A General Theory of Late Middlegame and Endgame Strategy

By Alexander Woo


I have observed that many Diplomacy players, as they near the end of a game, really do not understand how Diplomacy games end, and the implications of that for their strategy. Therefore, I am writing this article which provides a logical framework for planning and deciding strategy towards the end of the game. I am hoping to provide one or more follow-up articles with some examples of how these strategies apply themselves to specific situations.

A Diplomacy game ends when one power reaches a majority of the SCs, or when the remaining powers agree to a draw. Practically, however, with the exception of a few balanced 3-way draws, games are drawn when one power runs up against a stalemate line held by some opposing coalition. (I will go against the usual definition and define a stalemate line as holding 17 or MORE SCs, both for this and any follow up articles.) If this opposing coalition is large, there may follow attempts to eliminate one or more powers from it. (If you do not know what a stalemate line is, you should learn before proceeding further. A good source is the Diplomacy-Archive.)

If you aim to win the game, you must of course aim to be the strongest power on the board. However, if for any of a number of reasons a win is unlikely for you, you must aim for a draw, and, in that case, there is no potential gain for you in being the 2nd strongest power. Rather, you need only aim to be at least the 3rd strongest power, or even some smaller but essential power, at the moment when a coalition must be formed to stop the strongest power from winning. After all, every place in a draw is equal, whether you have 17 SCs at the time or only 1.

Note that the two aims, winning and drawing, require very different strategies. In most cases, it will be clear which aim to play for. In other cases, it will not be clear, but delaying a decision could potentially be disastrous, and you must make the best decision possible. Sometimes, it is possible to postpone a decision for a year or two, but inevitably, there comes a point where you must make moves that will close off one possibility or the other.

I will first focus on the strategy for a draw, for several reasons. First of all, a win depends on the failure of the opponents to force a draw, and seeing the strategy for a draw helps in learning how to play for a win. Secondly, you will be forced to play for a draw far more often than you can play for a win. Finally, the strategy for a draw is less intuitive and far more often misplayed.

To repeat, a draw generally occurs when one power attempting to win (I will call this the "potential winner") runs up against a stalemate line held by an opposing coalition. At this point, the strongest power is unable to force a win, and all of the other players keep their place by an implied threat to throw the game to the strongest power if any attempt is made to eliminate them. Thus the situation becomes static and a draw is agreed to or imposed by the GM. Therefore, in playing for a draw, you have two aims both of which must be simultaneously achieved. You must encourage the growth of some power to be a threat to win the game, and you must become and remain an essential part of the coalition that stops that power from winning. If you are not an essential part of the coalition, than your "allies" can and most likely will safely eliminate you to make for a smaller draw. If you are one of the stronger powers, then you may also aim to safely eliminate as many powers as possible to reduce the size of the draw.

To be an essential part of a win-stopping coalition, you generally need to control a part of some minimal stalemate line keeping the potential winner from winning. SCs or even mere units on the far side of the smallest stalemate line you need to hold will NOT, at the end, help you gain a place in the draw, since the other powers can simply allow or even support what was formerly the potential winner in taking those SCs and forcing those units to retreat. It may seem counterintuitive, but this does frequently mean giving up SCs to the strongest power on the board.

Furthermore, note that almost any force expended against the potential winner before you need to actually move to form the win-stopping line is likely to be counterproductive. If you are one of the weaker powers in the potential win-stopping coalition, you want the potential winner to become a direct threat to win while you still are an essential part of the coalition. This means you want the potential winner to grow as fast as possible, since your place in the draw will be secured by your implied threat to throw him the game, and the bigger this power is, the stronger your threat. If you are one of the stronger powers in the win-stopping coalition, you will generally want to expend your force in reducing the weaker powers in your coalition to cut down on the size of the draw. For these powers, it will sometimes make sense to place some units against the strong power in order to gain more time to cut down the size of the draw.

There are also situations where a power is not essential. To have any chance, this power must delay the potential winner while he becomes essential. Also, a weak power frequently wishes to keep some non-essential SCs to support units fighting to keep or take more essential SCs. Fortunately, this problem frequently has a diplomatic solution. If you are in this situation, the potential winner usually has an interest in helping you attack powers that threaten to hold a stalemate line against him, and may agree to leave your SCs alone while you do some fighting for him, since it will be easier for him to take your non-essential SCs later.

Note that the powers in the potential win-stopping coalition have diverging interests until they actually need to work together to form the win-stopping line. The stronger powers in the potential coalition will be trying to eliminate the weaker ones. Furthermore, the weaker powers will generally want the potential winner to quickly become an actual threat to win, since this will give the others less time to eliminate them. Conversely, the stronger powers (except of course for the potential winner) will generally want to slow down the potential winner, though, given a choice between using units to against the weaker members of the potential coalition or against the potential winner, they will usually be better off with the former. As mentioned, this frequently leads to a temporary alliance between the potential winner and the weaker powers in the potential win-stopping coalition.

Of course, it can be very hard to determine when "you need to actually move to form the win-stopping line." Move too early and you could get eliminated, or, conversely, be forced to accept a larger draw. Move too late and the strong power gets a solo win. You need a strong tactical ability to be able to get this correct. Fortunately, in most cases, any of two or three consecutive seasons work. If you are strong enough, moving too early might not hurt you at all.

Now that I have finished writing generalities about playing for draws, I will switch to the problem of playing for wins. A win will generally require some miscalculation, either tactical or diplomatic, among several players who might stop you which ends in the failure to form a win-stopping stalemate line. Sometimes, you may be able to convince some player that you would forego a particular chance at a win. From my experience, this is highly unlikely. Sometimes, however, a power, for whatever reasons, decides to help you to a win. If so, you need to conceal this fact for as long as possible. The assumption among the other powers will be that this power will eventually cooperate with them in forming a draw, and lacking the information that this power will not, they may miscalculate in the mean time.

To achieve the win, therefore, you generally need to somehow convince some of your opponents to make a mistake. There are several prominent possibilities here. One is the long range forced or semi-forced win. If your opponents against you have no stalemate line, no matter what they do, you have some sequence of correct guesses that will lead to the win. Another way of putting this is that, if you could see everyone else's moves ahead of time, there is no possible way everyone else could coordinate to deny you the win. Some of these positions have only 10-12 SCs, and may require only 1 or 2 correct guesses. (Of course, this will also depend on your opponents having the wrong mix, for them, of armies and fleets.) Therefore, if your opponents do not have a very solid grasp of stalemate lines, you may be able to get one of these positions, since they do not appear to be particularly threatening. Alternatively, some opponent may decide to permit you a position where you need some number of guesses to win in order to eliminate the 4th or 5th placed power, or some opponent may need to permit you such a position to have a chance at a place in a final draw. In these cases, you may be able to guess correctly and win, though usually, you will have to make several correct guesses in a row. To take advantage of these tactical situations (or to avoid them on the other side), you need a good grasp of stalemate lines. You should know where the most common stalemate lines are, and also to compute as needed in a game all the minimal (17+ SC) stalemate lines covering a given position. (I feel confident that I can do this correctly in any situation given an hour, but I don't feel like I can explain how to anyone who doesn't basically understand the process already. Most 17 SC stalemate lines are listed somewhere, but there are no sources I know of for minimal 18+ SC lines.)

Another possible mistake is diplomatic. Some power's threat to throw you the game might not be taken seriously, either due to tactical miscalculation, or because some player wrongly thinks the threat will not be carried out, or even because some player would rather lose than submit to such a threat. To encourage such miscalculations, it usually helps to keep as many powers alive as possible. (On the other hand, a large number of tactical mistakes come from powers going for a win that they don't actually have a chance for, and you need to eliminate players to encourage such mistakes.) This situation is where a good knowledge of your opponents may come in handy; you want to keep around combinations of players who will let you win.

Diplomacy well-played is really a very drawish game, especially with longer deadlines which allow players to calculate the correct moves. Amongst expert players, wins by tactical force very rarely occur. Games that are thrown to one player after a diplomatic miscalculation by another are somewhat more common, but we see many 3-way draws and occasionally larger ones. All these draws occur despite experts taking calculated risks most of us would not imagine.

In the next article, hopefully in the next issue, I will give some examples of how to play for draws.

(Many thanks to Jamie Dreier for some helpful comments.)

aka Alexander Woo


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