Diplomacy Wiki
Strategy Article


Was I Right?


Richard Sharp


The Numbers Game #17 (April 1992)


[Link to article]

The most obvious change in the Austrian openings over the years has been the swing away from the Balkan Gambit to the Southern Hedgehog in the late 1970s, and back again some eleven years later. I can certainly claim to have been largely responsible for the first of these phenomena. In articles in Dolchstoß and in The Game of Diplomacy I plugged the Hedgehog as being the best way for Austria to avoid an early bath; I didn’t claim it was particularly a winning opening, but there is no doubt that it is much more difficult to win if you have been eliminated in 1903...

The question is, was I right? To find out, I ran a programme comparing Austria’s results with the openings chosen, ignoring games in which Austria had resigned or dropped out, and ignoring abandoned games and those in progress. The facts of the case are as follows:

The Trieste Variation of the Balkan Gambit (F(Tri)-Alb, A(Bud)-Ser, A(Vie)-Tri), still the most popular Austrian opening overall and now making a comeback, is significantly less likely to produce a win (9.3%) than its main rivals the Southern Hedgehog (F(Tri)-Ven, A(Bud)-Ser, A(Vie)-Gal) and the Galicia Variation (F(Tri)-Alb, A(Bud)-Ser, A(Vie)-Gal), both 12.3%. The Galicia Variation is significantly more likely to produce a draw (25.1%) than either of the others (20.4% and 20.2%). Overall, then, the Galicia Variation avoids defeat in 37.4% of cases, as compared with 32.5% for the Southern Hedgehog and 29.7% for the Trieste Variation.

Mind you, these pale into insignificance compared with some of the less popular choices: of openings played 10 or more times, the best success rate (50%) is shared by the unbaptised F(Tri)-Alb, A(Bud)-Tri, A(Vie)-Tyr and, believe it or not, the Bohemian Variation of the Balkan Gambit (F(Tri)-Alb, A(Bud)-Ser, A(Vie)-Boh)! Shows what a load of rubbish statistics are, doesn’t it? For out and out winning prospects, F(Tri)-Alb, A(Bud)-Tri, A(Vie)-Tyr is the best at 20%, though this is based on a sample of only ten games (two wins, three draws, two 5ths and three 6ths). At the other end of the scale, three Austrian openings have been played 10-20 times without a single win: the Viennese Variation of the Galician Gambit (F(Tri)-Alb, A(Bud)-Tri, A(Vie)-Gal), the Italian Attack (F(Tri)-Ven, A(Bud)-Ser, A(Vie)-Tyr) and, surprisingly, the True Hedgehog (F(Tri)-Ven, A(Bud)-Rum, A(Vie)-Gal).

In terms of disaster avoidance (where “disaster” is arbitrarily defined as 5th or worse) the Trieste Variation scores 58.9%, the Galician 64.7% and the Southern Hedgehog 65.8%. The most successful of the more popular openings has been the Vienna variation of the gambit (F(Tri)-Alb, A(Bud)-Ser, A(Vie)Std.) with an impressive 76.3%; the least successful has been the Hungarian Roadhog (F(Tri)Std., A(Bud)-Ser, A(Vie)-Bud) with 44%. The best chance of coming 7th, if you want a quick game, is offered by the Budapest Variation of the Gambit (F(Tri)-Alb, A(Bud)-Ser, A(Vie)-Bud), closely followed by the Hungarian Hedgehog (F(Tri)-Ven, A(Bud)-Ser, A(Vie)-Bud), which I once described as “quite interesting” in an inattentive moment. It looks as though A(Vie)-Bud is a losing option. Among the minority openings, I suppose the worst is what I called the Dead Porcupine (F(Tri)-Ven, A(Vie) S A(Bud)-Gal), still only played once, and the only way of guaranteeing that you can’t get Serbia in 1901.

This is the frequency table for the three most popular openings:

Year Southern Hedgehog (%) Trieste Variation (%) Galician Variation (%)
1973 1.2 37.6 11.8
1974 1.1 31.2 15.1
1975 1.1 35.2 19.8
1976 5.0 27.3 20.7
1977 4.0 23.2 18.4
1978 16.7 19.0 15.1
1979 23.2 9.5 15.8
1980 26.1 19.3 17.0
1981 22.5 22.5 17.1
1982 27.3 21.2 25.3
1983 23.5 21.6 14.7
1984 24.6 19.5 17.8
1985 23.6 17.1 21.1
1986 26.1 18.9 14.4
1987 24.3 16.2 19.8
1988 28.0 28.0 18.3
1989 16.9 21.1 19.7
1990 21.7 43.5 13.0
1991 31.1 21.3 13.1