About the Convoy Paradox Resolution

The Readership goes berko... responses to last issue's article
Eliminating the Paradox in Diplomacy



Good grief, did this one stir you all up into a frenzy! (Mental note: Next time Manus asks me if I'll be the Guest Editor for an issue of The Pouch, before agreeing, check the previous issue to see if there are any articles about convoy paradoxes... if there are, don't bother saying no, just begin running for the hills straight away....)

Before we start seeing the plethora of responses regarding the article, it's interesting to note the statistics of your responses:

A total of 298 people responded to the survey. Of these, 45 changed the vote that they originally cast after reading the two essays. Here are the final numbers:

 Original VoteVotes Changed To Final Vote
Proposal 1Proposal 2Current RulesCan't Decide <-- original vote final vote 0 {'3': 1, '0': 4} 1 {'2': 7, '3': 1, '1': 50} 2 {'2': 209, '3': 4, '0': 9, '1': 11} 3 {'3': 1, '1': 1} -->
Proposal 158-71062 (21.0%)
Proposal 223311-49216 (72.5%)
Current Rules210-07 (2.3%)
Can't Decide5001-13 (4.4%)

[This just means that the proposal 1 people thought they didn't need to vote, since they were sure it was over the top. --Manus]

Anyway... into it then:


Editor's comment: The first set of replies each deal with an issue with resolution 2 which Simon overlooked when wording his rule:

From Brandon Clarke (bjc@stevensons.co.nz):

Having finished reading the article, I've voted for resolution 1... I've always thought the beleaguered garrison bit happens before the convoy gets through to London, so it's the bit that the resolution of the paradox has to begin with.

I have a major problem with resolution 2.

As Manus points out it requires two exceptions to the rules.

It also creates a third issue neither of you have mentioned:

Rule IX. The support order, 6. Holding and receiving support. Says:

"A unit not ordered to move (i.e. one that is ordered to hold, ordered to convoy, ordered to support, or not ordered at all) may receive support in holding. A unit ordered to move may receive support only for its attempted movement. It may not be supported in place in the event that its attempted move fails."
Simon's proposed resolution takes an Army that was ordered to MOVE (i.e. a unit which cannot receive support to hold) and changes it's status to an Army that was ordered to HOLD (i.e. one that can receive support to hold).

What's with that?

It seems simple enough to fix, and I feel strongly that this must be fixed, or else you'll end up with someone who adds to the situation with

FRANCE: A GAS S A BRE

GERMANY: A PIC S A PAR - BRE
GERMANY: A PAR -BRE

and since the convoyed army now holds, GAS S A BRE is valid, and means that the attack from PAR fails.

I suggest you change the wording of Proposal 2: to

"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, except that that army may not receive support to hold."
Brandon Clarke.
Simon's response: You are correct in what you say (somebody else caught this as well). That was not the intent of the rule, but you are right that that is the way it reads. I will be correcting this in a "response article" in the next issue. My rule meant for the army to stay where it was, but I had not meant for it to be able to receive support, though the current wording allows that. Thanks for the feedback.

- Simon


From Rod Spade (rodspade@acm.org):

I just read your excellent article about paradoxes in the Pouch, and I'm certain you're going to get a lot of feedback on this issue.

Editor's comment: Ya think!??? You ain't seen nothin' yet...
Manus stated that the 1982 rule introduced new paradoxes. I believe that the revised rule does eliminate all paradoxes, if you interpret it as saying that the support is not cut even if the fleet is dislodged. (This is similar to Manus' proposal for modifying the 1976 rule.) The 1982 rule is poorly worded, and it does not explicitly say that it overrides the rule that a dislodged unit can not provide support. However, it makes sense to give the 1982 anti-paradox rule precedence over all the other rules, because it was introduced after the other rules (and therefore supersedes them) and because its wording implies that it is a very specific rule that overrides the general-case application of the other rules.
Simon's response: Your interpretation is in direct contradiction with Section X which states that a support is cut if a unit is dislodged (Or did that section also change in the 1982 rules?).
Rod's reply: That's true, but really *any* interpretation of XII.5 is a contradiction of X - it is the rules themselves that are in conflict, not just my interpretation. Even the 1976 rule is an *exception* to the general precepts of cutting support.
I see no reason for the convoy rule to take precedence over all other rules just because it was introduced after all the other rules.
Rod's reply: Here's another reason: The special convoy rule is more *specific* than the other rules. The other rules govern the general case, and the convoy rule overrides the other rules in the scenarios where it applies. Another argument is that because XII.5 follows X, it overrules it. (You explain the general rules first, then introduce the special-case exceptions.)
Somebody buying the game now doesn't know about earlier rule books. In other places in the rules, if there is an exception to a rule, it is explicitly stated (such as the section on self-standoff). Otherwise, the rules are a "set" and there is not supposed to be any precedence or ordering in their application because adjudication is supposed to be simultaneous. For somebody who does not know about earlier rule books and convoy problems, to accept your interpretation is to accept a direct contradiction between two rules in two different sections, without any explicit indication in the rules as to which should be applied first or which takes precedence over the other.
Rod's reply: Even if you don't accept my interpretation, the rules still contradict each other, and you must choose to give one precedence.
Manus stated that he dislikes the 1982 rule because it forbids the army from cutting even more supports than the 1976 rule - that it can not cut support for or against any fleet in water. However, this is not wholly accurate - the support is not cut only if it is for or against a fleet that is convoying. (Of course, we could argue about the exact definition of "convoying fleet", but that's another matter.) It is more broad than the 1976 rule, but that is not a reason to discard the revised 1982 rule. Any rule that resolves all paradoxes must cover more cases than the 1976 rule, and the 1982 rule is a good solution.

So does the 1982 rule make sense? It implies that support for or against a convoying fleet succeeds or fails before a convoying army is able to cut the support. Note that I am not saying that attacks on convoying fleets occur "before" a convoyed army moves, only that the related supports are resolved before the army is able to complete its movement. Remember that a turn in Diplomacy is a six month ordeal, so if we wish to discuss a "realistic" basis for order adjudications, we should consider the actions of the units to be processes than span the course of six months, rather than discrete events that all happen at a single instant in time. It is not far-fetched to imagine a model where fleets secure support for/against convoys before convoying armies land enough forces to cut support. Some may argue that this violates the "simultaneity" principle of Diplomacy. However, I would argue that even the normal occurrence of cutting support is not truly simultaneous - attacks generally cut support before movements succeed or fail. (All attacks happen "simultaneously" during the six month movement period, but gathering forces prevent a defender from lending support elsewhere before a triumphant army finalizes its conquest of another territory and actually dislodges the defender.) It has always been true that we must consider the effects of some moves before finalizing the adjudication of others, and it is natural to extend this principle to the resolution of convoy paradoxes. If we accept the underlying concept of the 1982 rule, then it makes sense to apply it broadly and literally, which precludes paradoxes.

Simon's response: Even if I buy your justification based on realism (moves taking six months) and I acknowledge the fact that one must consider some rules before adjudicating others despite simultaneity, I still maintain that accepting your interpretation means accepting two rules that are in direct contradiction, and choosing to give one precedence over the other without anything explicit in the rules telling you that one has precedence over the other.

I think that the average gamer (not an avid fan of Diplomacy) would never interpret the rules the way you have, because there is no reason for it to occur to them to do so given the current wording of the rule book.

Rod's reply: That is true. When I first read the rule, it seemed so bizarre that I assumed it must apply only to the scenario spelled out in the example.
The rules do not say "This rule is an exception to the rule on support by dislodged units and takes precedence over it..." so there is no reason to think that it should.
Rod's reply: If we do not allow the rule to take precedence over any other, then we must throw it away entirely. Section X says that an attack (from a space other than the target of the support) cuts support. We accept that XII.5 is an exception that overrules this. XII.5 says that in a certain scenario "support is not cut." It does NOT say "unless the unit is dislodged". The provision about dislodgement is part of X, and XII.5 overrides X. If XII.5 does not override X, then XII.5 *never* applies. How can we cut apart the first sentence of X, and say that XII.5 is an exception to *part* of that sentence, but another part of the sentence overrules XII.5? Without an explicit exception to the exception, we must assume that XII.5 means exactly what it says - the support is not cut. Period. From this point of view, Manus' proposal is merely a clarification of the existing 1982 rules, not an actual change. Of course, that's only my interpretation. :-)
I think it is more likely that Avalon Hill fixed one kind of paradox without realizing they introduced another, than it is that the really fixed everything but worded a rule badly (in contradiction to another rule without pointing out the contradiction, and/or without making a statement of precedence).
Regarding Simon's proposal, if an army's movement would result in a paradox, we pretend that it was ordered to hold. Does this mean that it can receive support in holding? I suspect that this was not Simon's intent, in which case perhaps the wording of the proposal should be clarified.
Simon's response: You are correct. That was not my intent, but that is the way it reads. Somebody else caught that as well. I'll have to modify the actual wording in the rule (but the idea behind the rule still remains unchanged).
I look forward to reading everyone's replies to your article - it is bound to be a lively discussion!

Rod Spade
rodspade@acm.org

From Pre (pre@dalliance.net):

Simon says "The convoyed army is adjudicated as if it were ordered to hold".

I changed my mind during the arguments (while not being completely sure this isn't some test to see how many people do that or not, but then I'm distrusting in general). The thing that finally convinced me was wondering what would happen if the army in Bre were under attack from Par and Pic and supported to hold from Gas. Does this adjudication really act as if the unit were ordered to hold? Would a unit which was ordered to move really still be able to recieve support from a unit to stand still if that doesn't work? I'm sure the rule change could be constructed to work out that problem (assuming Simon even agrees it is a problem) too but as it stands at least it simply doesn't.

Thanks for the great work.

Pre........

Simon's response: Several people pointed this out.

You are correct in what you say (somebody else caught this as well). That was not the intent of the rule, but you are right that that is the way it reads. My rule meant for the army to stay where it was, but I had not meant for it to be able to receive support, though the current wording allows that. Thanks for the feedback.

- Simon

Editor's comment: Simon and I have since discussed this, and Simon came up with the following reworded rule:
"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."
This means the Army has still been issued with a move order, but that the move is unsuccessful, and therefore the Army cannot receive support to hold.

From Robin Walters (r.g.walters@sheffield.ac.uk):

Hi,

Enjoyed the discussion of the two proposed rules. A 3rd possible rule occurred to me (although I don't know if it would deal with all the possible paradoxes):

"A beleaguered fleet cannot convoy".
It certainly deals with the paradox highlighted in the article.

Robin

Editor's comment: The cheek of it!!! ... 1, 2, 3, 4, 5 WORDS!!!! Good grief Robin, Do you realise how many words Manus and Simon generated between them... and you come along and propose a solution that uses FIVE words!? That's just not on at all! You obviously have not grasped the essence of this paradox thing... it's SUPPOSED to generate endless discussion, not simple little fixes of FIVE words!!!

*evil grin*

Perhaps Manus and Simon would like to write a 10000 word follow up article on the pro's and con's of Robin's suggested rule. Rick Desper certainly was more in the spirit of things when he turned up to work one morning and wrote:

From Rick Desper (R.Desper@dkfz-heidelberg.de):

Regarding the "paradox" and the two proposals..., basically, there is no real justification to having the army hold. OTOH, there is a justification to letting the army go through the convoy.

What is this justification? It's a logical outgrowth of the "dislodged fleets cannot convoy"?

I would prefer it if dislodged fleets could convoy. This would remove all the problems with paradoxes and so forth. But probably Calhamer thought that this would make England too powerful, so he put in the rule about convoys being vulnerable to fleet attacks.

What exactly can we infer as a conclusion from this rule? Well, if convoys don't happen at all when a fleet is dislodged, it means that the convoy happens _later_ than the attack on the convoying fleet.

Once you adopt this attitude, everything else falls through nicely. This attitude is consistent with the rule in the book saying that a convoy cannot cut a support for an action which contains a convoying fleet. Why can it not cut this support? Because it's too late! The convoy didn't start until it determined the route was safe. Once it's
safe, that means the naval battle is over.

The problem here is that people who have some introduction to mathematical logic are viewing these events as logical variables which are invariant in time. E.g., somebody says something like "if the convoy gets through, the support gets cut, so the fleet is vulnerable and then the convoy doesn't get through, etc."

But we already know that dislodged fleets do not convoy. What is the inference? If the fleet is dislodged, the convoy does not happen. End of story. If the convoy happens, the fleet is not dislodged. Don't view these things as logical variable. View them as events that happen in time. Is the army sitting in Brest, calculating its impact upon the logical structure of the universe before deciding whether it can sail across the Channel? No! It's sitting there deciding whether the convoy is safe. Once it is safe, the convoy goes, and that's the end of it. The convoy doesn't disrupt its safeness. _That_ would be a paradox.

I really dislike the idea that the army should hold. Replace your example with an example in which the army is going to Liverpool instead. ) Add a fleet in Irish as necessary). Clearly in that case, the army goes through, right? So, why then should it not go through to London? In both cases you have a safe convoy route. In the Lpl
case, the route is clearly to be used. Why, then, not in the second case? Because the army is worried about destroying the logical structure of the universe? Such a profound concern is beyond the bounds of the army. Armies either move when they are ordered to do so, or they don't.

This reminds me of a typical problem with people who think time travel might be possible. Movies have this idea that Michael J. Fox can travel in time as long as he hides in the background and therefore doesn't meet himself at a period in time in which he met himself. But that's not how the universe works. The laws of physics do not depend upon the vagaries of fate. By this I mean, if you let somebody travel in time, you have to accept that he can then "change the past", since the "past" is now the present.

Anyway, I digress.

I think you are framing the discussion not quite perfectly. The dispute here is between two rules, one of which says that a convoy cannot cut support, and the other which says that a dislodged force cannot offer any support.

To accept the Proposal 1, all you have to do is say the former rule outweighs the latter. Jamie Dreier has in the past recommended this interpretation, as the former rule was added to the rule book later than the latter rule. Furthermore, adopting this attitude removes all paradoxes from the game.

The only thing which causes the paradox is the idea that the latter rule should supersede the former! So, really, this is the case in which you have two contradictory rules. If you choose one to supersede the other, you have no paradoxes. If you choose the other way around, you do have paradoxes. The choice should be clear.

To address a few fine points:

"The rule works fine for getting rid of paradoxes. But it asks you to accept that a destroyed unit can successfully support an attack. Granted, a destroyed fleet providing successful support is not a general rule, but only comes into play in a paradoxical situation in order to eliminate a paradox. But for me, saying it applies as an exception rather than a general rule does not make it easier to swallow. To me, a destroyed unit should simply not be able to successfully support an attack, any attack, at any time."

Well, you are adding a paradox to the rules by adapting this attitude. We have already got a rule which specifically says that the convoyed army does not cut this support. We already know that convoys do not go through fleets which are dislodged. The minimal-pain interpretation would be that the time for dislodging the fleet is over, and now we deal with the effects of the convoy.

"I speak only for myself here, and I have no doubt that there will be people who feel more strongly than I about this issue of realism, as well as people who are not particularly bothered by the idea. It's just me."

Well, look through deja.com and search the r.g.d. archive for "Ghost Ships". I wrote specifically on this topic a few years ago. I was defending what you call Proposal 2 and Jamie Dreier was defending Proposal 1. Jamie argued what I say above, a point-of-view that I have adopted. There is no rule book justification for denying the
convoy. If the convoying fleet is there, the convoy goes. If the convoy goes, then it dislodges. There is, however, a rule book justification for not having the support cut. All you have to do is decide that the convoy/support cut rule supersedes.

BTW, you gloss over the whole issue of the difference between the '76 rules and the '82 rules. I have, this week on r.g.d., argued (forcefully, I think) that the '82 rules really are necessary. In other words, you cannot just say that a convoy cannot cut support if the support affects that particular convoy. You have to say it cannot cut support if that support affects ANY convoy.

Here's the example:

England:
F Edi S Ger F Nth

France:
A Bre - Edi; F Eng C A Bre - Edi; F Iri C A Bre - Edi; F NAO C A Bre -
Edi; F Nwg C A Bre - Edi; A Cly S A Bre - Edi

Germany:
A Hol - Lon; F Nth C A Hol - Lon; A Yor S A Hol - Lon

Italy:
F Lon S F MAO - Eng; F MAO - Eng

Russia:
F Nwy S F Ska - Nth; F Ska - Nth

Let C1 = "convoy Bre-Edi succeeds", C2 = "convoy Hol-Lon succeeds"
Let A1 " "attack Ska - Nth succeeds", A2 = "attack MAO - Eng succeeds"
Let S1 = "F Edi support valid" and S2 = "F Lon support valid"
Use ~ for negation, => for logical inference.

Then, if you view these things as timeless logical variables, you have C1 => ~S1 => A1 => ~C2 => S2 => A2 => ~C1 and ~C1 => S1 => ~A1 => C2 => ~S2 => ~A2 => C1

Clearly this is ridiculous. How can we break one of these logical equivalencies? I think we have to leave in S2 <=> A2 and S1 <=> ~A1. And the rules clearly indicate A1 => ~C2 and A2 => ~C1. So, we are left with C1 => ~S1 and C2 => ~S2. They've got to go. Which is why the rules were changed between 1976 and 1982.

(A side note, I, personally, would prefer to remove the A1 => ~C2 and A2 => ~C1 connections. But that would be hard to justify with the current rule book.)

So, to summarize, I think it is valid to view this problem as the problem of choosing which already-existing rule supersedes the other, rather than the problem of deciding how to fix a paradox. If we choose one precedence, there is no paradox. If we choose the other, there is a paradox. Then the problem is compounded, as the solution "the convoy holds" is totally ad hoc, without any rule book justification.

Yes, adoption of Proposal 1 has the downside that you have to accept Ghost Ships. Well, we already have some "unrealistic" rules. Why should it take one season for an army to be convoyed from St Pete to Smyrna when it would take a fleet 4.5 years to make the same trip? How can a power support an army in Africa if the only SC he owns is Denmark? (What kind of magical supply lines are involved here?)

Rick

p.s. Thanks for releasing the new issue today! And I thought I would get some work done. :(

Editor's comment: And if you think Rick did well, just take a peek over at the scroll bar on your browser and imagine how much there is to go!!!!
Simon's response: First off, I agree that if you interpret the 1982 rules as you suggest, there is no paradox. What you are basically saying is that the rule that Manus suggested does not need to be added to the rule book, because it is already there. While it is convenient to take that interpretation, I'm not completely convinced.

The "1982 rule" (saying the convoyed army does not cut support if it attacks a fleet supporting an action into a body of water convoying an army, etc.) is in direct contradiction with Section X which states that a support is cut if a unit is dislodged.

You argue that the former should take precedence over preexisting rules just because it was introduced after all the other rules. I don't buy that argument for two reasons.

First of all, somebody buying the game now doesn't know about earlier rule books. Secondly, in other places in the rules, if there is an exception to a rule, it is explicitly stated (such as the section on self-standoff). In the absence of any statement of exception, the rules are a "set" and there is no explicit justification to put precedence or ordering in their application. For somebody who does not know about earlier rule books and convoy problems, to accept your interpretation is to accept a direct contradiction between two rules in two different sections, without any explicit indication in the rules as to which should be applied first or which takes precedence over the other.

I think that the average gamer (not an avid fan of Diplomacy) would never interpret the rules the way you have, because there is no reason for it to occur to them to do so given the current wording of the rule book. The rules do not say "This rule is an exception to the rule on support by dislodged units and takes precedence over it..." so there is no reason to think that it should.

There are two possibilities that I can see here.

Possibility 1: There was a paradox problem in the pre-1982 rules. They added a rule to fix it. They did not intend for a dislodged unit to be able to provide support, so there was no contradiction with section X of the rules. Unfortunately, they did not notice that they introduced a new set of problems with the new 1982 rule.

Possibility 2: There was a paradox problem in the pre-1982 rules. They added a rule to fix it. Their intent was that this rule should take precedence over section X that says no support from dislodged units. That makes the new rule in direct contradiction with section X, but they do not point out the contradiction in the rules, nor do they indicate that the new rule should take precedence over section X.

As you say above, it is convenient to look at the current rules and say "if you interpret it this way, there is no paradox". However, I simply don't think that is the interpretation they intended. I think they missed the new problem. If this is not the interpretation they intended, then interpreting it that way for convenience means introducing a contradiction that they had not meant to be there. If they had not intended that contradiction to be there, then your re-interpretation of the rules changes the spirit of the rules as much as adding a new rule would.

What it comes down to is what their intent was with that rule. Is it the intent summarized by possibility 1 above, or possibility 2?

To me, possibility 1 is much more likely. They meant the rule the way most people interpret it, and they didn't notice that they were causing new problems. Why do I think that? For reasons I've already mentioned. (1) If they were aware of the contradiction and meant for the 1982 rule to be an exception to section X, they would have said so explicitly as they do in other exception cases. (2) There is nothing in the rules to indicate that one rule should take precedence over the other. If they recognized the contradiction, they would not have left it open to two different interpretations and left it up to people to assume which takes precedence over the other. (3) The argument that the 1982 rule should take precedence over section X because it was the newest rule is completely weak. In 1982 to 1985, what do you think the ratio of boxes sold to first-time buyers vs. people who already owned the game... 10-1, 20-1? They're in the business of marketing and selling games. They knew perfectly well that a large percentage of games sold in a given year are purchased by first-time buyers, and people buying the game for the first time have no idea of earlier rule books. Again, if their intent was that the 1982 rules should take
precedence, it makes absolutely no sense for them not to say it explicitly and to assume that people will know what they mean because it's a new rule.

I think it is far more likely that Avalon Hill fixed one kind of paradox without realizing they introduced another, than it is that they really fixed everything but worded a rule badly (in contradiction to another rule without pointing out the contradiction, and/or without making a statement of precedence). While your interpretation is convenient because it eliminates the paradoxes without adding a new rule, I don't think that's what they intended, and I also think that interpreting the rules differently from what they intended to eliminate paradoxes is as much of a change in the rules as adding a new rule. Either way, your
outcome or mine is different from what they meant.

Now my argument assumes one intent, yours assumes another. The intent you interpret has two justifications. The "newer rule takes precedence" argument doesn't hold water for me. I also think the "that's what they meant because it eliminates the paradox" argument is weak, because if they were aware of the contradiction and that's what they meant, there are plenty of ways it could have been clarified so that it wasn't so open to misinterpretation. I think they patched one problem with a band aid, and didn't realize their bandied was defective itself.

Rick Desper wrote:

I really dislike the idea that the army should hold. [...] Because the army is worried about destroying the logical structure of the universe? Such a profound concern is beyond the bounds of the army. Armies either move when they are ordered to do so, or they don't.
There is another suggestion that has come up -- adding a rule that says a fleet in beleaguered garrison has a convoy disrupted. That results in eliminating paradoxes in all cases, and in all cases results in the same outcome as the rule I suggested. On the positive side, this rule has nothing to do with the logical structure of the universe. On the negative side, it affects *all* convoys by fleets in beleaguered garrison, even in non-paradoxical situations.

- Simon

Editor's comment: This discussion spilled over into other forums, and the result was that numerous others commented on the issues raised above. Here's some excerpts... but I warn you, the discussion goes back and forth, back and forth, and it could possibly turn your head inside out:
From: Jamie Dreier (James_Dreier@brown.edu):

Simon (et al.),

I think there are two reasonable principles to use in interpreting the rules of games. One is the Intent principle: that we should try to figure out what the authors of the rules intended. The other is the Charity principle: that we should interpret the rules so that they make as much sense as possible, consistent with the actual wording (so that vagueness, conflict, unclarity, and ambiguity should be resolved in the way that makes most sense). Both of these principles, I claim, support Rick's and my interpretation of the Convoyed Armies Do Not Cut Certain Supports rule.

Editor's comment: I must butt in here and state that I strongly disagree with Jamie here. I have always argued strongly on this issue that worrying about intent when looking at the meaning of the rules of a game is very dangerous. Interpreting rules is similar to statutory interpretation. When I did law intermediate at university (and I didn't go on to do a law degree, so I'm only too happy to be corrected by any experts in statutory interpretation, or, of course, even more happy to be backed up by them...) I learnt the following approach to statutory interpretation:

When looking at a law, the primary consideration when trying to determine the meaning of the law is the meaning of the words used to phrase the law. Since the people who wrote the law may or may not be available to answer questions about their intentions, we cannot rely on finding out for sure what their intentions were. We cannot attempt to guess at their intentions either, because that leads to a grey area where you could perhaps argue vary degrees of intent, or indeed different intents altogether. The citizen in the street ought to be able to read the law, and based on the meaning of the words in the law, work out what the law is, and so base their actions on that interpretation. It is the responsibility of the law makers to use language accurately when phrasing laws, so that the citizen in the street can do so.

So, what should the law be when the law makers have used language carelessly? Let's say they have used language which means one thing, when the intent was clearly another? In that case the law must be what the law says as it was written, i.e. the law is what the language used actually means, not what the law makers intended. The law makers have the ability to amend the law, and when it becomes clear to the law makers that the law is not what they intended, or, due to the changing nature of language it becomes clear that the common meaning of the language used has shifted from the time the law was drafted to the present day, they should amend the law, using more accurate language, to bring the meaning of the law into line with their intention. In the meantime however the law will be at odds with their alleged intentions.

When applying game rules the same principle must hold. If the game rules say one thing, but it seems the intention was something else, the rule is what the rule says it is, as per the meaning of the language used. Trying to come to conclusions about intent, or for that matter applying Jamie's Charity principle is the wrong approach.

Even if you disagree with my strongly "Rules lawyer-ish" approach here, I think it's clear that as well as Jamie's two approaches of Intent and Charity, there is a third approach which needs to be considered... that of Meaning.

Brandon Clarke.

Charity first.
Interpreting the rule in our way removes paradoxes. Other suggestions either leave paradoxes, or they simply invent rules that are obviously not written in the rule book. (The 'beleaguered convoys are disrupted' rule that you enigmatically mention later is in the latter category.) Interpreting the rules so that they are literally contradictory is about as uncharitable as you can get. So Charity very strongly supports our interpretation, at least until and unless someone else comes up with another feasible interpretation that avoids paradox.

Now Intent.
My best guess is that the authors of the 1982 rule did not notice that the rule contradicts Rule X. It didn't occur to them. So this is *roughly* your Possibility 1, thus:

Possibility 1: There was a paradox problem in the pre-1982 rules. They added a rule to fix it. They did not intend for a dislodged unit to be able to provide support, so there was no contradiction with section X of the rules. Unfortunately, they did not notice that they introduced a new set of problems with the new 1982 rule.
Only 'roughly' because I don't agree that this new rule 'introduces a new set of problems'. The contradiction with Rule X was already present in the old 'Convoyed attacks do not protect the convoying fleet' rule. Also, I think it is somewhat misleading to say "they did not intend for a dislodged unit to be able to provide support," although it is probably literally true. My guess is that they did not *intend* any such thing, but neither did they intend the contrary. It just didn't occur to them.

However, I think this doesn't really matter. It is very clear that the authors of the 1982 rule intended it *to resolve the paradoxes*. They intended to be writing a rule that resolves the Pandin paradoxes. Following the Intent principle, then, we should interpret the rule in such a way that it really *does* resolve the paradoxes, at least if there is any sensible interpretation of the English wording that succeeds in doing that.

When Intent and Charity conflict, I give precedence to Charity, myself. But in this case, I don't think they conflict.

There is a secondary issue: when a pair of rules contradict each other, should the later rule be taken to overrule the earlier one? I say it should. I will not argue for giving the later rule precedence in general, but only for this case.

Charity:
If the 1982 rule, as written, is not taken to overrule the old Rule X, then we still have a contradictory set of rules. This is a most extremely uncharitable reading of game rules.

Intent:
Surely the authors intended that, IF the new rule should conflict with any existing rule left unamended, the new rule should have precedence.

Simon Szykman responds:

I disagree here. You are reading into this an explicit intent for a new rule to have precedence over old ones. I don't agree with the conditional part of your sentence... the whole "IF" part. I think their intent was to introduce a new rule to fix paradoxes (period) without any intent for this rule to have precedence over other ones because I do not think they explored the possibility of conflicts.

I think the way the rules ended up being written implies that they didn't notice the conflict with rule X. If they had noticed, I would think something would have been said about it.

Jamie Dreier responds: I agree that they did not notice it! That's why I'm talking about *conditional* intention.

Imagine that we asked the authors of the rule: On the condition that this new rule turns out to conflict with some earlier one, do you intend for the new one to override the old?

So first they say, "Oh, we don't intend for it to conflict any other rule." "Right," we say, "we know that. Did you hear the 'if'?"

"Oh, ok," they say. "Well then...." Now, how would you complete the response? Do you think they would have said, "*IF* our new rule conflicts with any old one, then the conflict is to be resolved in favour of the old rule"? So, they would be saying that if their new rule conflicts with an old one, then the paradoxes may not be resolved at all. The whole point of the new rule would be defeated.

I find that extremely hard to believe, myself.

Simon Sykzman responds: I don't. I think that their rule was broken, and they didn't realize it. You are saying that if it were pointed out to them, they would say it should have precedence over section X. I disagree. I am saying that if they had realized that their rule was broken, and that the only way to fix it was to allow for the possibility of a destroyed unit providing valid support, that they would have scrapped their broken rule and looked for another solution.

Your interpretation of their rule is the only interpretation of their rule that eliminates the paradox problem. But that rule is not the only one that can eliminate paradoxes. Other rules can be developed that have more benign conflicts with existing rules.

Jamie Dreier responds: Absolutely, other rules can be developed.

BUT THOSE RULES ARE NOT WRITTEN IN THE RULE BOOK!

I may have misunderstood the point of the discussion. I thought the question was what the rule is now. That's why I mentioned the two interpretative principles (Intent and Charity). I am claiming only that those principles are relevant to interpreting what the rules are now. If we're talking instead about what rules there ought to be instead of the current ones, that's an entirely different matter.

You can argue your case based on charity if you wish, but I don't think you can argue it based on intent. Saying "they intended to fix paradoxes and introduced the 1982 rule to do it" is very different from saying "they intended to fix paradoxes and introduced the 1982 rule and intended for that rule to have precedence over other rules." To say that their intent was to have the 1982 rule override preexisting rules in the case of a conflict is unfounded. There is no indication that the latter is the case.
Jamie Dreier responds: The whole point of the 1982 rule is to resolve the paradoxes. That is the intent. There is no possible alternative explanation for the 1982 rule. According to your interpretation, the obvious intent of the rule is utterly thwarted. The rule is rendered pointless. The only thing it was intended to do, it does not do. According to mine, the obvious intent of the rule succeeds. It does what it was intended to do. I conclude that my interpretation is clearly, overwhelmingly supported by the intent.
They tried a fix, and their fix was unbeknownst to them, broken. You are basing your argument on Charity and Intent. But your Intent argument seems to say that (1) your interpretation avoids paradoxes, and that (2) because it does, that implies that they meant for the new rule to take precedence over others in case of conflict when there is no reason to think they considered conflicts let alone having an intent over what the precedence should be.

You are taking their broken fix and giving it an interpretation that fixes paradoxes because you know they wanted to fix paradoxes. I'm not saying your points are unreasonable, but that sounds a lot like Charity to me, despite your calling it Intent.

Jamie Dreier responds: Huh? You admit that their intent was to fix the paradoxes! My interpretation *obviously* accords with that intent! Yours obviously does not!

Maybe you mean something different by 'intent'. As I use this concept, when someone intends that P, and one interpretation makes it the case that P, and a second interpretation makes it the case that not-P, then the first interpretation better accords with intent. According to that standard, my interpretation is undeniably the one supported by intent.

Simon Szykman responds: I agree that their *intent* was to fix the paradoxes. After that our opinions diverge. You take the view that they succeeded, providing the rule is interpreted the way you suggest. I take the view that while their intent was to fix paradoxes, they failed to do so.
Jamie Dreier responds: But hold on. You agree, don't you, that if the rules are interpreted my way, then the paradoxes *are* removed? So don't you agree that if the rule is interpreted in the way I suggest, the authors *did* succeed in their intent ("to fix the paradoxes," as you have just agreed)?
You say that because their intent was to fix paradoxes, if you give the rule precedence over section X, it succeeds. I contend that since a conflict arises that they didn't consider, they failed to accomplish what they set out to do.
Jamie Dreier responds: I explained what I mean by interpreting in accord with the authors' intention, right? If the authors intended that P, then an interpretation according to which P is in accord with the intention, and an interpretation according to which not-P, is not. Since we both say that the authors of the rule intended that the rule resolve the Pandin Paradoxes, and it is indisputable that my interpretation of the rule is the one according to which the Pandin Paradoxes are, in fact, resolved, it follows immediately that my interpretation is the one that accords with the intentions.
I am not convinced that if they did see the conflict, they would done what you are doing... giving the new rule precedence over section X.

You seem to think that if they realized their rule was broken they'd want to keep it. I think there are better solutions and if they had realized it was broken they'd have looked for another one. I don't believe any rule can be developed that will not conflict with existing rules, but I do think that rules can be developed that have outcomes which are more acceptable. I am speaking for myself only here, but the idea of a destroyed unit providing valid support is just too contrary to my view of how the game mechanics work.


It is much more believable that they didn't notice the secondary effect of their new rule (a contradiction with existing rules). You seem to be agreeing that they probably didn't notice the problem with their rule, and in response I think you are being charitable in saying that's what they would have meant.

Otherwise they would not be sure what rule they were introducing, nor whether it would have the effect they were aiming for (removing the Pandin paradoxes).

I agree with you (Simon) that the authors could not reasonably expect the average gamer to know which rules were written or introduced before which others. I figure that if they had realized that the 1982 rule conflicted with Rule X, they would have amended Rule X so that it explicitly contains an exception. (This would be a much more sensible thing to do than what they actually sometimes do, which is to state the exception only in the later rule!) Nevertheless, *we* know which rule came first, so we are in a better position to interpret the rules than the average gamer.

Notice, by the way, that the old "Convoyed attack does not protect..." rule already conflicts *explicitly* with the original rule on Cutting Support. This conflict is not mentioned in any of the rules, new or old. I take it to be quite obvious that the "Convoyed attack does not protect..." rule was supposed to take precedence over the old Cutting Support rule, even though it 'comes later' only in the rules, and not in history. In any case, you (Simon) are mistaken when you claim that

Secondly, in other places in the rules, if there is an exception to a rule, it is explicitly stated (such as the section on self-standoff).
There *is* a conflict between the old Cutting Support rule and the old "Convoyed attack does not protect..." rule, and it is not explicitly noted in the rule book.

Finally:

There is another suggestion that has come up -- adding a rule that says a fleet in beleaguered garrison has a convoy disrupted. That results in eliminating paradoxes in all cases, and in all cases results in the same outcome as the rule I suggested.
I don't know what rule you (Simon) suggested. But this rule about beleaguered convoys does not seem to resolve Pandin paradoxes at all.
Here is an example.

England:
A Hol S A Yor -> Bel
F Lon S F Nth
F Nth C A Yor -> Bel
A Yor -> Bel

France:
F Eng C A Pic -> Lon
A Pic -> Lon
A Wal S A Pic -> Lon

Germany:
F Bre -> Eng
F Nwg S F Ska -> Nth
F Ska -> Nth
F Bel S F Bre -> Eng

There are no beleaguered convoying fleets here, so this suggested rule has no effect. But it is a Pandin Paradox. Of course, our (my and Rick's) interpretation of the 1982 rule does eliminate the paradox.

Possibly I missed some context that explains what you meant.

-Jamie

Simon Szykman responds: The beleaguered convoys rule was suggested by several other people as a solution to the paradox rule. You are right that it doesn't work in the case you mentioned -- it wasn't analysed well enough.

The problem I have with your interpretation of the 1982 rule is that it can result in a situation where a fleet is destroyed, and yet its support is never cut. I just don't find that palatable.

Jamie Dreier responds: If this just means that you would prefer a different way of resolving the paradoxes, I have no problem with that at all. (I would like to see a good alternative, myself.)
Simon Szykman responds: That's exactly what I'm saying. My idea for resolving paradoxes is the following rule:
"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."
This will eliminate paradoxes in all situations.
Jamie Dreier responds: It is not a determinate rule. It is ambiguous. What is the definition of an order "resulting in" a paradoxical adjudication?
I admit that this rule is clearly a bandied for fixing paradoxes. It is a rule that only is active when there is a paradox, and it serves no purpose other than to eliminate the paradox. But I like it better because it does not result in any outcomes that go against the grain of more typical game mechanics (meaning the valid support by a destroyed unit).
Jamie Dreier responds: It doesn't go against that particular rule, obviously, but it plainly goes against other rules!
An article in the latest issue of The Pouch is comparing the two rules and asking people which rule they like better, which outcome they like better. I don't know what the latest count is, but last time I checked, 70-something percent of the people preferred the rule I suggest. We didn't ask people why they preferred one over the other, but I think at least some significant percentage of people were bothered by the same support issue that I am bothered by. I don't think allowing another rule to take precedence over section X is something people are happy with.
Jamie Dreier responds: I find that the large majority of chess players don't like the *en passant* rule.

I must say that I am not at all attracted to alternative rules with clauses like "If a paradox would result, then...." They are *so* ad hoc, and (as I think Rick pointed out) they allow for an unappealing 'action at a distance', they allow global features of the position to affect the convoy.

But that's really just a matter of taste, I suppose.

Simon Szykman wrote: If it is reasonable for you to say:
Surely the authors intended that, IF the new rule should conflict with any existing rule left unamended, the new rule should have precedence.
then is it unreasonable for me to say the following:

Allowing a destroyed unit to successfully support an attack goes so much against the grain of the existing dynamics of the game that if the authors had realized that the only way their rule could fix paradoxes was to allow that, they might have reconsidered their rule?

- Simon

Jamie Dreier responds: Oh, sure. They might very well have reconsidered their rule, and written an entirely different one. I wonder what it would have been.

All I'm saying is that, as a matter of fact, they didn't write a different rule, they wrote the 1982 rule. I think it's a very good idea to follow that one, until and unless some people propose some really compelling alternative, have it discussed on r.g.d., write up a nice specific clear House Rule for Judge Play, and so forth. I have no particular objection to improving upon Avalon Hill's rules.
For example, I like the Judge rule on convoy syntax much better than Avalon Hill's.

-Jamie

Editor's comment: Phew! Maybe there's a mini series in there somewhere? As the Guest Editor Manus, I assume I get a cut on the royalties? Please note Jamie, I let you get the last word in there... you owe me ;o) Now let's see what every other person in the world had to say on the issue:


Editor's comment: The next set of replies all deal with the proposal Robin Walters put up above: "A beleaguered fleet cannot convoy". This solution, while it fixes the specific example given in the article, does not fix all convoy paradox situations. Simon wrote the following to me and suggested we start the section dealing with the "A beleaguered fleet cannot convoy" responses as follows:

Simon Szykman wrote: Several people had suggested a rule saying something to the effect of a fleet in beleaguered garrison cannot convoy an army, as a
fix to the paradox problem. For a while, Manus liked that rule and even added it as an optional rule on the DPJudge. However, while it works for the paradox Manus and I discussed in the previous article, because not all paradoxes involve beleaguered garrison, that rule does not work as a general paradox-eliminating rule. I don't know whether Manus ended up getting back to all the people who suggested that or not, but since numerous people did, it's probably worth highlighting the issue and the problem with that rule.

Jamie Dreier had whipped up an example of a paradox that did not include a beleaguered garrison, and I'd like to thank Jamie for bringing this to my attention.

This rule about beleaguered convoys does not seem to resolve Pandin paradoxes at all. (The example is included above...click here to see it.)

- Simon

From Malcolm Dow (malcolm@dowcorp.freeserve.co.uk):

About Rule 1 of the Paradox article at the pouch.

In the paradox article, you point out that the convoying fleet must be a beleaguered garrison in order to produce the paradox. You could go on to use this fact to further restrict the application of this rule.

Something which has always bothered me about this rule, (which sounds like Jamie Drier's interpretation of the 1982 AH rules), is that it changes some otherwise non-paradoxical cases where a convoyed army is forbidden to cut a support.

EG: your example

E: F wal -> eng ; F lon S F wal -> eng
F: A bre -> eng -> lon ; F eng C bre -> lon ; A yor S bre -> lon

Move the Germans
G: F pic -> eng ; F bel S pic -> eng

Add the Russians
R: A nwy -> nth -> bel ; F nth C nwy -> bel

Under 1970's rules, bel is cut, and wal dislodges eng. Not a paradox in sight. Under your proposed rule, and various discussions of the 1982 rules that I have read on r.g.d, bel is not cut, resulting in the paradox which requires london to be uncut and dislodged.

So add a new clause: "If an army convoyed via a fleet which is a beleaguered garrison attacks a fleet that is supporting an action in or into a body of water that contains a convoying fleet, that support is not cut by the convoyed army under any circumstance (but a convoying army does cut all other supports normally)."

The above clause restores the old outcome in the case with the russians, and without the russians, preserves the paradox resolution that you were discussing,

Malcolm Dow

From Darin McGrew (mcgrew@stanfordalumni.org):

I have a couple comments about the paradox article. (Great article, BTW!) First, I prefer resolution #2 overall.

However, I disagree with the perspective that this resolution produces "an army holding for no apparent reason." It makes more sense (at a gut level) to say that a unit doesn't move when trying to pass through an extremely chaotic region, than to say that a dislodged/destroyed unit can still provide uncut support.
I view this anti-paradox rule as just another way convoys can be disrupted.

Your analysis of F vs EG and E vs FG convinced me that the situation is really a three-way conflict: F vs E vs G. So, from France's perspective, it's F vs EG. (FWIW, England sees E vs FG and Germany sees G vs EF.) In such a situation, a "nothing happens" standoff makes a lot of sense; each force is effectively outgunned. Thus, I prefer the resolution that allows France's convoy to fail in a new way, rather than a resolution that allows France's attack to succeed thanks to support from the destroyed target of the attack.

Darin McGrew

Simon's response: I agree that from a "realism" point of view, you could argue that the army does not move because it's trying to pass through a chaotic region. When I said the army holds for no apparent reason, I meant that from the game mechanics point of view. There is nothing at all in the current rules of the game that would prevent that army from attempting to convoy, and yet we are arbitrarily adding a new rule that forces it to stay where it is. "No apparent reason" was a comment made in the context of the existing rules, not in the context of real life. But yes, from the realism perspective, if this were a physical army trying to physically be convoyed from one place to another, the argument you make is clearly a better justification than "no apparent reason".

Thanks for the feedback!

- Simon

From Brendan McClure (bmcclure@gpu.srv.ualberta.ca):

Now, am I missing something, or could there be a different small rule change leading to a THIRD solution? Suppose London is dislodged with it's support CUT (eliminating the problem of "phantom support"). That means German F Belgium would dislodge French F English Channel, invalidating the convoy.

Why not have the convoy succeed AND have F English Channel dislodged?

I imagine it as London's support is cut once a French soldier lands on English soil taking London, enabling F English Channel to be dislodged. This wouldn't disrupt the convoy, as it has already succeeded. So could a new rule say to the effect that in this case and only this case, a convoy can succeed even though the convoying fleet was dislodged? I admit this solution has it's own strengths and flaws, and I would appreciate your input on Resolution Three.

Brendan McClure

Simon's response: You are absolutely right. You could add a rule that allows the convoy to succeed and the convoying fleet to be dislodged.

As I said in my portion of the debate article, any rule you introduce will introduce some contradictions or inconsistencies with the current rules. Your suggestion directly contradicts the rule that says a convoy fails if one of the convoying fleets is dislodged.

I also said there's no right answer, but that it's a matter of coming up with the rule whose contradictions are most acceptable. Personally, I find this contradiction less acceptable than the one I suggested... I'd rather have an army not be able to move to avoid a paradox than to allow a dislodged fleet to successfully convoy to avoid a paradox when dislodged fleets cannot successfully convoy in any other circumstances. But that's just my opinion. Your suggestion is as effective in eliminating paradoxes as the ones discussed in the article. It just results in a different outcome, and one that I personally find more dissatisfying. But I agree that it works for eliminating paradoxes, and it's certainly worth bringing up in a followup in the next issue of the Zine.

- Simon

From Joe Carl(jcarl@neo.rr.com):

There is yet a third possible board configuration that neither of you considered.

It is possible that the English Fleet in London has it's support cut. The Germany Fleet successfully moves to the English Channel and the Army does not get convoyed.

Manus's response: Yes, this seems to be a popular choice as well. This and modifying Simon's proposal to say "beleaguered fleets cannot convoy."
To which Joe replied: No you can't simply reduce Simon's proposal to that.

Then a non-paradoxical situation would arise and have unsatisfactory results. i.e. What if the English Channel were convoy to Picardy instead. Kinda stupid move but you wouldn't want to say it didn't succeed in this case.

Manus's response: Actually, this is indeed what the proponents of the rule argue. Their point is that the fleet is too busy fending off multiple attacks to be able to convoy an army.
To which Joe replied: Hmmm, I don't like that at all. The fleet is not fending off attacks at all. The British and Germany were fighting all around them while the France deposited the army in Picardy with no resistance.

Gut feel says it should succeed

Editor's comment: I agree with joe here... the fleet is not busy fending off multiple attacks... just like in this situation:

A MUN H

A RUH S A BUR - MUN
A BUR - MUN

A TYR S A BOH - MUN
A BOH - MUN

A SIL S A BER - MUN
A BER - MUN

The Army in Munich is not busy running around fending off three attacks... that's not why the unit in Munich is not dislodged at all... it's because the attacks are all of equal strength, and all bounce off each other, leaving the petrified unit in Munich hunkering down and probably quite amazed that it managed to survive.

This is the most reasonable adjudication that I could come up with that is different than the two you came up with.

It seems to make a gut feel sense to me that the cut from the convoyed army should occur, that the fleet should be dislodged and the convoy does not happen.

This paradox is resolved if you assume all convoys are secondary effects of a battle. If and only if the Convoy fleets are not dislodged will army end up in the proper location.

This seems right on a gut level to me.

Germany isn't working with England or France, but France and England are clearly opposing each other. It makes sense that Germany succeeds because he is unopposed while England and France fail because they are opposing each other as well as Germany.

I believe the end result should be

Germany
Fleet English Channel
Fleet Belgium

France
Army Brest
Fleet English Channel (*Dislodged*)

England
Fleet London
Fleet Wales

Joe Carl

Simon's response: Joe Bronikowski made the same point. For my reply, click here.
It now occurs to me there is even a 4th outcome that could be reasonable argued.

The convoy works and the fleet is dislodged anyway.

The final outcome is

England:
F London (*Dislodged*)
F Wales

France
A London
F English Channel (*Dislodged*)

Germany
F English Channel
F Belgium

This could be argued that if simultaneous results are going to happen then this is the most simultaneous that can be resolved.

This paradox is eliminated by the following rule. If the convoyed fleet dislodged had been supported in instead and would have succeeded, then the convoyed army will succeed regardless of whether the fleet is dislodged or not.

P.S.
I haven't thought deeply on the rules I suggest for the 2 alternate paradox resolutions. But these were my initial thoughts, and felt I should share them and see what comes of it.

Joe Carl

Simon's response: Yikes!

This truly contradicts a fundamental rule of the game which says that convoys are disrupted (i.e. fail) when a convoying fleet is dislodged. I think this rule would throw the entire dynamic of the game out of whack. It removes incentive for defending convoying fleets since the convoy would succeed even if the fleet is dislodged. I think this not just affect tactics (for example it eliminates some well-known stalemate lines) but would hurt the playability of the game.

Comments are always welcome, though I would prefer your previous suggestion over this one.

From Andrew Goff (andrewgoff@hotmail.com):

I recently read your article(s) on the paradox.

Feel free to correct me if I have made a glaring mistake, but I would suggest a third way...

What I'm saying comes down to the fact that, intuitively, I think that the French get London AND the German's get the English Channel.

The change is that the first thing that happens - before any other orders are considered - is the convoy and any support directed to the province the attack is going to. Then everything else.

To put it into a rule situation...

"Any resolution involving a convoyed army's movement takes precedence over all other orders, except for any supports that are written regarding that army is moving to."
You don't even have to specifically reference it to the paradox since it won't effect any other situation in the game.

This assumes that the convoyed army is actually convoyed by a fleet somewhere, not just a "A Syr - Edi" hopeful from a dying turk :-)

What do you think?

-Andrew Goff

Simon's response: I would respond to your mail the same way I responded to Joe Carl (above).
From Jeff Chizever (jchizever@peco-energy.com):

I enjoyed your article titled "Eliminating the Paradox in Diplomacy".

I believe that result 2 (convoy fails) is better but for a different reason (rule) than the one you stated. I don't like the idea that the French fleet can complete its convoy since Germany attacked the fleet with more force. The German fleet may not succeed in dislodging the French fleet, because of the besieged garrison rule. But this should not prevent the disruption of the convoy. A convoying fleet is by its nature using its support to convoy the army. However the rules state that only a fleet that is dislodged prevents the convoy.

I would change the rule for convoying to be: "Even if not dislodged, a fleet that is attacked by more force is not able to convoy." Therefore, the convoy is disrupted and the French fleet remains due to the equal forces attacking it by Germany and England.

What do you think of my solution?

Jeff Chizever

From Glenn R (MyLittleCAR@aol.com):

Greetings and good day to you. I don't play diplomacy much but I think that you have the better solution to the paradox. The reason that I think yours is better is because I don't view the army as holding but as not being able to succeed. Think about a real life example: Two fleets enter a body of water (fleet being more than one ship), each supported by another fleet. Now we have three fleets in said body of water with two fleets gaining support from another. Now the fleet that is convoying has no unit supporting it to stay, however since it is the native fleet it can avoid the other two fleets, but cannot make it through the two fleets. Thereby not allowing the convoyed army to be able to land. The army on the land that is supporting the convoying army cannot offer any support to the naval unit to help it navigate. Now the two fleets that were moving into the sea can't both stay so they leave and the native fleet gets to stay. I apologize if this is confusing, I'm not real good at writing what-ifs. But it I think is more true to real life and to your physics example. (only one body can occupy one space at one time, or something like that).

Any response would be appreciated, even if it is to tell me that I'm crazy to think like this.

Have a great Day,

Glenn R.

From Brian E. Tanner (betanner@usa.net):

Simon, Manus --

Having just read your articles and arguments for each of your proposed paradox-solutions, I just wished to tell you that I liked both rules, but that I preferred the second one (the Army Holds one, Simon's) for a reason entirely different from the ones argued.

It's not a good reason, but it is a reason. I think Diplomacy moves should have some resemblance to real life. In real life, despite the beleaguered garrison rule, a Fleet which had been attacked from multiple directions by multiple forces, even though it was not destroyed, would not have time or resources to successfully convoy an Army. Therefore, I like Simon's rule not so much because of the arguments he makes for it, but rather because it has a more "real life" feel to it. The result is the same either way, but I think the rule is better not because it forces the Army to H instead of move, but because it means that the Fleet, though technically not Dislodged, was under sufficient duress to make it unable to carry out the Convoy.

Just something else to think about!

Simon's response (To Jeff, Glenn and Brian): Your suggestion is along the lines of a rule suggested by other people. That suggested rule was that fleets in beleaguered garrison could not convoy. In your case, you are proposing that as a justification for my rule. In their case, they proposed that as an alternative rule. That actually works well in the situation Manus and I used as an example in the paradox article. However, there are some situations where you have a paradox but you have no fleets in beleaguered garrison. The rule proposed by others breaks down in that case. Similarly, though you didn't suggest it as a rule, the justification you make won't quite work to justify my rule in paradoxical situations that don't involve beleaguered garrison. I believe my rule will still work in those cases, but your justification for my rule may not correspond to the situation the way it does in the example we used.

Thanks for the comments!

- Simon


From Fred Leusch (fred.leusch@unb.ca):

Hey Manus,
Just read your article (and Simon's) about the convoy paradox. I think there is a third possible arbitration of that solution ... here it is:
The French convoy from Bre is broken because the French fleet in ENG is dislodged. It doesn't matter that the German and English fleet bounce, the ante for ENG that turn is 2 units, and the French fleet has no support, so it cannot keep ENG. Do you see what I am trying to say?

Cheers,
Fred Leusch

From Stephen Szikora (stephen.szikora@sympatico.ca):

I'm not an expert at the rules and have only just begun internet play after not having played the game in 20 years. However, this may give me a clearer view. It seems to me that the intuitive result is neither option 1 or 2 as described. I would think that where 2 powers standoff against a third and each could clearly have beaten the third (as here in the channel) that not only should the convoy fail, but also the convoying fleet should be dislodged (in this case leaving the Channel unoccupied.) My reasoning is first, that the convoy must occur before the attack, and secondly, that it makes little sense to allow the incumbent to remain after being overwhelmed twice from 2 different directions. What the rules would like implementing this solution, I have not worked out, but they may offer fewer problems and/or objections than those encountered by you first 2 options.

Stephen Szikora
Guelph, Ontario, Canada

Simon's response: You are saying that since Germany and England both could clearly have beaten France, France should be dislodged leaving Eng unoccupied. The first problem is that this directly contradicts what the rules say about standoffs and beleaguered garrison. However, the other suggested rules for eliminating paradoxes also have their own contradictions, so that alone can't be the basis for not considering a different rule.

However, the problem with your suggestion is that it concerns a situation where a fleet is in beleaguered garrison, and there are some paradoxes that arise in situations that don't involve beleaguered garrison. Your suggestion therefore doesn't work as a general rule for eliminating paradoxes.

Thanks for the comments though!

- Simon

From Eric Hunter (Elladan915@aol.com):

Manus and Simon,

Thank you for the convoy paradox article, it was an interesting read. I voted in favour of the second resolution, but I would phrase the exception rule somewhat differently. Since a supported attack on a convoying fleet would result in the convoy failing, due to the convoying Fleet being dislodged, and being unable dock to allow the Army to embark, I would argue that Rule XII.3 DISRUPTING A CONVOY (AH '82) should be revised to read: "If a Fleet ordered to Convoy is dislodged during the move, OR IS SUBJECT TO MORE THAN ONE SUPPORTED ATTACK, BUT NOT DISLODGED DUE TO THE BELEAGUERED GARRISON RULE, the army to be convoyed remains in its original province, and has no effect on the province to which it was ordered."

This would resolve the same way as Simon's rule in the paradox situation, but would actually result in more failed convoys, since eliminating EF Wales, but adding IF Iri S IF MAO -> Eng, would also cause the French attack on London to fail. This seems to eliminate the paradox, and produce a consistent rule set, though. Of course, I could easily be overlooking something here.

Respectfully Submitted,
For Your Consideration,

Eric Hunter.

Manus' responses: Wow. Yours is the fourth (I think) e-mail to suggest that the rule be changed to say that "beleaguered fleets may not convoy." I haven't
taken the time to see if this handles all paradoxical cases, but it does seem to incorporate Simon's solution in a slightly different way,
and therefore it should.

Simon, what do you think?

Simon wrote: Yes, I think it will handle all paradoxes because all paradoxes that I can think of are a involve support cutting (or not support cutting) in a situation where one result causes a disrupted convoy and the other causes a beleaguered garrison.

I think it will result in the same adjudication as my suggestion in all paradoxical cases. As has been pointed out, it will also result in other convoys failing that do not fail under the current rules, in situations that are not paradoxical. While that may be seen as a negative, it can also be seen as a positive point that this rule is always in effect, rather than being a magical rule that only applies in paradoxical cases... a drawback of both of the rules that Manus and I suggested.

Of course, what the beleaguered garrison rule does suffer from, which both the other rules suffered from, is a direct contradiction with the current rules. Specifically, the rule in Section XXI.3 that states "An attack on a convoying fleet which does not dislodge it does not affect the convoy." So one still has to accept a rule that contradicts the existing rules. It'll again be a matter of what people find most palatable, or least unpalatable. From the elegance point of view, I actually find this rule more appealing than mine. But I still don't know how I feel about a general rule that essentially says a convoy is disrupted if a fleet is in beleaguered garrison, when the rules say that only a dislodged fleet causes a disrupted convoy. I'll have to think about it some more.

- Simon

To which Eric replied: Clearly, XII.3 needs to be rewritten, then. ;^} Something like:
"If a fleet ordered to convoy is dislodged during the move, or holds as a beleaguered garrison, the army remains... An attack lacking the strength to dislodge a convoying fleet, does not affect the convoy."
From Tim Yocum (darkwolf@communitygate.net):

Here is an idea that crossed my mind when I saw your discussion on paradox resolution: It is based on the definition of support and cutting of support.

Why is support cut when a supporting unit is attacked by an outside unit?

Besides providing a darned nice way of creatively reducing support for an area you want at the moment, and thus enhancing game play, I believe I remember correctly that it is analogous to the fact that a unit is too busy defending itself from said outside attack to support some OTHER unit.

Well, think about the convoy. It is under the same sort of duress when it is dislodged. Thus its convoy is cut. That is easy to see. And the rules have decided that a single attack that does not dislodge is not enough duress to successfully stop a convoy. Makes sense. Convoys tend to be armed convoys, and thus some ships are used to transport and others to defend. Easy enough.

But in this case, the paradox comes from the inclusion of the Beleaguered Garrison rule. France is standing with only 1 fleet there, and had England or Germany attacked alone, that fleet would have been dislodged and the army never convoyed (tho some might question due to Pandin's, if England had attacked alone, did the French army make it across to cut the support? :) But that is not this case).

In this case you have England and Germany offsetting each other, and thus the Beleaguered Garrison rule kicks in and says France's fleet is saved. Since the Beleaguered Garrison rule is in fact a sort of "special case exception", then in fact, modifying IT, instead of the basic fundamentals might not be a bad option. Thus in this case I would modify the Beleaguered Garrison rule to include the following:

"If the unit being beleaguered is a fleet convoying an army the fleet is NOT dislodged, as per the B.G. rule, but the convoy fails." This can be analogous to the convoy being overwhelmed and having just enough troops to keep control of the waters, but not enough to actually move units across it.

What say ye both? Do I have a suitable option 3? :)

I look forward to your responses :)

--Tim Yocum

Manus's response: A few other people have suggested your rule idea -- that beleaguered garrisons may not convoy. In fact, it seemed (Simon?) to implement Simon's solution so well that I have in fact implemented it as an optional rule in the DPjudge.

Manus

From Thomas S. Mowle (Thomas.Mowle@RANDOLPH.AF.MIL):

Simon

I am not personally happy with the idea of a beleaguered fleet being able to convoy at all. I understand that a beleaguered garrison may be able to hunker down (batten down) and hold its ground if only because the forces arrayed against it are expending some effort fighting each other for local superiority. Manus' rule implies that a fleet can embark an army, carry it under a crossfire to a defended coast, and deposit it there safely. Your solution -- either the army never is able to embark because of the enemy pressure, or the fleet is "bottled up" close to the home coast -- makes a lot more sense to me than Manus' does. Perhaps this could take place if the convoy faced only pressure from one fleet of each power. But for the paradox to occur, at least four fleets would have to be fighting for that SLOC. That seems a bit much -- using Manus' own criteria of "gut feel."

Thomas S. Mowle, Maj, USAF

Manus's comment: Just to let you guys know, the DPjudge now has an optional "rule" you can set that says that armies cannot convoy through beleaguered fleets.
From Simon Szykman (ss8c@andrew.cmu.edu) replying to Alexander K. Woo (awoo@math.berkeley.edu):

Alexander,

Sorry for taking a while to reply to your message. We've gotten quite a volume of response to the article so I decided to wait until the weekend to reply.

Alexander K Woo wrote: I find it interesting here to look at Pandin's Paradox proper, which has two different, equally consistent adjudications. That can be constructed from the given situation by adding a new unit and order:

English F IRI S F WAL - ENG

What is interesting is that both proposals (assuming that this s also a "paradoxical adjudication" under rule #2), the French fleet in ENG is dislodged. This does not have to be the case - there are 3 alternative adjudications here - one is that Eng F LON is dislodged (A), one that Fr F ENG is dislodged (B), and one that no unit moves (C).

This leaves 6 possible different resolutions.

It would seem unreasonable for F LON not to be dislodged in the first case (without Eng F IRI) but to be dislodged in the second (where England has an extra unit in the situation), so that leaves 5 resolutions.


I'm not exactly sure how you are "counting" resolutions. As I see it, there are four possible outcomes. Either the army gets to Lon, or it remains in Bre. Either F Eng is dislodged or it is not. So the possibilities are: (1) Army stays in Bre, F Eng is not dislodged, (2) Army stays in Bre, F Eng is dislodged, (3) Army reaches Lon, F Eng is not dislodged, and (4) Army reaches Lon, F Eng is dislodged.

But aside from that, let's look at rule 1 and rule 2 proposed by Manus and I. With my rule, there is no problem with adjudicating the above situation. The army in Bre stays where it is, therefore it never hits Lon, therefore Wales attacks Eng with strength three (against German attack of strength 2) dislodging Eng. No problem... no bizarre results, no paradox. With Manus' rule there is also no problem. The army trying to reach Lon cannot cut Lon's support, therefore Wal is attacking Eng with strength three no matter what, the attack succeeds, Eng is dislodged and the convoy fails. Both rules succeed in eliminating the paradox and in this case both leave the army in Bre. That is option 2 in the above list of four options.

I don't have any problem with that outcome. If you think a different outcome makes sense, the trick would be to come up with a rule that gives you the outcome you want and also is a successful paradox eliminating rule that works in other situations such as the one presented in the article Manus and I wrote as well as other classical paradoxical situations. Ideally, that rule would be simple and concise, and any of the inevitable contradictions with the current rules will be minor enough so as to be
acceptable.

You seem to be posing a question, but I'm not sure that you are expressing any dissatisfaction with the current outcome. Personally, I think that a rule that results in option 4 (allowing the convoy by a dislodged fleet) wouldn't make sense. We've already got two paradox eliminating rules that result in outcome 1. If you that outcome 2 or 3 in the list of 4 makes more sense, and you can come up with a paradox eliminating rule that produces that outcome, then we can it up for comparison in this situation and others and see what seems best. I say "seems best" because as I said in the Zine article, any rule you introduce to eliminate paradoxes will contradict the current rules, so there is no "right" answer... it's just a matter of deciding which produces the most appealing outcomes in a given set of situations. And of course, there will never be 100% agreement, so the best you can hope for is that some kind of consensus emerges.

For consideration, also look at the same situation where France has another unit:

French F MAO S F ENG

Does your opinion change?

This doesn't really change the situation. As before, with either rules 1 or 2 Eng is being attacked by England with strength 3, the attack succeeds, Eng is dislodged, the army stays in Bre. The rules work... the question is whether you think some other outcome is more right than this one, and if so, whether you can come up with a simple paradox eliminating rule that results in that outcome.

- Simon

From Charles-Antoine Allain (caallain@arobas.net):

Hello,

I read the paradox article with great interest. My gut reaction was to vote for solution no. 2, but I grew dissatisfied with this choice, especially that I did not find that the case made in the explanation was very good -- as the author himself seems to acknowledge. Despite the immense respect that both authors inspire me, I will dissent and offer a third solution, which I think follows the spirit of the rule more closely that the 2 other solutions.

If I were to rule on that, I would let Germany enter the English Channel, disrupting the convoy. And before I explain why, I would like to comment on both arguments presented.

I believe many Diplomacy rules can be seen as solving a paradox or problem. And they have been written with the spirit of the other rules in mind. For instance, rule no. XII.5. (2nd edition), upon which Mr. Hand draws extensively for his argument, is an extension in spirit of Chapter X. It only states that a convoying army, for the purposes of cutting a support, must be viewed as attacking from the space into which the convoy is located. This can easily be understood intuitively, since otherwise the supports are always cut by an attack from a space adjacent to them.

But on the contrary, I would argue, not unlike Mr. Szykman, that the proposed change, an extension of XII.5., goes directly against the spirit of Chapter X regarding dislodged units. Since I think that, all things being equal, Chapter X is more important than article XII.5., I think this is not admissible, or that at least it considerably weakens his argument.

I am not more sympathetic with the Mr. Hand's unit counting approach. First, I have to point out that convoying units should not be counted for this purpose, since they add nothing to the force of the attack. Should France be convoying its army all the way from Tuscany with 4 fleets, it would not make his position any stronger. In fact, article XII.5. itself is a good example of this. But anyhow, Diplomacy is full of examples where a player with fewer units is able to standoff a player with more units -- one of the signs of a good player indeed, or at least of a good position. I argue that this kind of point should not be used to settle this matter; I do not find anywhere else in the rules a trace of such a justification having ever been used.

As I said earlier, I do not think that the second proposal's case is much better, although I did vote for it. First, I do not like the way the new rule is phrased : "a paradoxical adjudication" is a very vague phrase. All of the other Diplomacy rules are worded quite precisely, indeed that is why none of the two solutions can fully satisfy them. Were they couched in such a language that could lead to interpretation, one would get rid of paradoxes, as it would probably be the case that any situation would be covered by one or more rules, but it would greatly multiply contradictions between the rules. Hence, we might very well find that the proposed new rule would not be satisfactory in another situation.

And, as the author points out, there is no example anywhere in the rules of an action being annulled because it could create a weird situation. It is not France's fault that things happened the way they did, and I feel it is both risky and unnecessary to create a whole new game principle that would negate him the expected benefits of its actions. In fact, it is possible to do just that relying on existing principles of the game, as I will show in a moment.

As I said earlier, my proposal would be to rule that Germany should enter the English Channel, thus disrupting the convoy. I could extrapolate on example 10 of chapter X to argue that the support is cut by the planned attack even though the convoy is itself disrupted to justify myself. However, I feel that this would introduce a additional element of temporality and conditionality in the game, and I find that I would not have a very good case.

But there is another very important principle in the game that amply justifies my proposition, and it is contained in rule IX.3. Indeed, one could easily argue that the London support is but an elaborate form of self dislodgement. The situation is presented quite obviously in propositions 1 and 2 : either the London fleet's actions causes it to self dislodge, or we need to create a new game principle to resolve the matter.

Since creating new game principles should be avoided when possible, and because self dislodgement is explicitly prohibited, I find that the London support is void in and of itself. The English Channel fleet is not a beleaguered garrison, since the German forces outnumber the English forces. The German fleet moves and the convoy is disrupted.

One could argue that the position of proposition number 2 could still be another outcome of this proposition on the basis that the London support could "still be valid for other purposes". I personally think that you allowing the fleet to remain means that the London unit is self dislodging (c.f. article XII.3.) and that there is only one purpose here which is to prevent that. However, I would be very open to hear what people have to say on this.

In fact, I would be delighted if you found my proposal interesting enough that you would take the time to answer me. I enjoyed thinking about this problem and I hope you enjoyed my thoughts too. Please excuse my lousy writing, English is not my first language.

Charles-Antoine Allain

From (steinker@cs.colorado.edu):

I just read the article about resolving paradoxes. Here are the points that I found to be most significant or convincing.

The support of the London fleet must not be cut. Both proposals agree on this.

The first proposal argues that it doesn't add a second exception while the second proposal does, but I don't think that's true. Both proposals add a single exception.

proposal 1: Destroyed units can lend support in this special situation.

proposal 2: Armies ordered to move with nothing preventing their movement can be made to hold in this situation.

My first gut reaction was that the French army shouldn't move, but when I read the rationale for the rules I changed my mind. I think I would have liked that resolution if there were something stopping the French army other than "to avoid paradoxes."

There are two reasons I like why the French army should move. One game reason, and one realism reason.

The French army attack on London should be considered as coming from the English Channel, and so shouldn't cut support. All it needs is a small exception that the support isn't cut even though the fleet is destroyed.

I don't mind a destroyed fleet lending support in this situation because of a realism reason. The London fleet is not being distracted by some other battle, it is being destroyed by fighting in the same battle where it is lending support.

The Wales and Belgian fleets sail out into the English Channel and the London and North Sea fleets stay nearby to help. The French fleet picks up the French army and starts sailing for London. Somewhere in the middle they all meet. They are three way adversaries and everyone is firing at everyone else so the French fleet doesn't get run off. The London fleet fights the French fleet as it sails up to the coast and drops of ground troops.

As for realism I see this as a single battle off the London coast. The French have enough armies to drive English ships away from the ports, but those ships still get plenty of chance to fire at other ships in the English Channel before having nowhere to run and being destroyed.

From (alkaline13@yahoo.ca):

Once again, the Pouch is a valued part of my year. Too bad it only comes along only five times...

In regards to the article by Manus Hand and Simon Szykman entitled "Eliminating the Paradox", perhaps a simpler, common sense approach is needed to solve the paradox. I agree with proposal 2 (French A Brest holds). We too often forget that Diplomacy is a war game based (loosely) on historical events. If the events described in any given game were to have actually happened at some point (such as during WWI), would the French convoy actually be able to slip through the Channel to English soil, what with the battles going on between English, French and German navies in that area? I would imagine that there would simply be too much conflict in the area to allow the French convoy to pass. Therefore, I feel the convoy should fail. But that's only my perspective.

keep up the good work!

From Joe Bronikowski (jbronikowski@software.rockwell.com):

I disagree with both your proposed resolutions, and therefore with your subsequent paradox eliminating rules. I would say that France's army Brest does cut London's support of Wales - English Channel (because Wales is not attacking the army), even if the army doesn't successfully attain London. Therefore England fails to bounce the German supported attack, so German Fleet English Channel is the proper end result.

I think I can boil this down to a single, simple rule (that admittedly may effect non-paradoxical situations too):

A convoyed army DOES cut a support order for a unit in it's destination province, even if the convoy is foiled by dislodging one of the convoying fleets.
All other results should follow from this. The only case where this is not true is when the supporting unit is supporting an attack on the ARMY's source province, because that basic Dip rule takes over (a unit can't cut support on an attack of it's own province).

What I don't know is how the Dip rules or the judge implementation handle disrupted convoys in non paradox situations. I have a feeling that if the convoy doesn't go through, it can't cut a support. But the general rule that a dislodged unit can still cut support ("before" it's dislodged) seems to align with my premise that a disrupted convoy can still cut support.

Joe Bronikowski

Manus's response: So you are saying whether or not a convoy fleet is disrupted, the moves should assume the convoy to succeed for the purposes of cutting support of any unit at the destination?

I will have to think about this one....

Yes. Actually I think the addition of the German fleets helped crystallize my thinking on the matter. While England and France are engaged in a 1 1/2 vs 1 1/2 paradox, Germany's two units should win the day :-)

Joe B

Simon's response: I'd be interested to get comments from other people to see if this rule holds up under other paradoxical situations. I have a nagging feeling that this rule will solve the paradox in the situation Manus and I discussed, but that it will lead to a new paradox that doesn't exist with the current rules, in a different situation. I haven't come up with an example of a paradoxical situation under this new rule, but I just have a feeling that such a situation exists. I'll think about it some more.

Assuming it does hold up, the problem I have with it is how to justify the rule at all. Either the army reaches its destination, or it doesn't. Quite clearly, the convoyed army never reaches Lon. One reason it doesn't is that its convoy was disrupted. The second reason it doesn't is that if it had, it would successfully take Lon since it had support for the attack. So given that the army never reaches Lon, on what basis do you argue that the convoy managed to cut London's support?

It's quite clear that this rule is not quite reconcilable with the current rules without some kind of contradiction, but then again, neither were my rule and Manus' rule. It again comes down to the point I made in my portion of the paradox article. Which contradiction are you most willing to live with to make the problem go away.

Possibly a more significant problem is the fact you already acknowledged above--that this rule will affect other situations in the game. I think that a general rule which allows armies whose convoys were disrupted to still successfully cut support at the destination end would certainly change the resolution of orders in numerous types of circumstances. The true effect of this rule would be difficult to judge ahead of time. At the very least, it would change the tactics surrounding situations involving convoys. Whether such a rule would affect the general playability of the game would remain to be seen.

- Simon

To which Joe replied: Well, I'm not the "deepest" thinker when it comes to designing games. It just seems to me that a convoyed army ought to have the same strength as if the fleet itself were attacking. If the fleet can cut a support without attaining landfall, why can't a convoyed army do the same? I'll leave further investigation in your hands....


From (cgv@penn.com):

Excellent article, and you both have arrived at acceptable and workable resolutions. I would not object to either one if I were a player involved in the paradox and you adjudicated it in the manners you described.

Personally, I would tend to fall back on the "realism" idea when resolving this paradox. The fact of the matter is that an amphibious operation is one of the most complex tasks that a military force can undertake. It seems to me that, given this complexity, it should be more easily disrupted than a simple cross border attack. Therefore, in the paradox discussed, I would give the benefit of the doubt to the London defenders, under the guise that the sudden arrival of FOUR hostile fleets would certainly have
interrupted any real life amphibious operation.

From Marc Léotard (marc.leotard@fucam.ac.be):

Actually, the outcome which I prefer is neither of the two you propose.

I consider the French (Yor-Lon) support to be effectively "tying" the activity of English F Lon, so as to prevent its effective support to F Wal.
Hence, the English try into the Channel is too weak to prevent the German to succeed.

So, I consider the German fleet to have succeeded into Channel! French fleet dislodged, situation in England unchanged.

Manus's response: So you are saying that the Yor support should be considered like an attack, such that it cuts the London support? How would you state this as a rule?
I wouldn't say "support is like an attack" in all cases. I would just say that, *when the rules are not enough* (that is, they cancel each other out and lead to paradox), then the side with a little more force should be given credit for it.

Surely, a supporting army does a little of fighting. Paradoxes are not due to convoys, they are due to the mix of convoys, supports and fleet dislodgement. I have once written an article (in French, for postal zine Mach die Spuhl!) where I analysed the essence of paradox and different ways to come out of it. Basically, the culprit is the possibility of negating an order, and the vicious circle it creates: Attack negates Support, Support (by allowing Dislodgement) negates Convoy (and therefore the initial
attack). So you have to bend one of those rules a little. The French version of Diplomacy (published by Descartes) chose one way out (remove simultaneity). The 1992 AH rules chose another (convoyed attack does not cut). I discussed another yet (no dislodgement necessary to cut convoy). I ended up thinking AH92 to be the best, though some gaps remained.

Now, your paradox also exists when there is no Yorkshire support. AH92 says "support not cut". I agree, my humble opinion is in favour of Standoff: Fleet Channel remains in place because of an equality of forces. Descartes or my own try would yield no definitive solution.

The situation you propose is more complex, because: Att negates Sup, Sup negates Disl, Disl negates Conv (i.e. Att). AH92 does not solve it, because it doesn't mention supported convoyed attack. Just a matter of making it explicit: either it also fails to cut the support, or it works. I feel, in the general standoff created, that A Yor is left to add a tiny supplement of weight, that effectively crumbles the whole position.

Here, it actually leads France to dig his own grave, but other examples exist, without paradox, to the same effect. Blame the French player, not the ruleset.

NOW! How to write it as a rule? I dunno. Though I think it should be easier to write in code than put in a rules manual...

Marc Léotard

From Eric Amtmann (ewa@mindspring.com):

Dear Manus and Simon:

You've hidden the best argument in a postscript at the end!

I think Simon is close to a good solution with his "Paradox Resolution Proposal 2." However, as you state, both Proposals 1 and 2 do result in a contradiction to the existing rules that would have to be accepted in the event of a paradoxical scenario occurring. However, I do think there is an alternative proposal which prevents the paradox, but does not contradict the existing rules. Even before reading Simon's "Postscript" and thoughts about "time travel" I had started to think about what constitutes a "valid order." A related way to phrase it is to say an order is "void."

The Diplomacy rules are full of descriptions about what types of moves are allowed for the two types of units. By all accounts in my mind, and by a wide margin, adding one more of these rules is the "least distasteful" solution to eliminating paradoxes. In fact, it is not distasteful *at all* because it does not raise any "special exceptions"
or "contradictions to existing rules."

My proposal is related to Simon's postscript. However, I think if you discard the time travel stuff, a very legitimate argument can be made for this sort of new rule, an argument that can go far beyond simply saying, "paradoxes can't occur."

Proposal 3:

"Armies may move by convoy unless the move creates an adjudication scenario with multiple conclusions. Such a convoy order is invalid, and rendered void."
From there, the existing rules take over on how to adjudicate. This new logic corresponds with the existing rules in two ways:

Example 1:
Army Munich -> Switzerland

In FTF, this order would be invalid, but perfectly acceptable to write down. Army Munich would *not actually move* to Switzerland, but as it attempts to move, would not be able to receive support to hold. In contrast to Simon's Proposal 2, there is no need to retract the *move* order and replace it with a *hold* order. Replacing an order in this manner could result in an army receiving support to hold when in fact it attempted to move. It is simply invalid, and is processed as such.

Editor's comment: As stated in the first set of replies to the article, Simon acknowledges that the wording for his rule has this problem with it, and has proposed a reworded rule which eliminates this problem.
In PBEM, the situation is somewhat different with how the judge handles "invalid" orders. With Manus' DPJudge, the order would simply be rendered invalid at the time of adjudication as with all other invalid orders. With the Ken Lowe judge, this "invalid paradoxical convoy" order would also be rendered invalid at the time of adjudication. Obviously, since the moves of other players need to be known before the order becomes invalid, the order could not be refused by the judge at the time of submittal.

Which brings me to Example 2. Why is it logical and consistent with the existing rules to render a convoy order invalid or void only *after* all other orders are known?

Example 2:
Army Munich -> Ruhr
Army Burgundy SUPPORT Army Munich HOLD

In this example it's obvious that the support order for Army Burgundy is void. But the key point is that the order is only rendered void *after* the order for Army Munich is known. Voila! There is precedent in the existing Diplomacy rules for determining the validity of an order in consideration of the complete set of orders for all units on the board.

Counterpoint:

Perhaps the biggest argument against this rule questions the justification of rendering the convoy order invalid. As Simon states in his postscript, does the proposed rule simply say, "Well you can't order that, because, well you can't, because it creates a paradox." ? It could be argued that in the supporting examples I used above, the rules
in question have logical explanations for their justification.

Example 1 has its justification in the fact that in the early 1900's Switzerland was for all practical purposes physically impassable for large scale military operations. Therefore, orders to move armies into Switzerland are invalid.

Example 2 has its justification in the fact that moving armies can't receive support to hold. This in turn is justified by the logic that if a unit is attacking somewhere else, it shouldn't be allowed to defend itself with more than it's own defensive strength - it shouldn't get help it wasn't prepared for.

At their core, the Diplomacy rules are founded on a framework of logic that is consistent, and sensible. Once you've got the general hang of how the rules work, you've got a fair chance at figuring out a rule you may not be familiar with just based on the system of logic established elsewhere in the rules. In short, the rules are beautifully simple because they are predictable and they *make sense*.

That is my response to those who would argue ruling paradoxical convoys to be invalid is arbitrary. Quite simply, paradoxical convoys don't *make sense.* That alone is sufficient reason to invalidate them based on the system of logic constructed through the entirety of the Diplomacy rules. I submit that rendering paradoxical scenarios as simply disallowed is quite in character with the overall logic structure of the Diplomacy rules.

Conclusion:

With this new rule, a convoy order may be rendered void during adjudication when the orders for all units on the board are known. If the convoy order is rendered void, the army which receives the "invalid paradoxical convoy" order attempts to move, cannot receive support to hold, but never actually moves to the destination province, and therefore does not cut support of any kind in the destination province.

I think this proposed new rule fits into the logic of the existing rules, does not contradict any existing rules, does not create any new exceptions, does not change the submitted order for any unit, and works in all cases.

Simon's response: I think your suggestion of making the army's order void is effectively the same as my revised rule. Even though it's worded quite differently, it produces the same outcomes in the same situations. In other words, we each said what we meant differently, but we both advocate the same result in the given paradoxical situation.

Possibly the one advantage of your rule over mine is that it does not change the order for any unit. I like that aspect of it as far as wording goes. The one potential drawback to the rule is the possible problems in hand-adjudicated games in determining how to treat a unit that has a voided order. The Diplomacy rules indicate that a unit that receives an illegal order stands in place. It is not clear whether that means it holds (so that it can receive support) or whether that just means it doesn't move but if it attempted to move it can't receive support. I think the more accepted ruling is that a unit ordered to move can't receive support even if the order is void or illegal, but the rules themselves are not 100% unambiguous.

As written, your rule runs the risk of being interpreted by a GM as a unit that stands (as per the rules) and interpreting that to mean it can receive support even though the more common interpretation is that it can't receive support. I think your rule should make it clear that the order is void but that the moving army cannot receive support since it is attempting to move. Then it works out fine.

- Simon

From Steve Araps (saraps@erols.com):
I just read your joint article from the DP Zine on Resolving paradoxes. I prefer Rule 2 to Rule 1, but my thinking on all this is a little bit different.

First of all, I absolutely hate the AH rule to limit paradoxes, you know the one (XII.5):

"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."

This of course is in conflict with the support cutting philosophy evident throughout the rest of the game. I am actually a little surprised that the Rule 1 argument didn't state that there are no paradoxes with the AH rules. The above XII.5 *handles* Pandin's paradox. In your example, a convoyed army (bre) attacks a fleet (lon) which is supporting an action in a body of water (eng). Since the fleet in eng is a convoying fleet, the support is not cut per the rules. Convoy succeeds, lon is dislodged/destroyed yet its support is uncut. Ugly, but it is a solution AH came up with.

Unfortunately, this paradox medicine is also applied indiscriminately to non paradox situations as well, creating nutty results that never were controversial in the first place. Matt Self's article has a good example of this. So I hate the AH rule and I hate Rule 1 as well because all this one does is clarify that the AH treatment applies even when the supporting fleet is dislodged/destroyed. Add to this the idea that a certain attack cuts support depending on the type of support. That kind of thinking is foreign to my concept of the game and its movement. In my view a support is a support, there are not different types, some immune to being cut, others not.

Imagine the GM's dilemma for this adjudication...

Italy: wed c Tunisia
Italy: ties s wed
Italy: pie s tussle

France: spa(sc) s Leo
France: Leo c anything anywhere (let's say marts, but mar is empty)

According to the rules, everything bounces though most people would be more conceptually satisfied if the spa support was cut and Leo dislodged. I know I would be. Is this example contrived? Of course, but it could happen and here's the why. Italy is worried that wed could get dislodged (assume F fleet mao in the scenario for this threat), so to avoid it he delivers the support cut with a convoy. France sees the danger of the convoyed support cut and the futility of trying to defend Leo or attack wed conventionally. So he sets up the quasi-uncuttable support by issuing a fictitious convoy order for Leo. Pretty stupid, huh?

Some may question whether wed should have the privilege of delivering a support cut while allowing itself to be protected by a support in place. I don't. It needs two units to accomplish this task (itself and tun). Also, like it or not, this *is* the game treatment. If spa were ordering spa s mar, there is no question the support is cut. Why the special treatment for the support of Leo? The answer is simple: to try to find a way around convoy paradoxes. Period. This strikes me as a bad reason, since there are no paradoxes here. All this just begs for a more elegant solution. In the meanwhile, the GM will need to rule on this Italy/France adjudication which is likely to be disputed regardless of his decision.

So getting back to the article, I liked Rule 2 better, although it is flawed in its wording. We certainly would not want the orders to be reinterpreted as "bre hold." What if there was French fleet pic s bre (normally a void order!). What if there were also more English fleets: gas s Maori. Now under the normal rules, units ordered to move cannot be supported in place. Bre was ordered to move. But the treatment of Rule 2 could be interpreted so that bre could receive the pic support to hold in place, even though it was ordered to move. I don't like this too much.

A better wording of Rule 2 would be something like: "If a situation arises in which an army's convoy order results in a paradoxical adjudication, all moves by all units involved in the paradox fail." I like this much better.

Now the main reason I like this (Rule 2a?) best is that it only applies to paradoxes and does not infect other normal elements of the game mechanics. For this reason, I would love to see something like Rule 2a replace A's Rule XII.5. But I also like Rule 2a because the result is easier to conceptualize, as opposed to uncut support from dislodged fleets. I would even apply a rationalization that I can readily accept, which neither of the two arguments really approached. Here it is, in not so brief form:

Diplomacy is a game of simultaneous movements and actions. Most people accept this at the surface and in most cases that is all that is necessary. But when paradoxes occur, a lack of full appreciation for simultaneous actions causes conceptual misunderstandings. In other words, even though everyone knows Dip is a game of simultaneous movement, I think many of them mistakenly interpret that to mean 'instantaneous.' That is, even though a campaign takes a whole 'season' to unravel, people mentally think of the events happening in no time at all... requiring special rules to allow them to draw a conclusion they are not able to conceptually understand. So let's think about what simultaneous can really mean. How about the concurrent transfer of influence by all elements *over* a period of time. Kind of like when you push down on a waterbed in one spot, you will see it rise immediately everywhere else. There's no delay. It happens all at once. *But* it only rises at the speed in which you push. It does not jump up to the height it expects it should be when you are done pushing. It happens over time in a direct reaction to your action.

So let's apply that thinking to the example. Let's say that we picture the bre fleet just starting to make it across the channel. Like starting to push down on the waterbed. The effect of this push is that the lon support must degrade a little because it is in the process now of being dislodged. So now the strength of the wal attack on eng is getting degraded as well. This in turn means that the eng fleet, which had begun to acquire a sense that it was becoming safely beleaguered, now finds it is coming under pressure from the emerging imbalance in the coincident E/G attacks. That is, G is starting to win and the eng fleet is in the process of being dislodged. Now we all know that a dislodged fleet cannot convoy armies, so the bre army which previously had a sense of just barely beginning to succeed in its mission, is now hampered and begins to draw back. Even though they have full support from yor, their own convoy is beginning to fail to the point that the attack on lon is insufficient due to the German distraction. Of course, if the bre army must drawdown because of the G pressure on their convoying eng fleet, the lon support for the wal-eng attack gets a boost. Now all of a sudden, the tide turns in the sea battle. The G offensive which was starting to look decisive is now being effectively stood off by the English attack. Since a beleaguered fleet can still convoy, the bre army can perhaps renew their assault, with the cycle repeating. And repeating. And repeating.

Fast forward now to the end of the season when there has been all this endless probing and false starts. Where should everyone be? Answer: right where they all started. Nobody moves. From the very beginning, a system of equilibrium was at work and that equilibrium placed the convoyed army at its origin in bre. The army in bre can never get to lon because the moment it begins to make progress in that direction, it is met with an equal force that resists it exactly. Ditto for the G fleet in bel. It had fits and starts where it began to get the upper hand against the E attack, but these advantages always disappeared as fast as they appeared. This sounds like what everyone knows to be a normal bounce (e.g. sil-gal, gal hold = bounce). For the bel/wal bounce, this is easy to understand... equal forces at the same point preventing movement. Like pushing down on a tabletop instead of a waterbed. Won't go anywhere. The normal conceptual hangup is with the failure of the attack on lon. People get hung up on this because the force resisting the movement in a paradox does not reside in the destination province of the various involved convoys. The force is applied to the convoys themselves, an element in the movement that does not exist in normal movement. Maybe it could be thought of as the paradoxical bounce. Instead of a unit in the destination province *pushing* an attack back, there is a unit acting on the convoy to *pull* the attack back. Or getting back to the waterbed, even though you know you have enough strength to push down all the way to the platform, you never get to apply it because someone else is standing there holding your hand back. The someone in this case was the German fleet from bel. Not a normal bounce. A paradoxical bounce. You're not running into a brick wall...there are no obstacles in your path. But you are restrained by a tether. That German fleet. I see perfect logic as to how the German fleet both protects the French fleet involved in the convoy while at the same time preventing the convoy itself from occurring.

This is an awful lot of words to get across a relatively simple thought. But since the simple thought runs smack into ingrained perceptions on movement, it is normally not even considered. In fact, most people, even with all these words, will reject this rationalization simply because they have a different concept of simultaneous movement that does not allow for the type of thinking I apply.

Interestingly enough, I developed some of my own conceptual thinking of a Diplomacy move from an article Manus wrote (I forget where it is). In it he explained the campaign concept within a move, the idea that a move does not occur at a point in time, but takes six months to unfold with the ultimate result being a determination of who exercises *influence* within the province. This idea of the passage of time is what I am referring to, not the other aspects of the article (which suggested conventional clashes of armies and navies is an overly simplified or narrow interpretation of actual events that could otherwise involve guerilla fighting, assassinations, propaganda, that kind of thing). Imagine my surprise when I saw that Manus was the advocate for Rule 1. I had expected him to hold the other view. But let's see, Manus has just finished programming the DPjudge. I do suppose that Rule 1 would be easier to code up than Rule 2... :-)

Steve Araps

From James North (jamesdnorth@yahoo.com):

I am writing this in response to your article. I must tell you up front that I disagree with both outcomes. However, it may be that I am working from a different rule book. The Diplomacy I use was published in 1961 by Games Research. In your scenario you discounted the German fleet and some subtle rules that are not obvious on first reading.

The rules. There are two parts to a convoyed army. The fact of convoying and the fact of landing in the target province. If either of these are not successful, the entire move is voided. Lets look at the landing. It says in my rule book that you cannot attack a province that is not adjacent unless convoyed by an joining and adjacent fleets. When the army in YOR supports the army BRE move to London, the army in YOR is actually attacking LON not supporting an army convoyed by a fleet in the Channel.

Simon's response: This is not quite correct. The army in Yor is supporting the army convoyed from Bre to Lon. To give support, a unit does not have to be adjacent to the unit it is supporting, it has to be adjacent to the location the attacking unit is moving to. In this case, Yor is adjacent to Lon, which is where Bre is moving to, so it's perfectly fine to say that Yor is supporting the army into Lon.
I say this for two reason. First, the Army in YOR is not adjacent to the fleet or the convoyed army in the Channel. Second, because an army cannot support a fleet in the water because it could not otherwise move there. Therefore the landing would be successful because you would have a two unit attack on a one unit hold (I will come back to this).

The convoy. Because the army in YOR is supporting the Army in in BRE move to LON the fleet in LON can no longer support the fleet in Wales. It must theoretically prepare to defend. Therefore, the Fleet in WAL and the fleet in the Channel are a one unit stalemate, but this nullifies the convoy so the army stays in BRE. However, because the Army in YOR is supporting a unit to LON (even if that move is unsuccessful) the Unit in LON still has its support cut to the fleet that wants to move from Wal to the Channel.

Simon's response: It is incorrect to say that Yor is actually attacking Lon. The order "A Yor S A Bre - Lon" alone does not cut London's support for Wal. If you say that Yor is attacking Lon, you are saying that the support is being cut. A support order cannot cut support, only an attack can. I don't think there's anything in the rules (even the 1961 rules) that says that a support order can cut another unit's support.

Now if I've misunderstood you and you are saying not that the support order is cutting support but that the unsuccessful attack is cutting support, that's a different matter. However, therein lies the paradox. If that's your position, you are correct as far as you went. The army hits Lon, cuts support, England attacks Eng with 1, Germany attacks with 2, Germany wins, dislodging the fleet. But if the fleet is dislodged, then the convoy fails, so there's no way for the army to get to Lon to cut support, so Lon's support is successful, so the German fleet does not win the attack on Eng. There's the paradox, which is what the two new rules are trying to address.

My version of the rules imply that the Unit in LON must stop its support and prepare to defend itself and that is why its support is cut.

Germany: Because French and English troops are at a one unit stalemate, this leaves the two unit attack from Germany as the victor in this move.

I am sure that you think that my theory is all washed up. And I am sure that no one would want me to GM. However, my belief is that a troop can only do one thing and that its first priority is to defend it territory. Unless it is attack from the province that it is attacking it must stop and defend itself. The only way to prevent this from happening is to cut the attack from a different direction before it happens.

Let me know what your thoughts are.

From Tarzan (tarzan_monkeyman@bigfoot.com):

In regard to the Paradox article:

I had a different take. I look at it this way. Let's simply ignore the German units. What would happen? The English Fleet in Lon would dislodge the fleet in Eng, and disrupt the French convoy.

At this point let's "throw" the German units back into the puzzle. As was pointed out in the article (argument #1), the German units are neither clearly pro-England nor pro-France. Thus, IMO the effect of the German units is merely to bounce England's Lon-Eng move. According to the current Dip rules this would in effect "undislodge" the French convoy. So, in effect, Germany's orders "save" the French Fleet Eng from being dislodged. The only real question (in my mind) is do Germany's orders also "save" the convoy? My opinion is no (see below).

When faced with dilemmas like this I like to return to the fundamentals of the actual game of Diplomacy. That is, Diplomacy is not about units fighting one another (although many players perceive the game this way), but instead Diplomacy is getting your "neighbour" to "permit" you to place a unit in a strategically favourable (or unfavourable) location. Remember Diplomacy takes place PRIOR to any war breaking out (as a matter of fact, according to the rules, when a player reaches 18 SCs this is when war breaks out - presumable the 18 SC power at this point feels that he has the resources to start a war?).

Returning to our paradox situation, France is trying (to England's chagrin) to strategically place an army in Lon (via a convoy). Meanwhile, England is trying to strategically place his fleet in Eng (and "kindly" have the French fleet back off). Meanwhile, Germany is strategically trying to locate his own fleet in Eng. Continuing the thought process along these lines, it seems to me that everyone should fail. Thus, I support rule #2, that is, all units hold.

Lastly, I want to throw one other idea out there. Let's say we took the same paradox situation and added one extra French unit in Mid and ordered F Mid S Eng. Well, the results would be unambiguous. Namely, France's convoy would succeed! That is, both England and Germany would bounce in Eng (France provided his own support for Eng) and the convoy would succeed!! Does this make sense in the context I suggested above? Is it fair? I think the answer is "yes" to both questions. So, if France adds this extra fleet, then clearly the resulting adjudication would turn out the same as if rule #1 was in place. However, without this "extra" fleet, I simply don't feel comfortable saying that the same outcome should occur. Thus, I support the "all hold" results from rule #2.

Editor's comment: And just to show that as well as being wordsmith's, our readers are actually also paranoid Diplomats too...

From Andrew Schoonmaker (aschoonm@orion.ac.hmc.edu):

Just a quick question WRT your article on modifying the rules of Diplomacy to handle paradoxical situations to assuage my curiosity... Did it matter which way I voted the first time? I mean, you're obviously recording the data, so in that sense, yes, but was the order I was shown the arguments for the two proposals affected by my vote, was
it independent, or was it pre-set (which seems silly given you went to all the trouble of having buttons and all...)?

I'm not done reading it yet, but so far it looks like another great issue!

Andrew

Editor's comment: The same thing did actually occur to me as I read the article Andrew... Simon? What's the guts?
From Dave (rebelservice@rebelsport.com.au):

I have read the article about the Paradox with the convoy to London and I was just wondering what a judge would do in this situation? Have you tried it just to see what the computer thinks?

Simon's response: I don't think it's been tried out.

The judge adjudication rules for convoys are available at: http://devel.diplom.org/Zine/W1995A/Tactics/section3.htm#TheJudgeRules

Anyone is welcome to try to work out the results by hand. From what is said, the judge's implementation manages to avoid some paradoxes, but maybe not more complicated ones.

Or, somebody can actually set up the situation in an actual judge game and try it out for real. The game 'oracle' has been set up on the USTR judge (ustr@diplom.org) for exactly the purpose of finding out what the judge will adjudicate in give situations. Just list the game to get instructions on how to set things up. I'd be very interested to hear the results if somebody wanted to set up 'oracle' to test the judge on this situation. Please do and let us know what happens.

- Simon


Editor's comment: *shaking my head*... and then, just to prove that this discussion really does have no bounds, Nicholas Robins and Simon branch of into a fantasy world of parallel universes... imagine the possibilities for variants there!!!! And no, I didn't just use the "V" word... it's all a figment of your imagination... so I don't want to hear about it... okay?

From Simon Szykman (ss8c@andrew.cmu.edu) replying to Nicholas Robins (nrobbin1@swarthmore.edu):

Nicholas,

Nicholas Robins wrote: I read the article, and your postscript. My interests here is in your postscript. Have you considered the possibility of a iterative universe. For a simpler example, lets say that the time traveller uses his sword to destroy his own sword before he left (he then has no sword, and so cant destroy it. yada, yada, yada..) Viola paradox. Unless we look at it in the following way:

Man A goes back in time destroys Man B's sword. Man B lives for a while, goes back in time, and does not destroy Man A's sword. Who then goes back in time and destroys Man B's sword. All is well. We merely must separate these two people. By any measure they are not the "same" person. They have different memories. They consist, at least partially of different matter, they experience two different subjective realities.

I've read stories that are based on a premise like this. The premise of a multiple universe branching into multiple universes with different local realities can make the paradox go away, but my understanding is that the possibility of concept has not yet been reconciled with the laws of physics as we understand them. Basically, they violate conservation of mass and conservation of energy. Where would the matter and energy for two universes come from, when previously there was only one.

Another similar solution is the possibility of multiple universes that are not different matter as you said, but a superposition of quantum states. Superposition would allow the two universes to exist as two different quantum states which exist simultaneously. The problem with this solution is that superposition of quantum states only works for unobserved systems. As soon as an observation of the system is made, the system collapses into one state or the other, and the convenient superposition that allowed both states to exist goes away.

(Please note that if it sounds like I don't know what I'm talking about, it's only because I've read enough about physics to be dangerous -- I can sort of rephrase something I've read but can't repeat the detailed argument I'd originally read.)

- Simon

Editor's comment: And that is all she wrote.