by Stephen Agar

Editor's Note: This article was originally published in Diplomacy World issue #80 in late 1996, and is itself a concatenation of an earlier article and discussion published in Spring Offensive issues #14 through to #16. The discussion led to the creation of a new Improved Diplomacy variant in Spring Offensive issue #19, and has been cited by Baron Powell as an inspiration in his design of the popular 1900 variant.

We are therefore happy to present it here, with the gracious permission of both the author Stephen Agar and that of Douglas Kent, who is the current Editor of Diplomacy World. We hope you will find this seminal article both interesting and instructive. Even players who aren't that interested in variants will find the discussion worth following, since it addresses many features of standard Diplomacy.

This discussion was itself a compilation of messages exchanged on the topic, and Stephen's responses. In the Responses section below, commenters' remarks appear in italic text (just like this Editor's Note), while Stephen's replies are presented in regular text.

The Proposal (Stephen Agar):

Just for fun I am proposing that we all join forces to redesign (or should that be improve?) good old regular Diplomacy. My idea is that we all sound off about what changes we would make to the original game to improve play balance — nothing that changes the game radically, no new units, new powers, or anything like that, but changes to the map, SC layout, starting positions, etc. Having put together a list of dislikes about the regular game, we will design a version that tackles these problems head on and hopefully produces a very playable game.

To set the ball rolling, let me list the areas in which I think Diplomacy is defective (in no particular order) with some possible solutions:

  1. Too difficult for Italy to open against France; therefore why not enlarge GoL to border on Rome and give Italy a F(Rome);

  2. Austria is too vulnerable if Russia takes Galicia, so divide Galicia into two — one space bordering on War, Ukr, Rum, and Bud, the other on War, Vie, and Bud;

  3. The eternal Ven/Tri problem would be toned down if Italy had a F(Rome) and if Switzerland was a neutral SC. Failing that, separate Ven and Tri somehow;

  4. Too many stalemate lines and bottlenecks; therefore divide Moscow into two, one as per the current Moscow, the other bordering on Mos, StP, and Sev. Allow A(Spa)-NAf and vice-versa;

  5. Turkey (and to a lesser extent England) is too hard to eliminate; so complete the whole North African coastline and possibly put a new SC in Egypt (giving Turkey another land flank) and make Ireland passable and adjacent to the Mid-Atlantic Ocean;

  6. Russia doesn't have enough options with a F(StP)sc, so extend Finland northwards to separate Norway and StP and let Russia start with A(StP);

  7. The game slows down far too much towards the endgame, so maybe we should consider permitting all Powers to build in neutrals — no more problems with quick expansion because you can't get your units to the front in time.

Okay, many of these ideas are not new and I certainly don't claim that they are — but let me know your views and we can put them all together and see what we come up with.

The Responses

Toby Harris:

I had mixed feelings about your article on "New Improved Diplomacy", as follows:

The game, in my opinion, would be spoilt by trying to "improve" it — it would be like trying to improve the Mona Lisa by making her teeth show. I agree that your points are valid ones and that some alterations would make a good game — but more as a variant than as an improvement.

There is already such a variant with copes with all bar one or two points you raised — the "Milan" game. As a renowned fan of playing Italy, I was instantly taken in by this variant, comprising a simple map alteration around northern Italy. I am sure you have the details, but if not you can see the article in Smodnoc No. 50, which goes into some detail on this very simple variant. It deals with the points you raise as follows:

GoL is enlarged to border Rome in the Milan game. However, whilst Italy does not start with a fleet in Rome, there is the possibility of forcing Savoy (similar to the old Piedmont province) and two options for the autumn — Bur and Mar. Very effective if played with a reasonable degree of skill.

Austria is too vulnerable — not any more! Italy and Austria don't have adjacent supply centres in Milan — hence, Austria only has one neighbour to worry about too seriously in the first move. Italy still has the Tyrolian attack on offer, but it is less tempting (with the new leverage on France) and, hence, Austria is happier to open to Gal and possibly arrange a stand-off with Russia.

The Ven/Tri problem? Again — not any more! Making Switzerland a supply centre would probably do Italy no favours anyway — it is two moves away from Venice — as it is from the nearest French and Austrian units, and Germany has the centre on his doorstep.

The idea of making Switzerland a supply centre is not an attempt to do favours for Italy, but to provide a focus for action which involves France, Germany, Italy, and Austria. By making Switzerland passable the established stalemate lines dissolve and east-west conflict is made easier. It removes the Piedmont bottleneck which inhibits French-Italian aggression. After all, as someone pointed out, Switzerland was passable to Hannibal and Napoleon. Anyway, using the Milan variant would allow Italy to go for Switzerland in S01.

The Milan game makes the stalemate line much harder to form.

I can't see that it makes that much difference on its own.

Turkey & England are harder to take out, although I feel this is compensated by their lack of flexibility towards the endgame, where wins are that much harder with the two countries. After all, these two countries are more for the safe, steady player than the gambler. I don't see that this is something which needs changing.

If you permit builds outside home SCs then England and Turkey have a great deal of flexibility in an end-game. Completing the North African coastline gives another land route to Syria, and broadens the scope of Italian-Turkish conflict, which usually hinges around who controls ION.

Yes, there certainly are very few options open to F(StP/sc). This is one point where the Milan game cannot come to the rescue. I am all in favour of your suggestion to expand Finland to the north coast and give Russia the army — it also makes life tougher for England, perhaps covering the last point to some degree.

As John Wilman points out, that change is not historically accurate. Alternatively you could give England an extra neutral to go for (Ireland) and let the Russian player choose either F(StP/sc) or F(StP/nc) in S01.

Perhaps builds should be permitted in any owned supply centre at all stages of the game — after all, it will be 1902 before this rule could be made use of anyway. However, with the two map alterations (the Milan map and the extended Finland), the game would probably be much more flexible anyway.

I just dislike the fact that building way behind the front line slows the game down during the endgame. I don't think that Milan goes far enough to be of interest to all the players on the board, although I completely accept that it improves the regular game with regard to Italy/Austria. I freely admit that some of the suggestions I made were based on Milan (though that particular wheel has been re-invented several times over the years in different forms). Similarly the idea of Egypt is an old chestnut. I just like the idea of putting a few of these simple map change variants together, along with any new suggestions that people can come up with.

The Milan game was originally designed to vary Italy's opening options, although it had very serious knock-on effects on all the other six countries; it strengthened Austria, Turkey, & England and weakened the rest. France and Austria are clearly the most affected, and I feel that the former needs to be weakened slightly whilst the latter could seriously use a little favouritism.

The fleet in Rome option, I feel, is far too boring — I mean, what is Italy going to do with it? Yeah, surprise surprise — F(Rom) - TYS - Tun; F(Nap) - ION - Gre. The only other possibility I can see is to take Tunis with one of the fleets and stick the other into WMS/GoL/ADS/AEG/EMS in the Autumn. Plenty of choice, I suppose, but it does give the victim time to make their Autumn builds accordingly.

John Wilman:

A few comments on your seven (nice number, seven!) proposals to improve on regular Diplomacy.
  1. [Separating Tri and Ven] This is basically the Milan variant, and seems to work quite well. Toby Harris likes it so much that he carries a Milan conversion kit around with him; we played it at Tittlecon.

  2. [Splitting Galicia] I believe that this features in some Fred Davis variants. Fred is one of the best and most respected variant designers, but European history and geography do not seem to be his strongest points — witness the "frozen seas" idea in Abstraction, when every schoolboy knows that the northern Russian port of Murmansk is open all year round; it is the shallow Baltic which freezes in winter! If you do split Galicia, Ruthenia should be the new province. It has been the focus of many historical disputes as to who should own it, unlike Transylvania, which is only argued over by Hungary and Rumania (as if anyone would want it anyway).

On balance I don't think there is any need to split Galicia if the Milan variant is used.

  1. [Fleet Rome] Italy needs two armies for flexibility, especially if an extra neutral appears in Switzerland, which is in any case counter-historical for the period in which the game is set. The straight F(Rome) variant makes life easier for Austria, but it has little real effect on Italy, who just builds an army instead of a fleet in Autumn 1901!

On reflection, I agree. As far as Switzerland is concerned, I am not sure if that is any more counter-historical than allowing Swedish neutrality to be violated.

  1. [Stalemate Lines] Moscow is vital for the cohesion of the Russian forces, and has little effect on stalemate lines. Allowing movement from Spain to North Africa is a good idea if you don't mind four-way nodes.

Here we disagree. I see Moscow as vital in most stalemate lines, as it is a big space which can only by attacked from two spaces to the north (Livonia and St.Petersburg) and two to the south (Ukraine and Sevastopol). It nearly always features in stalemate lines being held by a single support from Livonia or Sevastopol. Dividing it in two would make no difference at all to the Russians (I envisage a straight north-south divide with "Moscow" retaining all its present connections).

  1. [Egypt] I too would like to see the Mediterranean coastline complete. It needs another sea space — Libyan Gulf — and gives some point to Syria, other than being the justification for the Lepanto. Ireland would have no strategic value without a supply centre.

  2. [Russia] you can't do that to Finland without re-drawing the map of 19th century Europe and re-writing history. Besides, Russia needs a southern fleet, and from the south coast it can strike at Germany; it doesn't have to lurch straight into Scandinavia.

I've been giving the F(StP) and Irish problems some thought and what I've come up with is the suggestion that Russia can elect either F(StP/nc) or F(StP/sc) with the S01 moves, but that England is compensated for the increased difficulties in Scandinavia that this is likely to cause by making Ireland a neutral SC. Incidentally, according to Vol XIV of the Cambridge Modern History, the 1920 border of Finland is shown touching the Barents Sea and separating Russia from Norway, with Petsamo being marked as a port within Finnish borders.

  1. [Building in Neutrals] This is the "Premium Build" rule used in Mercator. like the Key rule, it destabilizes stalemate lines, which Calhamer was rather fond of, if only to show that war is very often fruitless.

Put the best of these changes together, and you have something very like Abstraction, with four supply centres for each country, and a much more fluid game.

At the moment I am only proposing 3 more supply centres (Swi, Ire, and Egy) and only six new spaces in total. That's 37 SCs as compared to Abstraction's 46!

Allan Gordon:

I am intrigued by your ideas on "Improved Diplomacy"! I can't think of anything you've missed — possibly a re-adjustment of the English provinces, and I've never been keen on the over-omnipotent sea-spaces like Black Sea, North Sea, Baltic, etc., where there are too many SCs adjacent to them.

For England see above. It is difficult to separate BLA without making Turkey and Russia an additional space apart (east-west split) or turning a move to BLA into a guessing game (north-south split), though the latter might work. North Sea would be a bugger to split too — I suppose you could divide it along a Denmark-Yorkshire axis, but it may leave England a touch vulnerable. The other possibility would be to divide it Yorkshire-Holland and ditch Heligoland Bight. I have no problems with BAL as it is.

James Nelson:

Whilst I agree with you that something to defuse tension between Austria and Italy is desirable, I do not believe that enlarging GoL to border on Rome and giving Italy F(Rome) is required. This idea has the distinct disadvantage that it makes France very weak. By doing this Italy can very easily make a supported attack on Marseilles in Autumn 1901 — how many Powers in Diplomacy can make a supported attack on another player's home center and still be able to pick up a neutral (Tunis in Italy's case)? This is an important consideration because essentially this is what you'll be allowing Italy to do. What is the point of improving Italy, arguably the worst Power in Diplomacy, at the expense of creating a "new" Italy in France?

Err… surely Italy can already do exactly what you suggest (supported attack on another's home centre and a neutral in A01) already — A(Tyr) S A(Ven)-Tri, F(ION)-Tun. Maybe I would be making France into another Austria? However, I do take your point.

I agree that the Venice/Trieste problem is one which needs solving, but allowing Italy to have F(Rome) is not really the answer. You may be interested to know that some research has been done on the success of Fleet Rome as a variant (albeit without any other changes). Geoff Challinger examined the results of completed games in Home of the Brave No.42 (1984) which I updated in Variants & Uncles No.13 (December 1988). I don't believe any games have been played since 1988, as the variant has been largely discredited. The Calhamer Points for the completed games (10) were as follows:

Placing Country Rating

Make of these figures what you wish. I think that it shows that France and Russia do very badly, Italy fares no better, and that Germany and Austria are strengthened due to the weakness of France and Italy. I suspect that if the GoL was extended to be adjacent to Rome then there would be little change other than to make France even weaker and England stronger. Italy will then be squeezed between two strong powers (Austria and England or Germany).

The Venice/Trieste problem has an ideal solution which does not wreak too much trouble with the play balance of other powers: namely, the addition of a non-supply centre province, Croatia, between Venice and Trieste (which is thus renamed Zara). Whether this space should be adjacent to all three Austrian centres or just Vienna and Zara is open to debate. One variant even has a neutral supply center between the two, but as you can imagine this only increases tension, totally against the purpose of the change! In Croatia Diplomacy, Austria has more breathing space but is still vulnerable to foreign units being in Galicia and Croatia, whilst France's strength is not diminished. Furthermore, Italy's defensive strength is increased slightly as it is harder for Austria to spring a surprise attack on him.

I have no problem with this suggestion, save to note that I think it helps Austria a lot more than it helps Italy, and that in regular Diplomacy in the UK Austria wins more games than Italy already.

I don't believe either that Galicia is so much of a problem. Yes it is a crucial space, much similar to Burgundy and the problem that France has if he allows A(Mun)-Bur in Spring 1901, or the Black Sea if Russia captures it in Spring 1901. One could also point to important provinces such as the North Sea, which borders on six centres and is within one move of another three. Galicia is a diplomatic problem, not a design problem.

I looked at the addition of Switzerland, both as a supply centre and as a non-supply centre, several years ago in conjunction with Milan Diplomacy. I came to the conclusion that it had very few merits. Whilst it does allow the bottleneck of Piedmont to be solved, it does lead to additional tension between France and Germany because now not only do they have to be concerned about Burgundy, but also Switzerland which will be adjacent to Munich or Marseilles! Furthermore, a stalemate line still exists. And of course the addition of a neutral centre would make matters worse. How could either France or Germany trust the other to take it when it is adjacent to a home centre? "Forcing" Powers into conflict is not a good way to try and improve Diplomacy, as any improvements should be made by increasing the options available, not removing them.

A solution to this problem I think I can now see. By dividing Switzerland into two non-supply centres, the western half being adjacent to Burgundy, Marseilles, and Piedmont, with the eastern half being adjacent to Burgundy, Piedmont, Tyrolia, and Munich, you allow the Piedmont bottleneck to be eased without greatly increasing the tension between France and Germany in the early (alliance-forming) stage of the game. It also allows France, Germany, and Italy increased flexibility as now the three can use the two provinces to support attacks against either of the other two — making it easier, for instance, for Germany to head south against Italy, or vice-versa and allowing extra leverage for France and Germany to attack each other and thus ease the Rhine bottleneck.

I accept what you say about making Switzerland a neutral SC, but I don't think making it a single passable space is such a problem. After all, there are many examples of spaces being adjacent to two other home supply centres of different Powers — ENG, Bur, Rum, Pru, Pie, Sil, Tyr, Boh, Gal, BLA. As you remarked of Galicia above (which borders 3 home centres and a neutral), it would be a diplomatic problem, not a design problem. It would also increase options for France, Germany, and especially Italy.

I think your suggestion of splitting Moscow into two is very sensible, and I can think of no objection to it. Again, allowing A(Spa)-NAf allows more options, but why stop at that? Why not allow F(NAf)-Spa/sc? I can see no logical reason as to why direct passage rules should not apply to fleets as well as armies. The logic behind direct passage is that the stretch of water concerned is so small that an army can move across the strait in small flotillas, so surely a fleet should be able to accomplish the same!

I also agree with your suggestion to eliminate the corner positions enjoyed by England and Turkey. Adding a southern coastline seems to open up Turkey and makes the Lepanto easier whilst a new centre on the Med coastline would strengthen Italy, making your other proposed changes to Italy redundant. My suggestion would be to move the centre to Libya and allow Italy to start with two fleets (or possibly even better, the choice of unit in Rome).

I don't think that giving Finland a north coast is desirable. Whilst I agree with you that Russia has no options for F(StP/sc), Russia needs a naval presence in the Baltic in Spring 1901. This is for the simple reason that otherwise Germany can completely dominate this region by building a second fleet in Winter 1901 and there is no way that Russia can then compete in the Baltic nor Scandinavia.

Building in neutral centres is a sensible suggestion and one with which I readily agree. It certainly would make Diplomacy a more fluid, aggressive game as the forces at the front line could be immediately added to without the delay of several turns which often occurs in Diplomacy.

Toby Harris thought that Milan Diplomacy was the solution to your problems, but I have to say that it really does not achieve its stated aim, which was to improve Italy's poor position! Again, I have some statistics from games played to completion (Calhamer points again):

Placing Country Rating

Austria is very strong in Milan Diplomacy and France weak, which leads to the immediate weakness of Russia and the strengthening of England. Italy is weak in the mid-game — he can make good early progress against France, but then is sandwiched between a strong England and Austria.

Finally, I come to your proposed change to Ireland, making it not only passable, but adjacent to MAO and also a supply centre. Your proposed changes are very bad for England if Russia builds F(StP/nc) or A(StP), because it means that England no longer has a guaranteed build (because France can stand England out of Ireland). My solution would be to start Russia with F(StP/sc), make Ireland passable and adjacent to MAO as you suggested, but not to have a supply centre there. And of course, if Ireland is to be passable it really deserves a direct passage connection to Scotland.

Alternatively, you could just have a Liverpool-Ireland direct passage, so England could secure and SC in Ireland in S01.

As John Wilman points out, most of the changes you are putting forward turn the game into something like Abstraction II. Abstraction is a much better game than Diplomacy. It has two features which make it more attractive than Diplomacy. Firstly, it uses Army/Fleet rules which allow a more flexible and aggressive game; and secondly, due to the increased number of provinces, there are very few stalemate lines, and they are harder to construct due to the power of Army/Fleets. As Richard Sharp said in The Game of Diplomacy, "Abstraction is probably the only variant ever to improve upon the basic game".

What I am suggesting is nothing like Abstraction, in which the board is totally redesigned. Also, you are quoting Richard out of context — what he actually wrote was "This is the expert's variant, considered by some good judges to be the one example of a variant which has improved on the original game", which sounds more like fence-sitting (and of course he was writing fifteen years ago).

The other alternative is Davis Diplomacy, which adds a High Ocean Box connecting the central and eastern Mediterranean to the western side of the board (thus eliminating the corner positions of England and Turkey), Croatia between Venice and Zara, a 35th supply centre in Morocco (North Africa being split in half into Morocco in the west and Algeria in the east), and Army/Fleet rules. Ireland is also passable, though it is not adjacent to MAO, and it does have direct passage to Clyde.

Steve Rennie:

I am fascinated by the debate over an improved Diplomacy. The difficulty of adding options, strengthening Austria without weakening Italy too much, and the importance of not tinkering too much with a basically excellent game, all appear to have been addressed by your option of three new supply centres. I should say that in considering options I think history matters not a jot; the game is all!

The Result: Diplomacy II

Putting elements of this discussion together, Stephen created a new "Diplomacy II" variant based on his own proposals and the comments he had received. This new variant is described below.

All the rules of Regular Diplomacy apply, save for the following:

  1. The following new spaces are created (supply centers listed in bold):


    Milan is an Italian home supply centre in place of Venice, which loses its supply centre status. Iceland and Egypt are new neutral supply centres. Tuscany is not a space in this variant; it is instead absorbed into the territory of the Rome space.

  2. Italy starts the game with A(Milan) instead of A(Venice).

  3. Russia has the choice of starting with either F(StP/sc) or F(StP/nc). The starting position of the northern Russian fleet is revealed along with the Spring 1901 orders.

  4. Powers may build in any vacant neutral supply centres that they own, provided they still control at least one of their home supply centres.

  5. There are not 36 supply centres; control of 19 is needed for victory.

  6. Units may pass directly from Ireland to Clyde, and vice-versa. Ireland also borders directly on MAO.
  7. There is a 4-way junction at Gibraltar between MAO, Spa, WMS, and NAf. Units in any of these spaces may move into any of the other three.

The map for the variant appears below:

Stephen Agar

If you wish to e-mail feedback on this article to the author, and clicking on the envelope above does not work for you, feel free to use the "Dear DP..." mail interface.