by Stephen Agar

The problems with trying to rate Diplomacy games for Tournaments are many and varied. Any system needs to be able to rate the performance of a player in a particular game with the other players in that game as well as being able to use the rating as a basis for comparisons with players in other games. If that was not bad enough, few FtF games of Diplomacy are played to the bitter end, so the rating system has to take account of the fact that games end prematurely (E.g. 1911), a fact which I itself gives rise to questions such as when should the game end and is it desirable to encourage players to act differently in the final year of a limited game-year contest, then they would have acted had the game being allowed to continue (i.e. do you sanction the final year rush for centres?) 

There are so many difficult questions associated with rating Diplomacy games that it is difficult to know where to begin. Should there be a reduced victory criteria? Should coming in second be a partial success or a total failure? Should a player’s performance be determined, if only in part, by what happens on other boards? Should all those who participate in a draw share equally? What importance, if any, is attached to supply centre counts during the game? 

There are no easy answers. Better minds than mine have looked at this problem from every angle imaginable and no system has managed to command universal, or even general, acceptance. However, I would like to examine some basic assumptions and see how far it gets me. 

I think any good rating system will be simple and will keep to the spirit of the original game. Diplomacy played to its natural conclusion only focuses on the end result, it is irrelevant how you got there or how long it took. A player who got to 17 centres and is then gradually reduced to 8 or 9 centres has failed, he is not rewarded for having done well early on, he is penalised for not having managed to capitalise on his initial success. Therefore, the first principle I would adopt in a Tournament rating System is that it is the end result that counts, if someone shoots his bolt too soon then hard luck. Any system which gauges the quality of the play on the cumulative record of the player throughout the game is simply not true to the spirit of the game, because it can lead to the situation where the winner of the game is the person who was 2nd or 3rd at the end of the game! This is similar to saying that when a football team led the game 1-0 for the first 85 minutes and then conceded two goals in the last five minutes of the game, then they are really the winner (because they were ahead for longest) even though they lost 2-1! Perhaps the Labour Party should have formed the last government because they were ahead in the Polls for the vast majority of the last General Election campaign… Yes, it is hard on someone who has done well, but who has been stuffed towards the end of the game, but if he/she was a good player then they wouldn’t have let it happen! For example, consider the following game… 

	01	02	03	04	05	06	07	08	09	10	11	Pt
A	3	4	3	4	4	3	3	2	1	0	0	27
E	4	4	6	6	5	7	7	8	9	7	4	67
F	5	6	6	6	7	6	8	8	8	11	15	86
G	5	5	4	2	2	1	0	0	0	0	0	19
I	4	4	5	5	4	5	5	5	5	5	5	52
R	5	6	5	4	4	4	1	1	1	1	0	32
T	4	5	5	7	8	8	10	10	10	10	10	87

France is a clear winner (15 centres), Turkey 2nd (10 centres), Italy 3rd (5 centres), England 4th (4 centres). Yet if you use a continuous supply centre count then the winner is Turkey, France 2nd, England 3rd, Italy 4th! Is that fair? More importantly, is it true to the spirit of the game? 

You shouldn’t feel sorry for someone who peaked too early and exposed himself to a combine onslaught, because Diplomacy is a race and those that fall behind in the final stretch should be counted as losers who put up a good fight, not winners! Of course this will encourage a “stuff the leader” mentality, but that is only as it should be. The good player is one who times his ascent to the top of the greasy poll so that he can remain there. I believe that ongoing track record should count for nothing, just as it counts for nothing in a game of Diplomacy played according to the full rules.. Therefore, personally I would reject the systems used at MidCon and MasterCon (and maybe even ManorCon – I can’t say because I don’t understand there system) because they are based on a players track record over the whole game. Any scoring system that produces a result which says that the player who was second or third when the game end is really the winner is not at all true to the spirit of the game. 

One criticism of this view is that it encourages centre grabbing at the end, which is not how the game would be played if it was continuing to the bitter end. To this I can only reply that the game is not being played to the bitter end, so you can’t judge it by the criteria that would apply if that were the case, you can only try to approximate the spirit of the game. Certainly a race to be ahead at (say) Autumn 1911 is more akin to a race to 18 centres, than continuous assessment. I see nothing wrong with centre grabbing at all, provided everyone knows that the idea is to have the most centres at the end of the game. If players want to leave themselves open to stabs on the last move, then that’s up to them. That sort of situation is infinitely preferable to the results of using continuous SC counts as a way of assessing progress which leads to the situation whereby a player who had a dominating position in the mid-game and yet who was ultimately defeated can still score more than the players who ultimately defeated him! After all, why compensate the player who may have been ganged up on at the end of the game, when you don’t compensate the player who was ganged up on at the beginning of the game and eliminated. FtF Diplomacy would be far more exciting with a nail biting finish with everything to play for – the player who leaves himself open to a devastating stab on the last move does not deserve to win. 

Having concluded that final supply centre count should be the only criteria, how is this turned into a manageable rating system? I favour simplicity.. Give an outright winner 100 points, then rank behind such a winner all the possible permutations in the order in which you think they should have preference and allocate points accordingly. Now this seems far too simple not to have been thought of before and rejected. What am I missing? 

Position	Final SC Count	Points
Outright Win	18 or more	100
1st	17	92
1st	16	91
1st	15	90
1st	14	89
1st	13	88
1st	12	87
1st	11	86
1st	10	85
1st	9	84
1st	8	83
1st	7	82
1st	6	81
2-way draw	17	80
2-way draw	16	79
2-way draw	15	78
2-way draw	14	77
2-way draw	13	76
2-way draw	12	75
2-way draw	11	74
2-way draw	10	73
2-way draw	9	72
2-way draw	8	71
2-way draw	7	70
2-way draw	6	69
3-way draw	11	68
3-way draw	10	67
3-way draw	9	66
3-way draw	8	65
3-way draw	7	64
3-way draw	6	63
4-way draw	8	62
4-way draw	7	61
4-way draw	6	60
5-way draw	6	59
5-way draw	5	58
6-way draw	5	57
7-way draw	4	56
2nd	16	55
2nd	15	54
2nd	14	53
2nd	13	52
2nd	12	51
2nd	11	50
2nd	10	49
2nd	9	48
2nd	8	47
2nd	7	46
2nd	6	45
2nd	5	44
2nd	4	43
3rd	11	42
3rd	10	41
3rd	9	40
3rd	8	39
3rd	7	38
3rd	6	37
3rd	5	36
3rd	4	35
3rd	3	34
3rd	2	33
3rd	1	32
4th	7	31
4th	6	30
4th	5	29
4th	4	28
4th	3	27
4th	2	26
4th	1	25
5th	6	24
5th	5	23
5th	4	22
5th	3	21
5th	2	20
5th	1	19
6th	4	18
6th	3	17
6th	2	16
6th	1	15
7th	4	14
7th	3	13
7th	2	12
7th	1	11
Eliminated 11	0	10
Eliminated 10	0	9
Eliminated 09	0	8
Eliminated 08	0	7
Eliminated 07	0	6
Eliminated 06	0	5
Eliminated 05	0	4
Eliminated 04	0	3
Eliminated 03	0	2
Eliminated 02	0	1

On this basis, using the example on the previous page, France would have 90 points, Turkey 49, Italy 36, England 28, Russia 10, Austria 9 and Germany 7. That looks just about right to me. What do you think? 

Reprinted from Spring Offensive 24