C.4. BUILDING A FLEET ON AN INLAND PROVINCE

A player might try to build a fleet in a province that cannot contain fleets. Such situation can be handled in the following ways:

1. The build fails always.
2. An army is built on the specified province.

First of all, I do not consider such order ambiguous. Second, if it has to be changed, there are alternatives. The player may want a fleet on another province or the player may want to decline the build instead. Since this cannot be concluded based on the given order, I prefer that the build fails (choice a).

4.C.5. MISSING NATIONALITY IN SUPPORT ORDER

When the designation of the nationality of supported or convoyed foreign unit is omitted, the following could be decided:

1. The order is illegal.
2. The order is still legal.

In Diplomacy World No. 26, page 5, 1980 Allan B. Calhamer is quoted:

He gave me an empathic no, pointing out that just giving the location of the unit was enough, to specify it.

Note, that there are adjudicators that cannot parse the nationality in support orders.

Specifying the nationality in support orders is a rule that is often ignored. I prefer that the order is just legal (choice b).

4.C.6. WRONG NATIONALITY IN SUPPORT ORDER

When the designation of the nationality of supported or convoyed foreign unit is incorrect, the following could be decided:

1. The order is illegal.
2. The order is still legal.

Note that there are adjudicators that cannot parse the nationality in support orders.

I think a best effort should be made. Therefore, the order should just be legal (choice b). This means that the nationality is in fact ignored. You can argue whether the nationality has any use. However, if other parts of the order are unclear, the nationality might help.

4.D. TOO MANY AND TOO FEW ORDERS

4.D.1. MULTIPLE ORDER SETS WITH DEFINED ORDER

If a game is played by email or by post, the Game Master may receive multiple order sets. If the latest order set contains an order for all units, then the latest order set replaces any previous order sets. However, if the latest order set is incomplete, it could be handled in the following ways:

1. All order sets are combined to one set of orders.
2. All order sets are combined to one set of orders, unless it is evident that the latest order set replaces all earlier sets.
3. Only the latest order set is considered, unless it is evident that earlier sets must still be taken into account.
4. Only the latest order set is considered.

I think the interests of players are best served when in principle a new set replaces an old set and only earlier sets are taken into account when the player made it clear on the order set (choice c).

4.D.2. MULTIPLE ORDER SETS WITH UNDEFINED ORDER

Also, in face to face games strange things can happen. When two sets of orders are submitted in a face to face game, then it is probably not clear which order set is the latest one. The following could be decided:

1. All units hold.
2. All order sets are combined to one set.

I think a best attempt should be made to interpret the orders. Therefore, I prefer that all orders sets are combined (choice b).

4.D.3. MULTIPLE ORDERS TO THE SAME UNIT

It might also occur that a player orders a unit twice or even more, with at least two orders that are not the same (if the orders are the same, then it should just follow that order).

First of all, it should be decided what should be treated as order. This is discussed in issue 4.E.1, which recommends that any illegal order is ignored. That means that if a unit gets a legal and an illegal order, then there are in fact no multiple orders and only the legal order should be handled.

The following interpretations are possible:

1. The first order is used.
2. The last order is used.
3. All orders to the unit are illegal.

The 2025EE rulebook is the first one to clarify this: "When a player has submitted multiple different orders for the same unit, the first order written is executed." So, choice a. Steve Froud criticized the wording on the Board Game Geek Diplomacy page on a thread started 20^th of January 2025 on the Rules Forum:

as a GM I can see a player using the wording of this rule to argue that about when (in time) an order was written; for example: "I wrote the one at the bottom of the page first, and the one at the top of the page was squeezed in at the last minute, so you should use the one at the bottom of the page". Then along comes an opponent who argues that "the first order written" means the one nearest the top of the page" (which I believe is the intent).

There is no reason to deviate from the 2025EE rules, so, the first order should be used. It looks like that the 2025EE rules follow the principle that orders are processed from top to bottom for all phases and earlier orders take precedence over subsequent orders.

Of course, a Game Master can have his or her own way of handling and look at the whole set of orders. Taking the order that is best written or best fits with the other orders (see issue 4.E.3).

4.D.4. TOO MANY BUILD ORDERS

It might occur that a player orders too many builds.

The following interpretations are possible:

1. All build orders are illegal.
2. The first legal build orders are used.
3. The last legal build orders are used.

See test case 6.I.1.

Denying France two builds in Winter 1901, because it ordered three, would be devastating. Allan Calhamer writes in Games & Puzzles No. 21, 1974:

The final problem of organizing a seven-person game, was not solved until I started studying Law in 1953. There I became aware that players who failed to meet their responsibilities towards the game should be made to suffer light penalties such as the loss of a single move; so they are encouraged to comply, but are not normally wiped out by minor lapses. The game should be designed so that it could charge right on in spite of poorly written orders and the like.

Again, the 2025EE rulebook is the first rulebook to clarify this: "If a player submits too many build or disband orders, the first orders are executed and the rest are ignored."

The 2025EE text is further evidence that the orders should processed from top to bottom for all phases. In case of conflict, earlier lines take precedence. So, I prefer choice b, in agreement with the 2025EE rules.

4.D.5. MULTIPLE BUILD ORDERS FOR ONE PROVINCE

Each province can have only one unit. This makes an order for building both fleet and army in one and the same province illegal.

The following interpretations are possible:

1. Both build orders fail.
2. The first build order is used.
3. The last build order is used.

Following the idea that orders are processed from top to bottom, I prefer that first build order is used (choice b).

4.D.6. TOO MANY DISBAND ORDERS

Comparable with the situation that a user orders too many build orders, a player can also order too many disband orders.

The following interpretations are possible:

1. All disband orders are handled according to the civil disorder rules.
2. The first legal disband orders are used.
3. The last legal disband orders are used.

See test case 6.J.1.

The 2025EE rulebook: "If a player submits too many build or disband orders, the first orders are executed and the rest are ignored."

The same arguments apply as for issue 4.D.4, So, I prefer choice b.

4.D.7. DECLINING BUILDS

The 1971 and 1982 rulebooks are ambiguous about declining builds. Two interpretations are possible:

1. Declining builds is allowed.
   This is according to the 2000/2023 rulebooks which explicitly specify that a build can be declined: "A country can decline to build a unit that it is entitled to for whatever reason (usually a diplomatic one)."
2. Declining builds is not allowed.
   The disadvantage of not allowing to decline builds is that additional rules are required when someone forgets to build

Note that if the game is played by email or by post, it may require that the build is explicitly declined. In this way the declining of builds can be distinguished from no orders received. This is not an issue in face to face games.

In older DATC versions and many articles this is called "waiving" a build. The current version follows the terminology of the latest rulebook.

With the 2000/2023 rulebook this is not really an issue anymore. So, I prefer choice a.

4.D.8. REMOVING A UNIT IN CIVIL DISORDER

1. The 1971 and 1982 rulebooks are identical on this item and read: "If a country in civil disorder has to remove units, because it has lost supply centers, the unit farthest from home (most distant from the nearest home supply center as computed by the shortest available route, including convoys) is removed first, the fleet before the army. If more units are equally eligible for removal than should be removed, priority is established by the names of the space in which they are located, the earliest in alphabetical order coming off first."

The problem of above rules is that the description of the distance calculation is ambiguous. Can fleets move like armies and vice versa? Does a convoy of an army count as one step? Different adjudicators implemented this differently. The DPTG specifies that distance for fleets shall be counted with legal fleet moves, while armies can also walk over water. This was also promoted by the DATC (up to version 2) and became the consensus to a certain extend. Although, it could be argued that this interpretation is overly complex without much added value.

2. It was reformulated in the 2000 rulebook: "If a country in civil disorder has to remove units, the units farthest from the country are removed first. If units are equally distant, then remove Fleets before Armies and then in alphabetical order by the provinces in which they are located."

This reformulation does not fix the ambiguity, but make it worse. Instead of counting to the home "supply center", it counts to the home "country". The country contains more provinces than just the home supply centers. For instance, Burgundy is part of France as country, but is not a supply center. This rule change was completely ignored by the community. Adapting the adjudicators would require a lot of effort, because the country information has no purpose and is not always available in the map data. Also, the change has only a negligible effect.

3. The 2023/2025EE rulebook fixes the issues: "In the event of a power going into civil disorder or when a power does not order a required removal, (a) No unit on a supply center is to be removed unless there are no viable options and (b) Unit precedence for removal starts with furthest from an owned supply center, by counting absolute adjacent provinces regardless of the ability of the unit to move into it, are removed first. Fleets are removed before armies, and then units are removed in alphabetic order of the name of the province on the map."

Note, that the 2023 rulebook not only takes away the ambiguity, but rules also significantly different. The distance is calculated to "owned" supply center, rather than "home" supply center. One could criticize this change, because it breaks with tradition. However, it will probably give some better play and it is more logical in variants where one can build in any supply center (such as Era of Empire). By this change, the rule will be the same for the standard version and these variants. And finally, the awkward unnecessary detail of army and fleet movements, is removed (note that by using "owned" supply centers, this becomes necessary, because a situation may arise that a power owns one inland supply centers while having two fleets).

The wording is unnecessary verbose. The part "(a) No unit on a supply center is to be removed unless there are no viable options" can be removed. Since any unit on a supply center in the adjustment phase, will make the supply owned and therefore make the distance zero, lower than any other unit.

Since the change with older rulebooks is significant, the test cases in section 6.J do not cover the rules prior to 2023.

If two distances are the same, then the alphabetical order of the provinces becomes significant. This raises the issue of the language that should be used. For instance, a dispute may arise when a non-English board is used, but English versions of conference maps and rules were downloaded and printed from the internet.

I prefer that the English language is used by default. However, if the game is played face to face with one or more physical boards (conference maps and rulebooks not taken into account) and those boards are all in the same language, then I prefer that the language of those boards be used. In such case, a list of English names might not be available. If the game is not played face to face and played completely in another language then English, then I still prefer that the English language is used. A computer program might have a user interface without any single English word, internally it still may use English names, for instance in a communication protocol. To avoid trouble here, the default language is English and the exception is only in the case of face to face games. Furthermore, in a game that is not face to face, the availability of a list with English names is probably not an issue.

4.D.9. RECEIVING HOLD SUPPORT IN CIVIL DISORDER

When a country is in civil disorder, all its units hold and do not support each other. But what should happen when another country supports the hold of one of the units?

The following interpretations are possible:

1. The support fails.
   This means that there is a difference between an unordered unit and a unit of a country in civil disorder. This has as consequence that a new issue is introduced. When is a country exactly in civil disorder? In a face to face game this might not be that clear!
2. The support succeeds.
   You may argue that this is just according to the official rules. Since a unit that is not ordered to move may receive support to hold.

I see no reason for introducing new issues for this rather insignificant issue. Therefore, I prefer that the support succeeds (choice b). Also, in a face to face game, a support to hold the armies, as an act of good will, might end the civil disorder.

4.E. MISCELLANEOUS ISSUES

4.E.1. ILLEGAL ORDERS

Mark Berch writes in Diplomacy World No. 38 page 18, 1984:

Can a unit ordered to move nonetheless be supported in place?

What a ridiculous question, you say. Berch isn't going to waste pages in DW when IX.6 clearly states, "A unit ordered to move may receive support only for its attempted movement. It may not be supported in place in the event that is attempted movement fails."

As it happens, however, many GMs believe that there is an exception to this rule, a circumstance where a unit ordered to move can be supported in place. How common this view is I do not know; my guess would be that a majority of GMs do hold to such a position.

In January 1984, Mark Larzelere ruled that a player who ordered F Mid-Hol (impossible), F Bre & F Por S F Mid did receive support for F Mid, so that the attack F NAt-Mid S by F Iri & F Eng failed. Larzelere stated, "Not all GMs would agree on how to rule regarding the French F Mid, whether it is 'ordered to move' and thus can't be supported in place, or was given an 'impossible' order, and is treated as holding."

So, the question is whether an illegal move is still considered a move and which orders are actually illegal. If an impossible move cannot be supported in hold, then such order can deliberately used to deny a support (although this is a farfetched tactic). The 1971/1982/2000 rulebooks don't clarify this, but the 2023/2025EE rules settle this issue:

• The 2023 rulebook reads: "A unit given an impossible order results in the unit HOLDING so it can be supported in place."
• The 2025EE has the same text, but on a different place it added: "Any order that cannot be interpreted by the Gamemaster and/or other players, or an order that is impossible (such "Army Delhi - Moon" or "Fleet Tokyo - Moscow") is ignored entirely.

In general, the terms "illegal" and "invalid" are not precisely defined or used in the rulebooks or other sources. In this document the following definitions are used:

• An illegal order is an order not in conformity with the rules. Legality is determined on the board situation only, and does not depend on other orders.
• An invalid order is an order that doesn't match with other orders and fails immediately for that reason.

Invalid orders are further discussed in 4.E.2.

Inspired on the 2023/2025EE rulebooks, legal orders are considered as restrictive as possible in this document. Meaning, anything that doesn't make sense is illegal. For instance, "A London - Holland" is illegal if there is no fleet in the North Sea. The advantage of this definition is that it is short and simple. A definition that allows more orders to be legal (and let them fail instead), would require more text. For instance, one would have to consider multi-coast issues. Ken Woodruff reports about PEERICON VII in Diplomacy World No. 48 page 50, 1987:

Earlier this year I attended the world renowned PEERICON Diplomacy tourney (June, 1987). One game was played, which was ultimately ruled a three way draw between Austria, Italy, and England. However, the game was fundamentally unbalanced, for Austria (Peery) managed to manipulate a technicality of order writing to his inestimable advantage, one result of which was the essential elimination of Turkey as a power.

The event sparked a heated debate with Turkey acquiescing only in order to 1) get the game over with before July, and 2) it's more respectable to lose on a "technicality" than to be legitimately smashed. Regardless of my "giving in" at the game board, I still feel that the decision made there was incorrect, and Mr. Peery graciously offered me an opportunity to "have my say" in the hallowed pages of DIPLOMACY WORLD.

Here are the positions of relevant pieces in Spring 1903:
AUSTRIA: Fleet Aegean Sea, Army Bulgaria, Army Serbia
TURKEY: Army Ankara, Army Smyrna, Fleet Constantinople

Now for the moves--
AUSTRIA:
     Army Bulgaria-Constantinople
     Fleet Aegean Support Army Bulgaria-Constantinople
     Army Serbia-Bulgaria
TURKEY:
     Fleet Constantinople-Bulgaria
     Army Smyrna Supports Fleet Constantinople-Hold
     Army Ankara Supports Fleet Constantinople-Hold

Now there is an obvious flaw in Turkey's order---there is no coast specified in the Fleet move. The rules have the following to say about such ambiguous orders:

"An order which admits of two meanings is not followed."

This means, in my interpretation, that an ambiguous or illegal order, regardless of its wording, is treated as no order at all! Just as "A Mos-Lon" is illegal, so is "F Con-Bul."

The Backstabbr (as of 2024) adjudicator allows orders such as "A Moscow - London" to be entered. However, when the fleet in Constantinople is selected, Bulgaria(ec) and Bulgaria(sc) can be clicked, but not the entirety of Bulgaria. Meaning, that if impossible orders are allowed, variation and therefore discussion still exists. Just considering illegal anything that is not proper avoids this and allows the supports for Constantinople to succeed.

The next question is whether an illegal order should be changed in a hold order, or should be ignored. Consider the following orders:

The Germany player intended to write "Belgium - Burgundy" for the first line, but accidentally wrote "Holland". If this order is changed in a Hold, then indeed Holland will Hold. However, if the order is ignored, then the only remaining order for Holland is the support for Kiel. Changing something in Hold, means that one has to distinguish between scribbles that resemble something of an order and leads to a Hold and scribbles that cannot be interpreted at all. A rather mood thing for something with such subtle implication. Ignoring everything that isn't a proper order, avoids further debate. If a unit does not have an order at the end, it will default to a Hold.

This approach can also be applied to the retreat phase or adjustment phase. Suppose Russia has the right to build only one unit, but orders two:

According to the 2025EE rules, the first build should be taken (see issue 4.D.4). However, the first order cannot be executed, because of the missing coast specification. The question is whether the build order still occupies a slot and no builds are performed, or the Moscow build is executed. The principle of going from top to bottom and skipping anything that makes no sense, comes to the rescue. The first build order is ignored, and the Moscow build order will proceed. See also my article in Diplomacy World No. 170 page 51, 2025.

An adjudicator with an interactive user interface, may only show the legal moves where the player can choose from. In such case, the tactic of denying a support by an impossible order, cannot even be entered. If the system does allow to enter illegal orders, then the adjudicator can still allow a Hold support on an impossible move, in accordance with the 2023/2025EE rules.

In the variant where players cannot communicate with each other ("no-press" variant) then it is very significant which orders are accepted to be published. If this is restricted to legal orders (but this is not a must), then this limits the communication possibilities between the players via orders.

There are a few edge cases related to illegal orders. First of all, the order "F Black Sea Supports A Rumania - Armenia" is illegal. Although the Black Sea can support an action in Armenia it cannot convoy the Rumanian army at the same time. See test case 6.D.31.

Secondly, for a convoy order, the fleet should be part of at least one route for which the fleets are available. An order like "F North Sea Convoys Sevastopol - Ankara" makes no sense and is illegal. We can even be more restrictive and require that a convoying fleet should not be superfluous. There should be at least one route, whether ordered or not, for which the fleet is necessary for the convoy. For instance, for a convoy from Tunis to Albania one can imagine the route "Tunis - Ionian Sea - Adriatic Sea - Albania", but the Adriatic Sea can just be left out. Therefore, the order "F Adriatic Sea Convoys Tunis - Albania" is considered illegal in this document (an interactive system showing only legal orders, should not show this order). The path finding algorithm in the adjudicator can implement this by only considering proper routes. In a proper route the convoy will never jump to a province that could be reached from an earlier node in the chain. The jump from Adriatic Sea to Albania would not be considered, because Albania could be reached from the Ionian Sea earlier in the chain. See also test case 6.G.19

I prefer that legal orders are as restrictive as possible and that illegal orders are ignored entirely. This gives the most straight forward way of processing orders.

4.E.2. RESOLUTION TERMINOLOGY

When the orders are submitted, the adjudication has been executed, then it only matters which moves succeed and which do not. Any adjudicator output about support cuts, disrupted convoys, standoffs etc. is purely for the understanding of the players. How this output should look like has never been standardized nor has much been written about it. For this reason, the adjudicator developer has all the freedom as long as it is clear whether a move fails or succeeds. Still, there are a number of things to mention and the developer might think about it before starting to program. Adjudication output that is clear and to the point will certainly be appreciated by the players.

Rulebooks had the convention to underline failed orders. The zines used this practice and extended it further. The order of a dislodged unit was tagged with /r/ or /d/ which stands for retreat and dislodged respectively. If a unit could not retreat it was tagged with /a/ for annihilation. The term "annihilated" was used in the 1959 rulebook, but dropped in the 1971 edition (see discussion below regarding units without retreat options). The abbreviation "nso" for "no such order" was sometimes used for non-matching orders. Email judges in the early days of internet diplomacy were limited to ASCII and did not have the possibility to underline text. More verbose output could easily be generated, however. In modern user interfaces also colors can be used, such as red for a failed order. Finally, the 2025EE rulebook dropped the convention of underlining.

The following resolution related (but incomplete) terminology can be found in the rulebook:

• Legal/illegal orders
• Valid/invalid orders
• Standoff (or bounce) for a move order
• Cut of a support order
• Disruption of a convoy order
• Dislodgment

For legality, see the previous issue 4.E.1.

Validity is mentioned in the rulebook, but not precisely defined. In this document an invalid order is legal, but does not match.

A standoff is not a synonym for a failing move order. There is only a standoff when the move fails and the opposing unit did not move either. If the other unit succeeded to move, then one might call it a "defeat". If the move fails due to a countryman unit holding, then this is in general not called standoff. Because of these details, the adjudicator developer might opt to avoid the term "standoff" in the output or only used it in the secondary explanation of a failing move.

A failed support order has the consistent terminology of "cut". For a support order that is not cut it is less clear. Marking such order as "succeeds" might confuse some if the supported unit fails. In the test cases of this document "given" is used instead.

The situation is similar for convoy orders. A failed convoy order is "disrupted". However, a non-disrupted convoy might be idling, instead of participating in a convoying operation. The convoy can be disrupted on a different place in the chain or the army did not order the convoy. Therefore, the term "available" is used in the test cases. That is choice made in this document and not based on any prior art.

The term "dislodgment" is used consistently for any unit that must retreat. In the 1959 rulebook the term "annihilated" was used for a dislodged unit with no retreat options. However, the rulebook did not address the possibility of voluntarily giving up a unit when retreat is possible. Since the 1971 rulebook, the term "annihilated" was dropped and replaced by the more generic term "disbanded" which applies to both voluntary and involuntary removals, whether in the retreat phase or the adjustment phase. Still, it remains useful for players when adjudicator output makes it explicitly clear that no retreat options are available. For this reason, this document uses the tag "destroyed".

Summarized, this document uses the following resolution terminology in the test cases:

Hold

• dislodged
• stands

In the 1959 Rulebook the term 'stand' was used instead of 'hold'. From 1971 the rulebooks standardized on the term 'hold'. Yet, using a different term specific for the situation of resolution, helps fast reading the results.

Move

• illegal, move is not possible
• invalid, a convoy is required, but no convoy path is ordered
• fails, opposed by superior or equal forces or required convoy is disrupted
• succeeds

Support

• illegal, destination cannot be reached, or supported move is illegal, or supported move requires this unit to be part of the convoy
• invalid, the support doesn't match
• cut
• given

Convoy

• illegal, see issue 4.E.1
• invalid, when the convoy doesn’t match with move order. There is no check if there is a convoy path ordered. So, the move order might be invalid, while the convoy order is valid. This is a choice made in this document
• disrupted
• available

The order of a unit that is dislodged, gets the additional tag dislodged or destroyed when there is no retreat option. In the case of a Hold order this is only mentioned once (output like dislodged, dislodged is silly).

4.E.3. IMPLICIT ORDERS

An implicit order is an order that is deduced from a support order or convoy order in the same order set. For instance, when Germany orders its army in Ruhr to support a move from Kiel to Holland, then the move order from Kiel to Holland can be deduced as implicit order. It can be decided that:

1. Implicit orders are allowed.
   Note that an explicit order takes precedence over an implicit order. And an implicit order can never be deduced from another player's order set.
2. Implicit orders are not allowed.

This issue is part of a more general concept. Can a poorly written order be clarified by orders in the same order set, or should every order be interpreted individually? Game Masters have different opinions about this. In Diplomacy World No. 135 page 42, 2016 Matthew Shields writes:

Should a GM consider the player’s entire order set when interpreting a single unclear order? I personally would say yes. The other orders can offer useful context, especially in cases a word is simply very hard to read, but otherwise the orders are internally consistent.

In Diplomacy World No. 136 page 5, 2016 Tim Haffey reacts:

Matt seems to suggest that individual orders should be viewed and interpreted as a set of orders. Not really. An order is for one individual piece an each has its own order. If they don't work together like they might, it is not the job of the GM to try and read the player's mind and think: "hmmm, now what is the player trying to do here".
...
Good lord, GM correcting errors, reading minds, interpreting intentions? What is Matt thinking? We adjudicate the individual moves as written without mind reading, interpretation or correcting errors.

In case the game is played with a Game Master, it is up to him or her how the game is led. If played without Game Master, looking at the whole set opens an extra dimension of discussion and players might even try to take advantage of it. Therefore, I prefer that in such case orders are interpreted individually (choice b).

4.E.4. PERPETUAL ORDERS

When a player wants to quit the game, he maybe wants to give "Perpetual Orders". That are orders that will be repeated every turn. For instance, when Italy still has two armies in Portugal and Spain, he may want to give the order that they will mutually support each other as long as it is possible. It can be decided that:

1. Perpetual orders are allowed.
2. Perpetual orders are not allowed.
   Since the rulebook contains special rules for a country in civil disorder, it can be argued that perpetual orders are not allowed according to the rules.

Although I do not really object against perpetual orders, I think allowing perpetual orders should be decided before the game is started or all players should agree during the game. If it was not explicitly allowed and a player wants to give perpetual orders and another player opposes, then I prefer that it is not allowed (choice b).

4.E.5. PROXY ORDERS

A "Proxy Order" is an order that one does not order a specific unit by oneself, but that another specific player may give the actual order to the unit. I can be decided that:

1. Proxy orders are allowed, where the player should notify the Game Master that it wants to proxy a unit. This notification is before the submission of the orders and the player that receives the proxy has knowledge about it.
2. Proxy orders are allowed where a proxy order is given as part of the normal order set.
3. Proxy orders are not allowed.

It should be realized that allowing proxy orders is not just an optimization of the negotiations between the players without a material effect on to the game. A player allowed to order a unit of a foreign power may move the unit out of position creating the conditions of a stab. Something that cannot happen if proxy orders are not allowed. A dominant player may demand a proxy order in negotiations and treat a refusal as distrust. Proxy orders allow the orders to be sold and this changes the dynamics of the diplomatic relations. Allan Calhamer wrote once about multi-coast support orders: "I think allies should be in evident agreement of the moves." (see for reference issue 4.B.4). With proxy orders such evident agreement is not enforced.

Of course, the players are totally free to allow proxy orders for their game. However, I think it is fair to say, that it is not according to the rules and it is not a neutral change. Therefore, I prefer that without prior agreement they are not allowed (choice c).

4.E.6. FLYING DUTCHMAN

A "Flying Dutchman" is a unit on the board that is illegal. This can be due to an adjudication error or cheating. Since it is not possible to replay previous rounds, the situation has to be corrected on the board. There are numerous ways to handle this situation (for instance, just play until the next adjustment turn).

Allan Calhamer writes in Diplomania, No. 12, August 1966:

A variety of rough-and-ready tactics were developed at this time. One was the "Flying Dutchman", which consisted in playing with a piece to which you were not entitled. It was ruled that this practice was legal so long as it was a deception; i.e., any player had a right to demand restoration of the true position, but if moves had intervened, they could not be taken back. It was never clear what the rights were if the deception was discovered during a move. Players quickly learned not to challenge a "Flying Dutchman" unless its removal was in their interest; sometimes a player might let one survive for several moves, then challenge it when alliances began to shift. The easiest way to put on a "Flying Dutchman" is to raise one when other players are raising, or to "forget" to drop one; but they are sometimes just placed on the board when only yourself or allies are present. Also, pieces have been advanced or pushed back, armies have turned into fleets, and so on.

I would say that the Flying Dutchman is immediately removed when discovered. If the unit being the Flying Dutchman can be identified, it is removed otherwise the Civil Disorder rules are used. The last adjudication can still be appealed, any prior adjudications cannot. In contrast with the story of Calhamer, I think that adding a Flying Dutchman to the board deliberately is considered cheating nowadays.

5. THE PROCESS OF ADJUDICATION

The article "The Math of Adjudication" of the Diplomatic Pouch the Zine, Spring 2009 Movement contains additional information with more examples. However, this chapter was rewritten for version 3.0 of the DATC and is more recent than the article.

Writing a Diplomacy adjudicator program may look not more difficult than writing a program that checks the moves of a chess game. However, the contrary is true. A Diplomacy adjudicator that passes all test cases as described in this document contains many small and difficult details.

The first thing people think about when analyzing the rules of Diplomacy is the sequence in which the orders are processed. However, this is exactly that should not be done. None of the rulebooks describe or hint about the sequence in which the orders must be adjudicated. The rules are a set of "equations" or "conditions" that should be fulfilled. The first step is to make these equations more precise, more mathematically.

A key feature of these equations is that the result during adjudication will not change. An order succeeds or fails and that does not change during adjudication. However, when the adjudication is started its state is "unknown" or "unresolved". This in contrast to principles like a support is successful until it is cut or a move order succeeds until it bounces. Going that way makes writing a correct adjudicator very difficult.

When the equations are understood, the next step is to make an algorithm that finds a solution.

5.A. OVERVIEW OF DIPLOMACY EQUATIONS

There are three active orders that should be adjudicated in success or fail. The result can be kept in a Boolean, were success is True and fail is False.

• MOVE
• SUPPORT
• CONVOY

A unit is dislodged when it doesn't move and another unit successfully moves to the province. This doesn't need further explanation.

A move order has a PATH when the destination can be reached directly or via a non-disrupted convoy.

There are a number of different strengths that need to be calculated during adjudication. They often result in the same value, but may differ in some circumstances. They should not be confused.

• ATTACK STRENGTH
  For each unit ordered to move, the strength to attack and conquer the province moved to.
• HOLD STRENGTH
  For each province on the board the strength to prevent that a unit moves to that province. If the province is empty, the value is zero.
• DEFEND STRENGTH
  The strength of unit ordered to move engaged in a head-to-head battle to prevent that the opposing unit succeeds.
• PREVENT STRENGTH
  The strength of unit ordered to move preventing another unit ordered to succeed moving to the same province.

Typically, if a move is not engaged in a head-to-head battle the ATTACK STRENGTH must be greater than the HOLD STRENGTH of the province and the PREVENT STRENGTH of each of the units competing for the same province, to be successful.

If multiple units are ordered to move to the same province, then each order has its own success condition. So, there is not a mechanism to declare a grand winner of the contested province.

The success of the MOVE orders of Berlin, Warsaw and Bohemia are determined separately. The orders of Warsaw and Bohemia fail, because their ATTACK STRENGTH is one, while the PREVENT STRENGTH of Berlin is two. The MOVE order of Berlin succeeds, because the ATTACK STRENGTH is two and that beats the PREVENT STRENGTH of both the units in Warsaw and Bohemia.

Similarly, in the case of a head-to-head battle the MOVE of both units is determined separately.

In this example Germany probably reversed the support order in Munich by mistake. The MOVE of Berlin will fail, because the ATTACK STRENGTH is one, while the DEFEND STRENGTH of Kiel is two. The MOVE of Kiel will also fail, because the ATTACK STRENGTH of one is insufficient to beat the DEFEND STRENGTH of Berlin also one. Since Munich may not help in dislodging a unit of the same nationality, but may help in defending, the ATTACK STRENGTH and DEFEND STRENGTH of Kiel are different.

5.B. PRECISE DESCRIPTION OF THE EQUATIONS

With the general idea of the different terms given in the previous section, the conditions can be further detailed out. In these details we refer to the result of other orders and ignore the problem of the sequence in which the orders must be adjudicated. That will be discussed later. Note, that various rule issues, such as coasts or pre-scanning the orders are not covered here. They are not relevant for the algorithm.

5.B.1. MOVE

• In the case of a head-to-head battle, the move succeeds when the ATTACK STRENGTH is larger than the DEFEND STRENGTH of the opposing unit and larger than the PREVENT STRENGTH of any unit moving to the same province. If one of the opposing strengths is equal or greater, then the move fails.
• If there is no head-to-head battle, the move succeeds when the ATTACK STRENGTH is larger than the HOLD STRENGTH of the destination and larger than the PREVENT STRENGTH of any unit moving to the same province. If one of the opposing strengths is equal or greater, then the move fails.

5.B.2. SUPPORT

A support order is cut (fails) when another unit is ordered to move to the province of the supporting unit and the following conditions are satisfied:

• The moving unit has a successful PATH.
• The moving unit is of a different nationality.
• The destination of the supported unit is not the province of the unit attacking the support.

Or:

• The unit has a successful MOVE (dislodging the supporting unit).

5.B.3. CONVOY

The convoy order is successful if the fleet is not dislodged. So, no other unit has a successful MOVE to the province.

5.B.4. PATH

The PATH of a move order is successful when the unit can directly move to the destination and is not convoyed or is convoyed and there is a chain of adjacent fleets from origin to destination each with a matching and successful CONVOY order.

5.B.5. HOLD STRENGTH

• The HOLD STRENGTH is defined for a province, rather than for an order.
• The HOLD STRENGTH is zero when the province is empty, or when it contains a unit with successful MOVE order.
• It is one when the province contains a unit with a failed MOVE order
• In all other cases, it is one plus the number of units support it in hold with successful SUPPORT order.

5.B.6. PREVENT STRENGTH

• If the PATH of the move order fails, then the PREVENT STRENGTH is 0.
• In cases where the move is part of a head-to-head battle and the MOVE order of the opposing unit is successful, then the prevent strength is 0.
• In the remaining cases the PREVENT STRENGTH is 1 plus the number of units with a successful SUPPORT order.

5.B.7. DEFEND STRENGTH

The DEFEND STRENGTH of a unit with a move order is one plus the number of units supporting the move with a successful SUPPORT order.

5.B.8. ATTACK STRENGTH

• If the PATH of the move order fails, then the ATTACK STRENGTH is zero.
• Otherwise, if the destination is empty, or in a case where there is no head-to-head battle and the unit at the destination has a move order for which the move is successful, then the ATTACK STRENGTH is one plus the number of units supporting the move with successful SUPPORT order.
• If not and the unit at the destination is of the same nationality, then the ATTACK STRENGTH is zero.
• In all other cases, the ATTACK STRENGTH is one plus the number of units supporting the move, of different nationality as the unit on the destination and with successful SUPPORT order.

5.B.9. CIRCULAR MOVEMENT AND PARADOXES

It is possible that for an order set there is no solution for given condition or that there are multiple solutions. This is only possible when there is a circular dependency between the different conditions. That happens with circular movement or with a convoy paradox.

However, a circular dependency may also have just one solution. For instance, in the case of a circular movement, if one the unit moves for sure because of a support or a unit won't move because of a bounce, the circular movement has only one solution. Similar, a convoy paradox might lose its paradox status if an additional support or bounce is added. If there is only one solution that solution must be taken.

In the case of a circular dependency that has zero or two solutions the orders that are in the circular dependency must be examined. If they only consist of MOVE orders, then it is a circular movement. All MOVE orders become successful. If there are any CONVOY orders in the cycle, then there is a convoy paradox. The Szykman rule must be applied by setting the CONVOY to fail (ignoring the original condition for CONVOY order).

In case the adjudicator can also handle variants, other paradoxes may apply that do not fulfil above conditions. In such case it is best to let all MOVE and SUPPORT orders in the cycle fail.

5.C. FROM CONDITIONS TO ALGORITHM

The next step is to transform the conditions as described in the previous sections to an algorithm. When finished, the algorithm can be tested with the test cases. Note, that if a bug is found, then it is unlikely to be a problem of the conditions described in the previous section. They have proven to be stable from the first version.

The adjudication program needs to handle the following situations:

1. An order that is not indirectly dependent on itself.
2. An order that is indirectly dependent on itself, but there is still exactly one resolution.
3. An order that is indirectly dependent on itself and there are zero or two resolutions.

If it was only needed to adjudicate orders of category a, the algorithm would be simple. Just a recursive function that adjudicates an order and if it depends on a different order, it will call itself recursively. This also shows that it is not needed to be concerned about the sequence of orders. Rule of thumbs such as cut support first do not need to be programmed. It is also not difficult to detect whether the recursive hits a cyclic dependency. Applying the backup rule to the cycle will also not give too much trouble. The real challenge is to determine whether the cycle has just one solution as in category b (that solution should then be taken) or zero or two solutions as in category c (the backup rule should be applied).

There are two approaches to this problem:

• Making decision on partial information.
• Guessing different resolutions.

Both can lead to a correct algorithm. The difference can be best explained by an example:

This circular movement has only one solution, because MOVE of Ankara to Constantinople is guaranteed to succeed, due to the support. And this is exactly how an algorithm based on partial information works. Even if the MOVE order of Constantinople to Smyrna is still uncertain, it can be concluded that the HOLD STRENGTH of Constantinople is zero or one, but not higher. The ATTACK STRENGTH of Ankara to Constantinople is two and will always beat this allowing the MOVE to succeed. When this is concluded, the other MOVE orders can also succeed.

The guessing algorithm just makes a guess for one of the orders. For instance, it guesses that the MOVE from Smyrna to Ankara fails. As result the MOVE from Constantinople to Smyrna fails, but the MOVE from Ankara to Constantinople still succeeds. If then the MOVE of the Smyrna is adjudicated it is concluded that it succeeds which is inconsistent with the initial guess. The guess can be repeated with a successful MOVE and then a consistent adjudication is obtained.

For both approaches it is suggested to make two mutual recursive functions (in Python code):

Both functions take an order reference as input (in other languages this can be an index or pointer) and return the success or failure of the order as Boolean.

The 'adjudicate' function implements the conditions as discussed earlier. The function can be split up in adjudicate functions for MOVE, SUPPORT and CONVOY and separate functions for PATH and STRENGTH values. If the result of another order is required, then it will call the 'resolve' function. The 'adjudicate' function will not update any administration.

The 'resolve' function is a generic function and has no knowledge about the details of the orders. It updates the administration when an adjudication becomes final, prevents that the same order is adjudicated twice, detect cyclic dependencies and applies the backup rule if needed.

The adjudicate function will be roughly the same for both approaches, while the resolve function will be fundamentally different.

The final program is then just calling resolve for each order to ensure that every order is resolved.

5.D. THE TROUBLE WITH PANDIN'S PARADOX

For solving circular dependencies, it would be great if each order is part of at most one cycle. However, in convoy paradoxes, it can be complex. Consider Pandin's paradox:

The cycle of dependencies looks like this:

The arrows are in the direction of going deeper in recursion. To decide whether the convoying fleet in the English succeeds it is needed to know whether the fleet in Wales or Belgium succeeds. For calculating the success of Wales and Belgium, the result of the SUPPORT of London is required. And the SUPPORT of London depends on the successful CONVOY of the English Channel. Note, the move order of the army in Brest is a key ingredient of this paradox, but is not listed in the cycle. The question is whether Brest will cut the SUPPORT of London, but that is not dependent on the success of the Brest order (in fact, whatever paradox rule is used, without additional support the Brest MOVE will fail).

The fact that there is not a single cycle, makes the algorithms complicated. In case the algorithm can make decisions on partial information, then it can conclude that the unit in Wales will never succeed, because its ATTACK STRENGTH will never be enough to beat the PREVENT STRENGTH of Belgium. However, the algorithm should ensure that all decisions on partial information are finished, before acting on cycles.

For a guessing algorithm these situations are very difficult to handle. For instance, it should not start making a guess on Wales, because that order is not a key decision that influences the paradox.

There is a simple hack that ensures that only clean single cycle can happen. The adjudicate function for CONVOY would normally call resolve for any units attacking the convoy. That make the success of the fleet in the English Channel dependent on Wales and Belgium in above example. However, if the adjudicate function is called instead, the dependency is skipped and all cycles will be clean and simple. In the above example the cycle will only consists out of English Channel and London. If the Szykman rule is applied on that cycle, the English Channel will fail to convoy and all other orders can be resolved.

5.E. THE PARTIAL INFORMATION ALGORITHM

If the success or failure of any dependent order is still uncertain it might be needed to adjudicate an order twice. This can best be illustrated by an example:

Suppose furthermore that the SUPPORT of Belgium cannot be determined yet. To see whether the North Sea MOVE succeeds, the ATTACK STRENGTH is calculated. This may not exceed 1 and this is insufficient to dislodge Holland. However, this absence of success does not imply that the order fails. This can separately be checked.

In earlier descriptions of this algorithm the failure and success of an order in the adjudicate function was implemented separately, doubling the code. However, there is a neat trick to avoid that. Both adjudicate and resolve functions get an additional parameter:

If the optimistic parameter is True, then any uncertain information is assumed to have a value that is supportive for letting the order succeed. In the pessimistic scenario (the parameter is False) the opposite. STRENGTH values will return what the strength will be at most in the optimistic scenario and the least in the pessimistic scenario.

The adjudicate function will not check on the optimistic parameter, but just pass it to its subfunctions and to the resolve function, but sometimes inverse it. For instance, for a MOVE adjudication the ATTACK STRENGTH must be calculated. In the optimistic scenario the optimistic ATTACK STRENGTH is calculated but for the STRENGTH values of any opposing unit the pessimistic value is used. Similar, in the optimistic scenario of a SUPPORT or CONVOY order the pessimistic result is used for any unit that can let the order fail.

The resolve function (with cycle detection, but without the backup rule) becomes something like this:

As you can see the resolve function does not just pass the optimistic parameter to the adjudicate function. It will try both optimistic and pessimistic scenarios and if they are in agreement the order is resolved.

In the case where there is only one resolution of the situation, then the resolve function will always resolve the order (although, the resolution of any depending orders may still stay open). To add the backup rule for cyclic movement and the Szykman rule, we need to track which orders are in the cycle when no resolution could be found. We do this by keeping those orders in a global array 'cycle'. Example code:

If the cycle is not a clean single cycle, it may fail in general, although not with the standard rules. If the resolve function is called on Wales as in the situation of 5.D, it will call in recursion, London, English Channel, Wales, Belgium and London before retreating from recursion. Then it will conclude that there is a convoy paradox when it is back by the London order. But at that moment, the Wales order is still undecided, while it could be decided by partial information. With the standard rules this still goes right, however, in the case of variant rules things can be different. A way to fix this is to retreat in recursion up to the order that is the ancestor of the whole cycle (Wales in the example). This can be achieved by adding a global integer 'recursion_hits' that keeps track how many times the recursion hit previous visited orders:

This is a very robust way of handling, especially in variants, since all orders that can be decided on partial or full information will be decided that way (as players expect).

The disadvantage of this algorithm is that all orders are adjudicated twice, in optimistic and pessimistic mode. With modern powerful computers, this shouldn't be a problem, except when it is used in an AI engine. A simple optimization is to skip the pessimistic adjudication when the optimistic adjudication fails. In such case the pessimistic adjudication is guaranteed to fail also. A further optimization is to track whether the adjudication of an order was fully dependent on resolved orders. In such case, one adjudication suffices. This can be implemented by adding another global variable 'uncertain'. This variable is set to True if a dependent order does not have a resolution yet and kept unaltered otherwise.

One should be reminded that one adjudicate call may result in multiple resolve calls for the same order. For instance, for MOVE order the ATTACK STRENGTH is calculated and the HOLD STRENGTH of the destination. Both STRENGTH values depend on the MOVE on the unit on the destination. Especially in a cyclic movement, where orders cannot directly get a permanent resolution, there is a risk that one gets an exponential explosion. If the cyclic movement consists of 10 moves, they may lead to 2 to the power of 10 (1024) adjudications. In above resolve function this is prevented by the check 'if order in cycle:'. One could also program the adjudicate function in such way that double calls to resolve are not made.

5.F. THE GUESS ALGORITHM

In the guess algorithm different resolutions are tried and checked on consistency. The difficulty is to do this in a simple recursive function. The trick is to set the guess value when an order is visited for the first time and use this guess value when this order is visited again deeper in the recursion. If this happens orders are again added to the global variable 'cycle' when returning from the recursion.

Again, this algorithm is problematic when there is a cycle that is not a clean single cycle. This can be fixed with the hack described in section 5.D. There is no straight forward way to fix this in the generic resolve function in theoretical correct way. However, if the algorithm retreats to the situation where the order is the ancestor of the whole cycle, then in practice it will work.

If the resolve function is called on Wales as in the situation of 5.D, then the result will be negative, whether it guesses positive or negative, although the remaining paradox is not tried in two ways. If the resolve function is called on Belgium, London or English Channel, then the paradox is detected and Wales is added (incorrectly) to the cycle. However, the backup rule won't touch this order, so it still goes right. For implementation, again a global variable 'recursion_hits' is added. Additionally a Boolean variable 'guess_based' is necessary, because looking whether the cycle has increased is not correct if multiple cycles are possible.

6. TEST CASES

With some scripting the test cases can be extracted from the HTML source. Each test case is in a <div> tag of class "test-case". The tag has an additional class "spring" or "winter" which describes the phase of the test case. Test cases for retreats are always part of a Spring test case. The id of the <div> tag contains the test case number followed by ".orders".

6.A. TEST CASES, BASIC CHECKS

6.A.1. TEST CASE, MOVING TO A PROVINCE THAT IS NOT A NEIGHBOUR

Check if an illegal move (without convoy) will fail.

6.A.2. TEST CASE, MOVE ARMY TO SEA

Check if the adjudicator allows an army to be moved to open sea.

6.A.3. TEST CASE, MOVE FLEET TO LAND

Check if the adjudicator allows a fleet to be move to an inland province.

6.A.4. TEST CASE, MOVE TO OWN PROVINCE

Moving to the same province is an illegal move (2023 rulebook, page 7, "An Army can be ordered to move into an adjacent inland or coastal province.").

Program should not crash.

6.A.5. TEST CASE, MOVE TO OWN PROVINCE WITH CONVOY

Moving to the same province is still illegal with convoy (2023 rulebook, page 7, "Note: An Army can move across water provinces from one coastal province to another...").

The move of the army in Yorkshire is illegal. This makes the support of Liverpool also illegal and without the support, the Germans have a stronger force.

6.A.6. TEST CASE, ORDERING A UNIT OF ANOTHER COUNTRY

Check whether someone cannot order a unit that is not his own unit.

6.A.7. TEST CASE, ONLY ARMIES CAN BE CONVOYED

A fleet cannot be convoyed.

6.A.8. TEST CASE, SUPPORT TO HOLD YOURSELF IS NOT POSSIBLE

An army cannot get an additional hold power by supporting itself.

6.A.9. TEST CASE, FLEETS MUST FOLLOW COAST IF NOT ON SEA

If two provinces are adjacent, that does not mean that a fleet can move between those two provinces. An implementation that only holds one list of adjacent provinces for each province is incorrect.

An army can go from Rome to Venice, but a fleet cannot.

6.A.10. TEST CASE, SUPPORT ON UNREACHABLE DESTINATION NOT POSSIBLE

The destination of the move that is supported must be reachable by the supporting unit.

The support of Rome is illegal, because Venice cannot be reached from Rome by a fleet.

6.A.11. TEST CASE, SIMPLE BOUNCE

Two armies bouncing on each other.

The two units bounce.

6.A.12. TEST CASE, BOUNCE OF THREE UNITS

If three units move to the same province, the adjudicator should not bounce the first two units and then let the third unit go to the now open province.

The three units bounce.

6.B. TEST CASES, MULTI-COAST ISSUES

6.B.1. TEST CASE, MOVING WITH UNSPECIFIED COAST WHEN COAST IS NECESSARY

Coast is significant in this case:

6.B.2. TEST CASE, MOVING WITH UNSPECIFIED COAST WHEN COAST IS NOT NECESSARY

There is only one coast possible in this case:

Since the north coast is the only coast that can be reached, it seems logical that a move is attempted to the north coast of Spain. See issue 4.B.2.

I prefer that an attempt is made to the only possible coast, the north coast of Spain.

6.B.3. TEST CASE, MOVING TO IMPOSSIBLE COAST WHILE OTHER COAST IS POSSIBLE

If only one coast is possible, but the wrong coast can be specified.

If the rules are given a lenient interpretation, a move will be attempted to the north coast of Spain. However, this order is very precisely wrong. The order should be declared illegal and ignored. See issue 4.B.3.

6.B.4. TEST CASE, SUPPORT TO UNREACHABLE COAST ALLOWED

A fleet can give support to a coast where it cannot go.

Although the fleet in Marseilles cannot go to the north coast it can still support targeting the north coast.

6.B.5. TEST CASE, SUPPORT FROM UNREACHABLE COAST NOT ALLOWED

A fleet cannot give support to a province that cannot be reached from the current coast of the fleet.

The Gulf of Lyon cannot be reached from the north coast of Spain. Therefore, the support of Spain is illegal.

6.B.6. TEST CASE, SUPPORT CAN BE CUT WITH OTHER COAST

Support can be cut from the other coast.

The Italian fleet in the Gulf of Lyon will cut the support in Spain.

6.B.7. TEST CASE, SUPPORTING OWN UNIT WITH UNSPECIFIED COAST

It is a little bit harsh to reject this.

See issue 4.B.4.

I prefer that the support succeeds and the Italian fleet in the Western Mediterranean bounces. However, if orders are checked on submission (such as in web-based play), support without coast should not be given as an option.

6.B.8. TEST CASE, SUPPORTING WITH UNSPECIFIED COAST WHEN ONLY ONE COAST IS POSSIBLE

If coast is omitted while only coast is possible, it should be considered a poorly written order, that should be followed.

Support of Portugal is successful.

6.B.9. TEST CASE, SUPPORTING WITH WRONG COAST

It should be possible to specify a coast and that coast should match.

See issue 4.B.4. Coastal specification in Portugal support order does not match, making it invalid.

6.B.10. TEST CASE, UNIT ORDERED WITH WRONG COAST

A player might specify the wrong coast for the ordered unit.

If only perfect orders are accepted, then the move will fail, but since the coast for the ordered unit has no purpose, it might also be ignored (see issue 4.B.5).

I prefer that a move will be attempted.

6.B.11. TEST CASE, COAST CANNOT BE ORDERED TO CHANGE

The coast cannot change by just ordering the other coast.

6.B.12. TEST CASE, ARMY MOVEMENT WITH COASTAL SPECIFICATION

For armies the coasts are irrelevant:

If only perfect orders are accepted, then the move will fail. But it is also possible that coasts are ignored in this case and a move will be attempted (see issue 4.B.6).

I prefer that a move will be attempted.

6.B.13. TEST CASE, COASTAL CRAWL NOT ALLOWED

If a fleet is leaving a province from a certain coast while in the opposite direction another fleet is moving to another coast of the province, it is still a head-to-head battle. This has been decided in the great revision of the 1961 rules that resulted in the 1971 rules.

6.B.14. TEST CASE, BUILDING WITH UNSPECIFIED COAST

Coast must be specified in certain build cases:

See issue 4.B.7. Build fails, subsequent build orders may use up this right to build.

6.B.15. TEST CASE, SUPPORTING FOREIGN UNIT WITH UNSPECIFIED COAST

Opinions differ on this.

See issue 4.B.4.

Although the move to the north coast of Spain might be a surprise for France, it is hard to believe that England somehow tricked France. Therefore, I prefer that the support succeeds and the Italian fleet in the Western Mediterranean bounces. However, if orders are checked on submission (such as in web-based play), support without coast should not be given as an option.

6.B.16. TEST CASE, HOLD SUPPORT WITH WRONG COAST

For a fleet holding on a multi-coast province, the coast doesn't need to be agreed on.

See issue 4.B.5.

Although the English support order contains the wrong coast, the coast specification is not required. Therefore, I prefer that the support is valid.

6.B.17. TEST CASE, MOVE SUPPORT WITH WRONG COAST FOR DEPARTURE PROVINCE

For a fleet moving from a multi-coast province, the coast doesn't need to be agreed on.

See issue 4.B.5.

Also, for a move support the coast specification of the departing province is not required. Therefore, I prefer that the support is valid.

6.B.18. TEST CASE, CONVOY STARTING FROM MULTI-COAST PROVINCE

A convoy follows fleet movement. However, if starting at a multi-coast province, the army position (without coast specification) is not a legal fleet position.

6.B.19. TEST CASE, CONVOY ENDING AT MULTI-COAST PROVINCE

Similar to the previous test case, but now the convoy is ending at a multi-coast province.

6.C. TEST CASES, CIRCULAR MOVEMENT

6.C.1. TEST CASE, THREE ARMY CIRCULAR MOVEMENT

Three units can change place, even in Spring 1901.

6.C.2. TEST CASE, THREE ARMY CIRCULAR MOVEMENT WITH SUPPORT

Three units can change place, even when one gets support.

Since the move of Ankara will succeed in any case, a resolution where all units bounce is not consistent anymore.

6.C.3. TEST CASE, A DISRUPTED THREE ARMY CIRCULAR MOVEMENT

When one of the units bounces, the whole circular movement will hold.

6.C.4. TEST CASE, A CIRCULAR MOVEMENT WITH ATTACKED CONVOY

When the circular movement contains an attacked convoy, the circular movement succeeds. The adjudication algorithm should handle attack of convoys before calculating circular movement.

The fleet in the Ionian Sea is attacked but not dislodged. The circular movement succeeds.

6.C.5. TEST CASE, A DISRUPTED CIRCULAR MOVEMENT DUE TO DISLODGED CONVOY

When the circular movement contains a convoy, the circular movement is disrupted when the convoying fleet is dislodged. The adjudication algorithm should disrupt convoys before calculating circular movement.

Due to the dislodged convoying fleet, all Austrian and Turkish armies fail to move.

6.C.6. TEST CASE, TWO ARMIES WITH TWO CONVOYS

Two armies can swap places even when they are not adjacent.

6.C.7. TEST CASE, DISRUPTED UNIT SWAP

If in a swap one of the unit bounces, then the swap fails.

6.C.8. TEST CASE, NO SELF DISLODGMENT IN DISRUPTED CIRCULAR MOVEMENT

Self dislodgment is prohibited as usual in circular movement.

6.C.9. TEST CASE, NO HELP IN DISLODGMENT OF OWN UNIT IN DISRUPTED CIRCULAR MOVEMENT

Helping to dislodge your own unit is prohibited as usual in circular movement.

6.D. TEST CASES, SUPPORTS AND DISLODGES

6.D.1. TEST CASE, SUPPORTED HOLD CAN PREVENT DISLODGMENT

The simplest support to hold order.

6.D.2. TEST CASE, A MOVE CUTS SUPPORT ON HOLD

The simplest support on hold cut.

6.D.3. TEST CASE, A MOVE CUTS SUPPORT ON MOVE

The simplest support on move cut.

6.D.4. TEST CASE, SUPPORT TO HOLD ON UNIT SUPPORTING A HOLD ALLOWED

A unit that is supporting a hold can receive a hold support.

6.D.5. TEST CASE, SUPPORT TO HOLD ON UNIT SUPPORTING A MOVE ALLOWED

A unit that is supporting a move can receive a hold support.

6.D.6. TEST CASE, SUPPORT TO HOLD ON CONVOYING UNIT ALLOWED

A unit that is convoying can receive a hold support.

6.D.7. TEST CASE, SUPPORT TO HOLD ON MOVING UNIT NOT ALLOWED

A unit that is moving cannot receive a hold support for a situation in which the move fails.

The support of the fleet in Prussia does not match.

6.D.8. TEST CASE, FAILED CONVOY CANNOT RECEIVE HOLD SUPPORT

If a convoy fails because of disruption of the convoy or when the right convoy orders are not given, then the army to be convoyed cannot receive support in hold, since it still tried to move.

There was a possible convoy from Greece to Naples via the Ionian Sea, but the convoy order was not given. This means that the order of Greece to Naples is legal, but invalid. It cannot receive hold support.

6.D.9. TEST CASE, SUPPORT TO MOVE ON HOLDING UNIT NOT ALLOWED

A unit that is holding cannot receive a support in moving.

The support of the army in Albania does not match

6.D.10. TEST CASE, SELF DISLODGMENT PROHIBITED

A unit may not dislodge a unit of the same great power.

6.D.11. TEST CASE, NO SELF DISLODGMENT OF RETURNING UNIT

Idem.

6.D.12. TEST CASE, SUPPORTING A FOREIGN UNIT TO DISLODGE OWN UNIT PROHIBITED

You may not help another power to dislodge your own unit.

6.D.13. TEST CASE, SUPPORTING A FOREIGN UNIT TO DISLODGE A RETURNING OWN UNIT PROHIBITED

Idem.

6.D.14. TEST CASE, SUPPORTING A FOREIGN UNIT IS NOT ENOUGH TO PREVENT DISLODGMENT

If a foreign unit has enough support to dislodge your unit, you may not prevent that dislodgment by supporting the attack.

6.D.15. TEST CASE, DEFENDER CANNOT CUT SUPPORT FOR ATTACK ON ITSELF

A unit that is attacked by a supported unit cannot prevent dislodgment by guessing which of the units will do the support.

The support of Constantinople is not cut.

6.D.16. TEST CASE, CONVOYING A UNIT DISLODGING A UNIT OF SAME POWER IS ALLOWED

It is allowed to convoy a foreign unit that dislodges your own unit.

Rod Walker writes in Diplomacy World No. 28 page 18, 1981:

The Rulebook does not say (as if often alleged) that a country cannot participate in its own dislodgment. The Rules are actually sparkling clear on this point. Rule IX.3 is the only Rule (with Examples 1 and 2) that applies. It states that a player can't dislodge one of his own units by attacking it, and if he supports a foreign unit in an attack on a space occupied by one of his own units, the support (but not the attack per se) is invalid if that unit does not move out. (Example 2 shows that if the foreign unit has enough support of its own to succeed, it will, regardless.) Nowhere does the Rulebook state that a player can't convoy in an attack that dislodges one of his own units. Nor am I aware of any reason why it should.

6.D.17. TEST CASE, DISLODGMENT CUTS SUPPORTS

A support is not cut if it is supporting an attack on the attacker, unless it is dislodged.

The house rules of War Bulletin, July 1971 state:

22. Boardman's dilemma. Is the support of a unit dislodged by an attack from the space into which is it is giving support still valid?
GERMANY: A Pru-Sil S by A Mun. A Ber-Pru S by F Bal.
RUSSIA: A War-Pru S by A Sil (A Sil /d/)
Yes it is, because to cut support the attack must come from some other space. This ruling is Boardman's Rule. The opposite ruling, cutting the support of A Sil and allowing A Ber-Pru to succeed, is Miller's Rule.

Miller's rule made it to the 1971 rulebook.

6.D.18. TEST CASE, A SURVIVING UNIT WILL SUSTAIN SUPPORT

Idem. But now with an additional hold that prevents dislodgment.

Note that the fleet in Ankara has no province to retreat to and is therefore destroyed.

6.D.19. TEST CASE, EVEN WHEN SURVIVING IS IN ALTERNATIVE WAY

Now, the dislodgment is prevented because the support comes from a Russian army:

The Russian fleet in Constantinople is not dislodged, because one of the supports is of Russian origin.

6.D.20. TEST CASE, UNIT CANNOT CUT SUPPORT OF ITS OWN COUNTRY

A unit attacking another unit of the same Great Power will not cut support.

The army in York does not cut support.

6.D.21. TEST CASE, DISLODGING DOES NOT CANCEL A SUPPORT CUT

Sometimes there is the question whether a dislodged moving unit cuts support (similar to the dislodge rule). This is the case.

Although the German army is dislodged, it still cuts the Italian support.

6.D.22. TEST CASE, IMPOSSIBLE FLEET MOVE CANNOT BE SUPPORTED

If a fleet tries moves to an inland province it seems pointless to support the fleet, since the move will fail anyway. However, in such case, the support is also invalid for defense purposes.

The German move from Kiel to Munich is impossible (fleets cannot go to Munich), making both German orders illegal.

6.D.23. TEST CASE, IMPOSSIBLE COAST MOVE CANNOT BE SUPPORTED

Comparable with the previous test case, but now the fleet move is impossible for coastal reasons.

The French move from Spain north coast to Gulf of Lyon is impossible (wrong coast), making both French orders illegal.

6.D.24. TEST CASE, IMPOSSIBLE ARMY MOVE CANNOT BE SUPPORTED

Comparable with the previous test case, but now an army tries to move into sea and the support is used in a beleaguered garrison.

The French move from Marseilles to Gulf of Lyon is impossible (an army cannot go to sea), making both French orders illegal.

6.D.25. TEST CASE, INVALID HOLD SUPPORT CAN BE SUPPORTED

If an adjudicator fails on one of the previous three test cases, then the bug should be removed with care. A failing move cannot be supported, but an invalid hold support, because of some preconditions (unmatching order) can still be supported.

Although the support of Berlin on Prussia is invalid (because of unmatching orders), the support of Kiel on Berlin is still valid.

6.D.26. TEST CASE, INVALID MOVE SUPPORT CAN BE SUPPORTED

Similar as the previous test case, but now with an unmatched support to move.

6.D.27. TEST CASE, INVALID CONVOY CAN BE SUPPORTED

Similar as the previous test case, but now with an unmatched convoy.

The convoy order in the Baltic Sea is unmatched, making it invalid. However, the support of Prussia on the Baltic Sea is still valid.

6.D.28. TEST CASE, IMPOSSIBLE MOVE AND SUPPORT

An impossible move is "illegal" and should be ignored.

See issue 4.E.1. Illegal orders are ignored. Without an order, Rumania holds and receives support.

6.D.29. TEST CASE, MOVE TO IMPOSSIBLE COAST AND SUPPORT

Similar to the previous test case, but now the move is "illegal" due the wrong coast.

See issue 4.E.1. Illegal orders are ignored. Without an order, Rumania holds and receives support.

6.D.30. TEST CASE, MOVE ORDER WITHOUT COAST SPECIFICATION AND RECEIVING SUPPORT

Similar to the previous test case, but now the move is "illegal" due to missing coast specification.

See issue 4.E.1. Illegal orders are ignored. Without an order, Constantinople holds and receives support.

6.D.31. TEST CASE, A TRICKY IMPOSSIBLE SUPPORT

A support order can be impossible for complex reasons.

Although the army in Rumania can move to Armenia and the fleet in the Black Sea can also go to Armenia, the support is still not possible. The reason is that the only possible convoy is through the Black Sea and a fleet cannot convoy and support at the same time. Therefore, the order is marked as "illegal".

This is relevant for computer programs that show only the possible orders. In the list of possible orders, the support as given to the fleet in the Black Sea, should not be listed.

If there is a second order for the Black Sea, that order should be executed (see issue 4.E.1), because illegal orders are ignored.

6.D.32. TEST CASE, A MISSING FLEET

The previous test cases contained an order that was impossible even when some other pieces on the board were changed. In this test case, the order is impossible, but only for that situation.

The German order to Yorkshire cannot be executed, because there is no fleet in the North Sea. In other situations (where there is a fleet in the North Sea), the exact same order would be possible. This is considered "illegal" (see issue 4.E.1). The order should be ignored and the support of the French fleet in London is given.

6.D.33. TEST CASE, UNWANTED SUPPORT ALLOWED

A self standoff can be broken by an unwanted support.

Due to the Russian support, the army in Serbia advances to Budapest.

6.D.34. TEST CASE, SUPPORT TARGETING OWN PROVINCE NOT ALLOWED

Support targeting the province where the supporting unit is standing is illegal.

Russia and Italy wanted to get rid of the Italian army in Prussia (to build an Italian fleet somewhere else). However, they didn't want a possible German attack on Prussia to succeed. They invented this odd order of Italy. It was intended that the attack of the army in Livonia would have strength three, so it would be capable to prevent the possible German attack to succeed. However, the order of Italy is illegal, because a unit may only support to a province where the unit can go by itself. A unit can't go to the province it is already standing, so the Italian order is illegal.

Note, the army in Prussia is destroyed, because there is no province to retreat to.

6.D.35. TEST CASE, DISLODGMENT CUTS SUPPORTS ALLOWING ENEMY TO ADVANCE

Similar to test case 6.D.17, but now the army in Armenia successfully moves due to support.

6.E. TEST CASES, HEAD-TO-HEAD BATTLES AND BELEAGUERED GARRISON

6.E.1. TEST CASE, DISLODGED UNIT HAS NO EFFECT ON ATTACKER'S PROVINCE

An army can follow.

6.E.2. TEST CASE, NO SELF DISLODGMENT IN HEAD-TO-HEAD BATTLE

Self dislodgment is not allowed. This also counts for head-to-head battles.

6.E.3. TEST CASE, NO HELP IN DISLODGING OWN UNIT

It is not possible to help a foreign power dislodge own unit in a head-to-head battle.

6.E.4. TEST CASE, NON-DISLODGED LOSER STILL HAS EFFECT

If in an unbalanced head-to-head battle the loser is not dislodged, it still has an effect on the province of the attacker.

The French fleet in the North Sea is not dislodged due to the beleaguered garrison. Therefore, the Austrian army in Ruhr will not move to Holland.

6.E.5. TEST CASE, LOSER DISLODGED BY ANOTHER ARMY STILL HAS EFFECT

If in an unbalanced head-to-head battle the loser is dislodged by a unit not part of the head-to-head battle, the loser still has an effect on the province of the winner of the head-to-head battle.

The French fleet in the North Sea is dislodged but not by the German fleet in Holland. Therefore, the French fleet can still prevent that the Austrian army in Ruhr will move to Holland.

6.E.6. TEST CASE, NOT DISLODGE BECAUSE OF OWN SUPPORT STILL HAS EFFECT

If in an unbalanced head-to-head battle the loser is not dislodged because the winner had help of a unit of the loser, the loser still has an effect on the province of the winner.

Although the German force from Holland to North Sea is one larger than the French force from North Sea to Holland, the French fleet in the North Sea is not dislodged, because one of the supports on the German movement is French.

6.E.7. TEST CASE, NO SELF DISLODGMENT WITH BELEAGUERED GARRISON

An attempt at self dislodgment can be combined with a beleaguered garrison. Such self dislodgment is still not possible. This situation may lead to a discussion, but the outcome is not disputed. This was discussed on Usenet group rec.games.diplomacy with a thread starting on the 8th of August 2001.

Although the Russians beat the German attack (with the support of Yorkshire) and the two Russian fleets are enough to dislodge the fleet in the North Sea, the fleet in the North Sea is not dislodged, since it would not be dislodged if the English fleet in Yorkshire would not give support. This is a typical bug that can happen if a grand winner is calculated of a contested province (instead of calculating every move separately). The Russians are the grand winner of the North Sea with a strength of three, but this doesn't mean that they can advance.

6.E.8. TEST CASE, NO SELF DISLODGMENT WITH BELEAGUERED GARRISON AND HEAD-TO-HEAD BATTLE

Similar to the previous test case, but now the beleaguered fleet is also engaged in a head-to-head battle.

6.E.9. TEST CASE, ALMOST SELF DISLODGMENT WITH BELEAGUERED GARRISON

Similar to the previous test case, but now the beleaguered fleet is moving away.

6.E.10. TEST CASE, ALMOST CIRCULAR MOVEMENT WITH NO SELF DISLODGMENT WITH BELEAGUERED GARRISON

Similar to the previous test case, but now the beleaguered fleet is in circular movement with the weaker attacker. So, the circular movement fails.

6.E.11. TEST CASE, NO SELF DISLODGMENT WITH BELEAGUERED GARRISON, UNIT SWAP WITH ADJACENT CONVOYING AND TWO COASTS

Similar to the previous test case, but now the beleaguered fleet is in a unit swap with the stronger attacker. The unit swap succeeds. To make the situation more complex, the swap is on a province with two coasts.

Note, due to the success of the swap, there is no beleaguered garrison anymore.

6.E.12. TEST CASE, SUPPORT ON ATTACK ON OWN UNIT CAN BE USED FOR OTHER MEANS

A support on an attack on your own unit still has an effect. It can prevent another army from dislodging the unit.

The support by Serbia of the Italian army prevents that the Russian army in Galicia will advance.

6.E.13. TEST CASE, THREE WAY BELEAGUERED GARRISON

In a beleaguered garrison from three sides, the adjudicator may not let two attacks fail and then let the third succeed.

6.E.14. TEST CASE, ILLEGAL HEAD-TO-HEAD BATTLE CAN STILL DEFEND

If in a head-to-head battle, one of the units makes an illegal move, then that unit is still able to defend against attacks with a strength of one.

The move of the Russian fleet is illegal, but can still prevent the English army from entering Edinburgh.

6.E.15. TEST CASE, THE FRIENDLY HEAD-TO-HEAD BATTLE

In this case each unit in the head-to-head battle prevents that the other unit from being dislodged.

This case is especially difficult for sequence based adjudicators. They will start adjudicating the head-to-head battle and continue to adjudicate the attack on one of the units which is part of the head-to-head battle. In this process, one of the sides of the head-to-head battle might be cancelled out.

6.F. TEST CASES, CONVOYS

6.F.1. TEST CASE, NO CONVOY IN COASTAL PROVINCES

A fleet in a coastal province may not convoy.

The convoy in Constantinople is not allowed.

6.F.2. TEST CASE, AN ARMY BEING CONVOYED CAN BOUNCE AS NORMAL

Armies being convoyed bounce on other units just as armies that are not being convoyed.

6.F.3. TEST CASE, AN ARMY BEING CONVOYED CAN RECEIVE SUPPORT

Armies being convoyed can receive support similarly to any other move.

The army in London receives support and beats the army in Paris.

6.F.4. TEST CASE, AN ATTACKED CONVOY IS NOT DISRUPTED

A convoy can only be disrupted by dislodging at least one necessary fleet. Attacking is insufficient.

6.F.5. TEST CASE, A BELEAGUERED CONVOY IS NOT DISRUPTED

Even when a convoy is in a beleaguered garrison it is not disrupted.

6.F.6. TEST CASE, DISLODGED CONVOY DOES NOT CUT SUPPORT

When a necessary fleet of a convoy is dislodged, the convoy is completely cancelled. So, no support is cut.

6.F.7. TEST CASE, DISLODGED CONVOY DOES NOT CAUSE CONTESTED PROVINCE

When a fleet of a convoy is dislodged, the landing province is not contested, so other units can retreat to that province.

6.F.8. TEST CASE, DISLODGED CONVOY DOES NOT CAUSE A BOUNCE

When a fleet of a convoy is dislodged, then there will be no bounce in the landing province.

6.F.9. TEST CASE, DISLODGE OF MULTI-ROUTE CONVOY

When a fleet of a convoy with multiple routes is dislodged, the result depends on the rulebook that is used.

If the 1971 rules are used (see issue 4.A.1), this will disrupt the convoy and the army will stay in London.

When later rulebooks are used (which I prefer) the army can still go via the North Sea and the convoy succeeds.

6.F.10. TEST CASE, DISLODGE OF MULTI-ROUTE CONVOY WITH FOREIGN FLEET

When the 1971 rulebook is used "unwanted" multi-route convoys are possible.

If the 1971 rules are used (see issue 4.A.1), this will disrupt the convoy and the army will stay in London. Without the "help" of the Germans the convoy would have succeeded!

When later rulebooks are used (which I prefer) the army can still go via the North Sea and the convoy succeeds.

6.F.11. TEST CASE, DISLODGE OF MULTI-ROUTE CONVOY WITH ONLY FOREIGN FLEETS

With the 1971 rulebook one could adopt a rule (DPTG) that foreign fleets are not used when not necessary, but this doesn't prevent an "unwanted" convoy when all convoying fleets are foreign.

If the 1971 rules are used (see issue 4.A.1), this will disrupt the convoy and the army will stay in London.

When later rulebooks are used (which I prefer) the army can still go via the North Sea and the convoy succeeds.

6.F.12. TEST CASE, DISLODGED CONVOYING FLEET NOT ON ROUTE

When the 1971 rule is used for multi-route convoys (see issue 4.A.1), the convoy is not necessarily disrupted when one of the fleets ordered to convoy is dislodged.

In all rulebooks after 1971 the convoy will succeed, because there is at least one route available. However, even in the case of the 1971 rules the convoy should succeed, because the fleet in the Irish Sea is not part of any route, although it can be reached from the starting point London.

6.F.13. TEST CASE, THE UNWANTED ALTERNATIVE

This situation is not difficult to adjudicate, but it shows that even if someone wants to convoy, the player might not want an alternative route for the convoy.

If France and German are allies, England want to keep its army in London, to defend the island. An army in Belgium could easily be destroyed by an alliance of France and Germany. England tries to be friends with Germany, however France and Germany trick England.

6.F.14. TEST CASE, SIMPLE CONVOY PARADOX

The most common paradox is when the attacked unit supports an attack on one of the convoying fleets.

According to all rulebooks (including the Szykman rule which I prefer), the support of London is not cut. See also issue 4.A.2.

6.F.15. TEST CASE, SIMPLE CONVOY PARADOX WITH ADDITIONAL CONVOY

Paradox rules only apply on the paradox core.

The adjudication of the paradox in the English Channel should not interfere with the adjudication of the Italian convoy.

6.F.16. TEST CASE, PANDIN'S PARADOX

Named after Tony Pandin. The unit that is attacked by a convoy protects the convoying fleet via a beleaguered garrison.

On the 11^th of July 2009 Chris Babcock forwarded an email from Edi Birsan on Usenet group rec.games.diplomacy:

For example the Pandin Paradox- I was there when it actually occurred in a FtF game (DipCon 2- which was a Youngstown game by the way and it happened in the Ionian not the Channel as typically it is thought of). That has been the only time in 44 years of play and thousands of games that I have ever seen it occur.

DipCon 2 took place in 1969 in Youngstown Ohio.

According to all rulebooks (including the Szykman rule which I prefer), the support of London is not cut and the overall result is the same. Note however, that the Szykman rule targets the convoy and makes it "disrupted", while the rulebooks target the support of London and will leave the convoy "available". See issue 4.A.2.

6.F.17. TEST CASE, PANDIN'S EXTENDED PARADOX

In Pandin's extended paradox, the attacked unit protects the convoying fleet by a beleaguered garrison and the attacked unit can dislodge the unit that gives the protection.

When the 1971/1982/2000/2023/2025EE rules are used (see issue 4.A.2), the support of London is not cut. That means that the fleet in the English Channel is not dislodged. The convoy will succeed and dislodge the fleet in London. One can argue that this violates the dislodge rule, but one may assume that the paradox convoy rule takes precedence over the dislodge rule.

If the Szykman rule is used (which I prefer), the convoy is disrupted and the fleet in London is not dislodged (which I think is a more appealing adjudication).

6.F.18. TEST CASE, BETRAYAL PARADOX

The betrayal paradox is comparable to Pandin's paradox, but now the attacked unit directly supports the convoying fleet. Of course, this will only happen when the player of the attacked unit is betrayed.

This type of paradox was reported on Reddit on the 30^th of April, 2020 with title: "Convoy Paradox Spotted in the wild".

The 1971, 2000 and 2023 rules do not give an answer on this (see issue 4.A.2).

According to the 1982 rules the French support on the North Sea will not be cut. So, the fleet in the North Sea will not be dislodged by the Germans and the army in London will dislodge the French army in Belgium.

If the Szykman rule is followed (which I prefer), the convoy in the North Sea is disrupted. Without the convoy, the move of the army in London will fail and the support in Belgium will not be cut.

6.F.19. TEST CASE, MULTI-ROUTE CONVOY PARADOX

The situation becomes more complex when the convoy has alternative routes.

Now, two issues play a role. The rule about disruption of multi-route convoys (issue 4.A.1) and the determination of how paradoxes are resolved (issue 4.A.2).

If the 1971 rulebook is used then a multi-route convoy is disrupted when one of the routes is disrupted. That makes this situation paradoxical and the 1971 paradox rule kicks in. The support of the fleet in Naples is not cut and the fleet in Rome dislodges the fleet in the Tyrrhenian Sea.

With the 1982 rulebook, the support of Naples is not cut, because it is supporting an action in a body of water that contains a convoying fleet. This means that the fleet in Rome dislodges the fleet in the Tyrrhenian Sea.

According to the 2000/2023/2025EE rules the fleet in the Tyrrhenian Sea is not "necessary" for the convoy and the support of Naples is cut and the fleet in the Tyrrhenian Sea is not dislodged.

If the Szykman rule is used (which I prefer), then there is no paradoxical situation. The support of Naples is cut (the same as with the 2000/2023/2025EE rules) and the fleet in the Tyrrhenian Sea is not dislodged.

6.F.20. TEST CASE, UNWANTED MULTI-ROUTE CONVOY PARADOX

The 1982 paradox rule allows some creative defense.

Again, two issues play a role. The rule about disruption of multi-route convoys (issue 4.A.1) and the determination of how paradoxes are resolved (issue 4.A.2).

If the 1971 rulebook is used, then a multi-route convoy is disrupted when one of the routes is disrupted. This makes the situation paradoxical. However, since the fleet in Naples is not supporting an attack on a convoying fleet, the paradox rule does not apply and the 1971 rules do not give answer to this situation.

With the 1982 rules the support in Naples is not cut, because it is supporting an action in a body of water that contains a convoying fleet. That means that the fleet in the Ionian Sea is not dislodged.

The paradox rule of the 2000/2023/2025EE rules, does not kick in, because the support is not a support that attacks the convoying fleet. However, with these rules a multi-route convoy is only disrupted when all routes are disrupted, which prevents that this situation is a paradox. So, the support of Naples is cut and the fleet in the Ionian Sea is dislodged by the Turkish fleet in the Eastern Mediterranean.

If the Szykman rule is used, then there is no paradoxical situation. The support of Naples is cut and the fleet in the Ionian Sea is dislodged by the Turkish fleet in the Eastern Mediterranean.

As you can see, the 1982 rules allow the Italian player to save its fleet in the Ionian Sea with a trick. I do not consider this trick as normal tactical play. I prefer the Szykman rule as one of the rules that does not allow this trick.

6.F.21. TEST CASE, DAD'S ARMY CONVOY

The 1982 paradox rule has as side effect that convoying armies do not cut support in some situations that are not paradoxical.

With all rules, except the 1982 paradox rule, the support of the fleet in Clyde on the North Atlantic Ocean is cut and the French fleet in the Mid-Atlantic Ocean will dislodge the fleet in the North Atlantic Ocean. This is the preferred way.

However, with the 1982 paradox rule (see issue 4.A.2), the support of the fleet in Clyde is not cut. That means that the English fleet in the North Atlantic Ocean is not dislodged.

As you can see, the 1982 rule allows England to save its fleet in the North Atlantic Ocean in a very strange way. Just the support of Clyde is insufficient (if there is no convoy, the support is cut). Only the convoy to the province occupied by own unit, can do the trick in this situation. The embarking of troops in the fleet deceives the enemy so much that it works as a magic cloak. The enemy is not able to dislodge the fleet in the North Atlantic Ocean any more. Of course, this will only work in comedies. I prefer the Szykman rule as one of the rules that does not allow this trick. According to this rule (and all other paradox rules), the fleet in the North Atlantic is just dislodged.

6.F.22. TEST CASE, SECOND ORDER PARADOX WITH TWO RESOLUTIONS

Two convoys are involved in a second order paradox.

Without any paradox rule, there are two consistent resolutions. The supports of the English fleet in London and the German fleet in Picardy are not cut. That means that the French fleet in the English Channel and the Russian fleet in the North Sea are dislodged, which makes it impossible to cut the support. The other resolution is that the supports of the English fleet in London the German fleet in Picardy are cut. In that case the French fleet in the English Channel and the Russian fleet in the North Sea will survive and will not be dislodged. This gives the possibility to cut the support.

The 1971/2000/2023/2025EE rules (see issue 4.A.2) do not have an answer on this.

According to the 1982 rule, the supports are not cut which means that the French fleet in the English Channel and the Russian fleet in the North Sea are dislodged.

With the Szykman rule (which I prefer) the convoys are disrupted. This has as consequence that the supports are not cut and therefore, the overall result is the same as with the 1982 rules.

6.F.23. TEST CASE, SECOND ORDER PARADOX WITH TWO EXCLUSIVE CONVOYS

In this paradox there are two consistent resolutions, but where the two convoys do not fail or succeed at the same time.

Without any paradox rule, there are two consistent resolutions. In one resolution, the convoy in the English Channel is dislodged by the fleet in the Mid-Atlantic Ocean, while the convoy in the North Sea succeeds. In the other resolution, it is the other way around. The convoy in the North Sea is dislodged by the fleet in Edinburgh, while the convoy in the English Channel succeeds.

The 1971/2000/2023/2025EE rules (see issue 4.A.2) do not have an answer on this.

According to the 1982 rule, the supports are not cut which means that the none of the units move.

With the Szykman rule (which I prefer) the convoys are disrupted. This has as consequence that the supports are not cut and therefore, the overall result is the same as with the 1982 rules.

6.F.24. TEST CASE, SECOND ORDER PARADOX WITH NO RESOLUTION

As first order paradoxes, second order paradoxes come in two flavors, with two resolutions or no resolution.

When no paradox rule is used, there is no consistent resolution. If the French support in Belgium is cut, the French fleet in the English Channel will be dislodged. That means that the support of London will not be cut and the fleet in Edinburgh will dislodge the Russian fleet in the North Sea. In this way the support in Belgium is not cut! But if the support in Belgium is not cut, the Russian fleet in the North Sea will not be dislodged and the army in Norway can cut the support in Belgium.

The 1971/2000/2023/2025EE rules (see issue 4.A.2) do not have an answer on this.

According to the 1982 rule, the supports are not cut, which means that the French fleet in the English Channel will survive and but the Russian fleet in the North Sea is dislodged.

If the Szykman rule is used (which I prefer), the convoys are disrupted, which gives the same result as with the 1982 rules.

6.F.25. TEST CASE, CUT SUPPORT LAST

For manual play the rule of thumb is, cut support first. However, in the example below the support of Holland is some of the last orders to be adjudicated.

The fleet in Sweden fails to disrupt the convoy in Skagerrak. The move from Denmark to Norway succeeds and cuts the support of Norway. The fleet in the Norwegian Sea fails to disrupt the convoy in North Sea. The move from Yorkshire to Holland succeeds and cuts the support of Holland. The move from Ruhr fails to dislodge the army in Belgium.

6.F.26. TEST CASE, CONVOY PARADOX WITH ADDITIONAL RELATED CONVOYS

The Szykman rule only affects the convoying fleets part of the paradox, even if other convoying fleets have some relation with the situation.

According to all rulebooks (including the Szykman rule which I prefer), the support of London is not cut. See also issue 4.A.2. The Szykman rule does not affect the fleets in Mid-Atlantic Ocean, North Atlantic Ocean and Irish Sea.

6.F.27. TEST CASE, TWO INDEPENDENT PARADOXES

Can the adjudicator handle two paradoxes at the same time?

6.F.28. TEST CASE, SIXTH ORDER PARADOX

By John Fisher, Usenet group rec.games.diplomacy, 2^nd of September, 2002.

If played under the 1971/2000/2023/2025EE rules (see issue 4.A.2) there is no resolution.

Under the 1982 rules or the Szykman rule (which I prefer) all attacked convoying fleets are dislodged.

6.F.29. TEST CASE, SIXTH ORDER BUTTERFLY EFFECT

The same as previous test case, but now there is a support in Portugal for the Mid-Atlantic Ocean. This has effect as far as the Gulf of Bothnia.

If played under the 1982 rules, none of the supports are cut. As result all attacked convoying fleets are dislodged, except for the Mid-Atlantic Ocean due to the additional support from Portugal.

Since there is no convoy paradox the other rulebooks (which I prefer) have a resolution. All the supports are cut and none of the convoying fleets are dislodged. Since there is no convoy paradox, the Szykman rule doesn't apply.

6.F.30. TEST CASE, THE ATTACK COMES FROM THE STARTING POSITION

Brannan's rule has not been adopted.

One could argue that there is a head-to-head battle between Belgium and English Channel with both units having strength two resulting in a standoff. This is the interpretation under Brannan's rule.

Calhamer opposed Brannan's rule (see issue 4.A.4) and the rule has not been adopted. In Erehwon No. 65, 1972 a letter was placed from Calhamer where he gives the adjudication of above test case not following Brannan's rule:

2. ENGLAND: A Lon-Bel, F Eng C A Lon-Bel /r/, F Nth S A Lon-Bel.
   FRANCE: F Bel-Eng, F Bre S F Bel-Eng.

Here the supported attacker just drives out the convoying fleet; and since the fleet is dislodged, the convoy does not succeed.

6.G. TEST CASES, CONVOYING TO ADJACENT PROVINCES

6.G.1. TEST CASE, TWO UNITS CAN SWAP PROVINCES BY CONVOY

The only way to swap two units, is by convoy.

If explicit adjacent convoying is used (DPTG, see issue 4.A.3), then it is just a head-to-head battle. However, all rulebooks (which I prefer) allow that convoy intent is given by a convoying fleet of same country.

6.G.2. TEST CASE, KIDNAPPING AN ARMY

Germany promised England to support to dislodge the Russian fleet in Sweden and it promised Russia to support to dislodge the English army in Norway. Instead, the joking German orders a convoy.

See issue 4.A.3. If the 1971 rulebook is used with the recommended handling, then the army in Norway is kidnapped and swaps with the army in Sweden. In all other rulebooks (which I prefer), kidnapping is prevented and the armies fail to move.

One could argue that the Skagerrak order should be marked "invalid". That would make sense in the case where the 'via convoy' tag is also required when convoying with own fleets (as in the DPTG).

6.G.3. TEST CASE, AN UNWANTED DISRUPTED CONVOY TO ADJACENT PROVINCE

One can try to convoy an army unwanted with a fleet that is almost certainly dislodged. However, this trick should not work.

See issue 4.A.3. The 1982/2000/2023/2025EE rulebooks (which I prefer) dictate only use of the convoy route if intent is clear. The army in Picardy will successfully move by land route to Belgium. In the case of the 1971 rulebook it is unclear. However, if the recommended handling is used the convoy is not used for two reasons, the convoy is disrupted and there is no unit in Belgium moving in opposite direction.

6.G.4. TEST CASE, AN UNWANTED DISRUPTED CONVOY TO ADJACENT PROVINCE AND OPPOSITE MOVE

In the situation of the previous test case, it was rather clear that the army didn't want to take the convoy. But what if there is an army moving in opposite direction?

See issue 4.A.3. In the case of the 1971 rules, it is unclear whether the French army in Picardy will take the land route. However, with the recommended handling, the convoy is not taken since it is disrupted. The move of the army in Picardy will succeed.

With the 1982/2000/2023/2025EE rulebooks (which I prefer), kidnapping is prevented.

6.G.5. TEST CASE, SWAPPING WITH MULTIPLE FLEETS WITH ONE OWN FLEET

One fleet is sufficient to show the intent to convoy.

If explicit adjacent convoying is used (DPTG, see issue 4.A.3), then it is just a head-to-head battle. However, all rulebooks (which I prefer) allow that convoy intent is given by a convoying fleet of same country. So, the swap should happen.

6.G.6. TEST CASE, SWAPPING WITH UNINTENDED INTENT

The intent is questionable.

Here England intended to convoy via the French fleets in the Irish Sea and the North Sea. However, the French did not order the convoy. The alternative route with the Russian fleets was unintended. The English fleet in the English Channel (with the convoy order) is not part of this alternative route with the Russian fleets.

See issue 4.A.3.

If the 1971 rules are used, the intent is not important and the units are swapped.

In the case of the 1982/2000/2023/2025EE rulebooks (which I prefer) England still intends to convoy and the armies should swap.

When explicit adjacent convoying is used (DPTG), then the English army did not receive an order to move by convoy. So, it is just a head-to-head battle and both the army in Edinburgh and Liverpool will not move.

6.G.7. TEST CASE, SWAPPING WITH ILLEGAL INTENT

Can the intent be made clear with an impossible order?

See issue 4.A.3 and 4.E.1.

In case the 1971 rules are used, the intent is not important and the units in Norway and Sweden swap.

With the 2023/2025EE rules (which I prefer) impossible orders are ignored. Also, with some modern web-based adjudicators, impossible orders cannot be given at all. With this, there is no intent to convoy and the units in Norway and Sweden fail to move.

If explicit adjacent convoying is used (DPTG) there is also no convoy and none of the units move.

6.G.8. TEST CASE, EXPLICIT CONVOY THAT ISN'T THERE

What to do when a unit is explicitly ordered to move via convoy and the convoy is not ordered?

The French army in Belgium intended to move convoyed with the English fleet in the North Sea. But England changed its plans.

See issue 4.A.3.

In the case of the 1971 or 1982 rulebook, this test case not applicable, because they don't have the notion of 'via convoy'.

For the 2000/2023 rulebooks and the DPTG, the question is whether the land route should be used as "fallback".

The 2025EE reformulation doesn't allow a fallback. I prefer the latest rules.

6.G.9. TEST CASE, SWAPPED OR DISLODGED?

In the following situation the English army in Norway will end in all cases in Sweden. But whether it is convoyed or not has effect on the Russian army. In the case of a convoy the Russian army ends in Norway and in the case of a land route the Russian army is dislodged (see issue 4.A.3).

If played according to the DPTG, then an army is only convoyed to an adjacent province if it is tagged with 'via convoy'. This means that the Russian army in Sweden is dislodged by the army from Norway.

If played according to any of the rulebooks (which I prefer) then the move of Norway is via convoy and the armies swap.

6.G.10. TEST CASE, SWAPPED OR AN HEAD-TO-HEAD BATTLE?

Can a dislodged unit have effect on the attacker's province, when the attacker moved by convoy?

Since England ordered the army in Norway to move explicitly via convoy and the army in Sweden is moving in opposite direction, there is no head-to-head battle. It is clear that the army in Norway will dislodge the Russian army in Sweden. Since the strength of three is in all cases the strongest force.

The army in Sweden will not advance to Norway, because it cannot beat the force in the Norwegian Sea. It will be dislodged by the army from Norway.

The more interesting question is whether the French fleet in the Norwegian Sea is bounced by the Russian army from Sweden. This depends on the interpretation of issue 4.A.7. If choice a is taken, then a dislodged unit cannot bounce a unit in the province where the attacker came from. This would mean that the move of the fleet in the Norwegian Sea succeeds. However, if choice b is taken (which I prefer), then a bounce is still possible, when there is no head-to-head battle.

6.G.11. TEST CASE, A CONVOY TO AN ADJACENT PROVINCE WITH A PARADOX

In this case the convoy route is available when the land route is chosen and the convoy route is not available when the convoy route is chosen.

See issue 4.A.2 and 4.A.3.

In the case of the 1971 rulebook with the recommended handling the move from Sweden to Norway is not a convoy (because Norway is not moving in opposite direction) and the English fleet in Norway is dislodged and the fleet in Skagerrak will not be dislodged.

In the case of the 1982/2000/2023 rulebook, the question arises whether the land route is the fallback of the convoy route. If so, then there is the paradoxical situation that the convoy route is available when the land route is taken, but not if the convoy route is taken. Applying the Szykman rule results in the fleet failing to participate in the convoy. The land route will be taken, and similar to the 1971 ruling, the English fleet in Norway is dislodged and the fleet in Skagerrak will not be dislodged.

The 2025EE (which I prefer)rules are explicit that there is no fallback. So, only the convoy route is considered. This results in a simple convoy paradox.

6.G.12. TEST CASE, SWAPPING TWO UNITS WITH TWO CONVOYS

Of course, two armies can also swap when they are both convoyed.

6.G.13. TEST CASE, SUPPORT CUT ON ATTACK ON ITSELF VIA CONVOY

If a unit is attacked by a supported unit, it is not possible to prevent dislodgment by trying to cut the support. But what, if a move is attempted via a convoy?

First it should be mentioned that if for issue 4.A.3 the 1971 rulebook is chosen with the recommended handling, the move from Trieste to Venice is just a move over land (because Venice does not move in opposite direction). In that case, the support of Venice will not be cut as normal.

With the 1982/2000/2023/2025EE rulebooks the attack is via convoy and it should be decided whether the Austrian attack is considered to be coming from Trieste or from the Adriatic Sea. If it comes from Trieste, the support in Venice is not cut and the army in Trieste is dislodged by the fleet in Albania. If the Austrian attack is considered to be coming from the Adriatic Sea (Brannan's rule), then the support is cut and the army in Trieste will not be dislodged.

See issue 4.A.4. Brannan's rule has not been adopted. The attack comes from Trieste and the support is not cut.

6.G.14. TEST CASE, BOUNCE BY CONVOY TO ADJACENT PROVINCE

Similar to test case 6.G.10, but now the other unit is taking the convoy.

Again, the army in Sweden is bounced by the fleet in the Norwegian Sea. The army in Norway will move to Sweden and dislodge the Russian army.

The final destination of the fleet in the Norwegian Sea depends on how issue 4.A.7 is resolved. If choice a is taken, then the fleet advances to Norway, but if choice b is taken (which I prefer) the fleet bounces and stays in the Norwegian Sea.

6.G.15. TEST CASE, BOUNCE AND DISLODGE WITH DOUBLE CONVOY

Similar to test case 6.G.10, but now both units use a convoy and with some support.

The French army in Belgium is bounced by the army from Yorkshire. The army in London move to Belgium, dislodging the unit there.

The final destination of the army in the Yorkshire depends on how issue 4.A.7 is resolved. If choice a is taken, then the army advances to London, but if choice b is taken (which I prefer) the army bounces and stays in Yorkshire.

6.G.16. TEST CASE, THE TWO UNIT IN ONE PROVINCE BUG, MOVING BY CONVOY

If the adjudicator is not correctly implemented, this may lead to a resolution where two units end up in the same province.

See decision details 5.B.6. If the 'PREVENT STRENGTH' is incorrectly implemented, due to the fact that it does not take into account that the 'PREVENT STRENGTH' is only zero when the unit is engaged in a head-to-head battle, then this goes wrong in this test case. The 'PREVENT STRENGTH' of Sweden would be zero, because the opposing unit in Norway successfully moves. Since, this strength would be zero, the fleet in the North Sea would move to Norway. However, although the 'PREVENT STRENGTH' is zero, the army in Sweden would also move to Norway. So, the final result would contain two units that successfully moved to Norway.

Of course, this is incorrect. The fleet in the North Sea will bounce.

6.G.17. TEST CASE, THE TWO UNIT IN ONE PROVINCE BUG, MOVING OVER LAND

Similar to the previous test case, but now the other unit moves by convoy.

6.G.18. TEST CASE, THE TWO UNIT IN ONE PROVINCE BUG, WITH DOUBLE CONVOY

Similar to the previous test case, but now both units move by convoy.

6.G.19. TEST CASE, SWAPPING WITH INTENT OF UNNECESSARY CONVOY

Can intent be made clear by the order of a fleet that is not necessary?

See issue 4.A.3 and 4.E.1.

In the case the 1971 rules are used, the intent is not important and the units in Marseilles and Spain swap.

The point of interest is that there is a convoy route from Marseilles, Gulf of Lyon, Western Mediterranean to Spain. However, the fleet in Western Mediterranean is not necessary for this convoy and not necessary for any other convoy route. Therefore, this order could be considered illegal. Web-based adjudicators should not give this order as an option.

With the 2023/2025EE rules (which I prefer) illegal orders are ignored. The fleet in Gulf of Lyon is foreign and foreign units cannot express intent. But this position is a little bit pedantic.

If explicit adjacent convoying is used (DPTG) there is also no convoy and none of the units move.

6.G.20. TEST CASE, EXPLICIT CONVOY TO ADJACENT PROVINCE DISRUPTED

If a move to adjacent province was explicit via convoy, and the convoy is disrupted, should it fall back to the land route?

This situation is not applicable for the 1971 and 1982 rulebooks, because they don't have the notion of 'via convoy'.

With the 2000/2023 rulebooks the question arises whether the army in Picardy will fall back to the land route, since the convoy route is disrupted. See issue 4.A.3.

The 2025EE rulebook (which I prefer) is clearer, no fallback.

6.H. TEST CASES, RETREATING

6.H.1. TEST CASE, NO SUPPORTS DURING RETREAT

Supports are not allowed in the retreat phase.

The Austrian support order is illegal. Both dislodged fleets are disbanded.

6.H.2. TEST CASE, NO SUPPORTS FROM RETREATING UNIT

A retreating unit cannot give support.

Although the fleet in Holland may receive an order, it may not support.

6.H.3. TEST CASE, NO CONVOY DURING RETREAT

Convoys during retreat are not allowed.

The convoy order is illegal. The army in Holland is disbanded.

6.H.4. TEST CASE, NO OTHER MOVES DURING RETREAT

Of course, you may not do any other move during a retreat. But look if the adjudicator checks for it.

6.H.5. TEST CASE, A UNIT MAY NOT RETREAT TO THE PROVINCE FROM WHICH IT IS ATTACKED

Well, that would be of course stupid. Still, the adjudicator must be tested on this.

6.H.6. TEST CASE, UNIT MAY NOT RETREAT TO A CONTESTED PROVINCE

Standoff prevents retreat to the province.

6.H.7. TEST CASE, MULTIPLE RETREAT TO SAME PROVINCE WILL DISBAND UNITS

There can only be one unit in a province.

6.H.8. TEST CASE, TRIPLE RETREAT TO SAME PROVINCE WILL DISBAND UNITS

When three units retreat to the same province, then all three units are disbanded.

All three units are disbanded.

6.H.9. TEST CASE, DISLODGED UNIT WILL NOT MAKE ATTACKER'S PROVINCE CONTESTED

An army can follow.

6.H.10. TEST CASE, RETREATING UNIT CANNOT BOUNCE IN ATTACKER'S PROVINCE

Similar to the previous test case, but here the dislodged unit tries to bounce a unit in the province from where the attack came.

The English attempt to retreat to Berlin will not bounce the unit from Prussia.

6.H.11. TEST CASE, RETREAT WHEN DISLODGED BY ADJACENT CONVOY

If a unit is dislodged by an army via convoy, the question arises whether the dislodged army can retreat to the original province of the convoyed army. This is only relevant in the case where the convoy was to an adjacent province.

If the 1971 rules are used with the recommended handling, then the convoy is ignored and the land route is taken and there is normal dislodgment. Marseilles may not retreat to Gascony.

With the other rulebooks, the army in Gascony takes a convoy and does not pass the border of Gascony with Marseilles (it went a completely different direction). Now, the result depends on which rule is used for retreating (see issue 4.A.5). The 2023/2025EE rules (which I prefer) explicitly allow this.

6.H.12. TEST CASE, RETREAT WHEN DISLODGED BY ADJACENT CONVOY WHILE TRYING TO DO THE SAME

The previous test case can be made more extra ordinary, when both armies tried to move by convoy.

Both the army in Liverpool as in Edinburgh will try to move by convoy, but the convoy from Liverpool to Edinburgh is disrupted.

If the 1971 rules are used with the recommended handling, then the move from Edinburgh to Liverpool will go successfully over land and there is no dislodgment.

For the other rulebooks the army in Liverpool will fail to move and be dislodged. The question arises whether the army in Liverpool may retreat to Edinburgh (see issue 4.A.5). The 2023/2025EE rules (which I prefer) explicitly allow this.

6.H.13. TEST CASE, NO RETREAT WITH CONVOY IN MOVEMENT PHASE

The provinces where a unit may retreat to must be determined during the movement phase. Care should be taken that a convoy ordered in the movement phase cannot be used in the retreat phase.

6.H.14. TEST CASE, NO RETREAT WITH SUPPORT IN MOVEMENT PHASE

Comparable to the previous test case, a support given in the movement phase cannot be used in the retreat phase.

The support given by the English Channel will not be effective in the retreat phase.

6.H.15. TEST CASE, NO COASTAL CRAWL IN RETREAT

You cannot go to the other coast of the province from which the attacker came.

6.H.16. TEST CASE, CONTESTED FOR BOTH COASTS

If one of the coasts is contested, the other is not available for retreat.

6.H.17. TEST CASE, DESTROYED WHILE HOLDING

If an army cannot retreat the adjudicator can make that clear in the output.

Note, other test cases (with different purpose) have a unit destroyed with an order different than Hold.

6.I. TEST CASES, BUILDING

6.I.1. TEST CASE, TOO MANY BUILD ORDERS

Check how program reacts when someone orders too many builds.

See issue 4.D.4. I prefer that the build orders are just handled one by one until all allowed units are build.

6.I.2. TEST CASE, FLEETS CANNOT BE BUILT ON INLAND PROVINCES

Only army builds are allowed in inland provinces.

See issue 4.C.4.

6.I.3. TEST CASE, SUPPLY CENTER MUST BE EMPTY FOR BUILDING

You can't have two units in a province. So, you can't build when there is a unit in the supply center.

6.I.4. TEST CASE, BOTH COASTS MUST BE EMPTY FOR BUILDING

If a province is occupied on one coast, the other coast cannot be used for building.

6.I.5. TEST CASE, BUILDING IN HOME SUPPLY CENTER THAT IS NOT OWNED

Building a unit is only allowed when supply center is a home supply center and is owned. If not owned, build fails.

Russia captured Berlin in Fall, but left in the next year. Germany captured other supply centers, but without recapturing Berlin it may not build in Berlin.

6.I.6. TEST CASE, BUILDING IN OWNED SUPPLY CENTER THAT IS NOT A HOME SUPPLY CENTER

Building a unit is only allowed when supply center is a home supply center and is owned. If it is not a home supply center, the build fails.

Germany owns Warsaw, Warsaw is empty and Germany may build one unit.

6.I.7. TEST CASE, ONLY ONE BUILD IN A HOME SUPPLY CENTER

If you may build two units, you can still only build one in a supply center.

6.J. TEST CASES, CIVIL DISORDER AND DISBANDS

6.J.1. TEST CASE, TOO MANY DISBAND ORDERS

Check how program reacts when someone orders too many disbands.

France has to disband one and has an army in Paris and Picardy (and no fleet in Gulf of Lyon).

See issue 4.D.6. I prefer that the disband orders are handled one by one.

6.J.2. TEST CASE, REMOVING THE SAME UNIT TWICE

If you have to remove two units, you can always try to trick the computer by removing the same unit twice.

France has to disband two:

6.J.3. TEST CASE, CIVIL DISORDER TWO ARMIES WITH DIFFERENT DISTANCE

When a player forgets to disband a unit, the civil disorder rules must be applied. When two armies have different distance to the owned supply centers, then the army with the greatest distance has to be removed.

6.J.4. TEST CASE, CIVIL DISORDER TWO ARMIES WITH EQUAL DISTANCE

If two armies have equal distance to the owned supply centers, then alphabetical order is used.

Livonia appears first in alphabetical order.

6.J.5. TEST CASE, CIVIL DISORDER TWO FLEETS WITH DIFFERENT DISTANCE

If two fleets have different distance to the owned supply centers, then the fleet with the greatest distance has to be removed.

The distance of the fleet in North Atlantic Ocean is three, the fleet in Skagerrak has distance two (via Norway).

6.J.6. TEST CASE, CIVIL DISORDER TWO FLEETS WITH EQUAL DISTANCE

Alphabetical order is used, when two fleets have equal distance to the owned supply centers.

Both distances are three. Gulf of Bothnia appears first in alphabetical order.

6.J.7. TEST CASE, CIVIL DISORDER TWO FLEETS AND ARMY WITH EQUAL DISTANCE

In removal, the fleet has precedence over an army. In this case there are two fleets, to make the test more complex.

The distances of the army and the fleets to one of the owned supply centers are two. The fleets take precedence above the army (although the army is alphabetical first). The fleet in the North Sea is alphabetical first compared to Skagerrak.

6.J.8. TEST CASE, CIVIL DISORDER A FLEET WITH SHORTER DISTANCE THAN THE ARMY

If the fleet has a shorter distance than the army, the army is disbanded.

The distance of the army to Warsaw is three while the distance of the fleet is two.

6.J.9. TEST CASE, CIVIL DISORDER MUST BE COUNTED FROM BOTH COASTS, SOUTH COAST SHORTEST

Distance must be calculated from both coasts.

The distance of the fleet to St Petersburg(nc) is three but to St Petersburg(sc) is two.

6.J.10. TEST CASE, CIVIL DISORDER MUST BE COUNTED FROM BOTH COASTS, NORTH COAST SHORTEST

Similar to the previous test case, but now the distance to the north coast is the shortest one.

The distance of the fleet to St Petersburg(sc) is three but to St Petersburg(nc) is two.

6.J.11. TEST CASE, CIVIL DISORDER COUNTING CONVOYING DISTANCE

For calculating the distance for armies all provinces must be considered.

The distance from Albania to owned supply center is four over land. However, for distance calculation it can go over water and arrive in two steps.

6.J.12. TEST CASE, DISTANCE TO OWNED SUPPLY CENTER

The 2023 rules say that distance must be calculated to owned supply center instead of home supply center (as it was in the older rulebooks).

The army in Tuscany is removed and Italy will continue defending its supply center in Warsaw. Under older rulebooks the army in Tuscany was kept.