Select Modern2 Gunboat data

[Doug7878]

The following is based on data from small sample size of only 23 Modern2 Gunboat games, all of which had a pot size of at least 500 D, were not terminated prematurely, seemed to be reasonably well-played and did not have game-altering CDs. Because some judgement was involved with game selection, and the sample size is small, the results are of limited value, but here they are:

Average SC's at end of game
1. 12.391 Italy *
2. 8.391 Egypt
3. 7.348 Spain
4. 7.087 Poland
5. 6.217 Germany
6. 5.826 Russia
7. 5.000 Britain
8. 4.957 Turkey
9. 3.522 France
10. 3.261 Ukraine

Note: Wins in games sampled are (5) Italy *, (2) Poland and (1 ea.) France, Poland and Russia
* In this sample, the standard deviation of Italy's SC count results is roughly 30% higher than any other country

Significant Correlations in country results, based on SC counts at EOG
Please note that positive correlations are relatively rare in a zero-sum game

Positive correlations (above .100 in EOG SC count):
1. .500 Spain and Turkey
2. .362 France and Ukraine
3(tie) .254 Britain and Italy
3(tie) .254 Egypt and Poland
5. .170 Poland and Spain

Negative correlations (below -.200 in EOG SC count):
1. -.564 Egypt and Italy
2. -.389 Egypt and Turkey
3. -.386 Italy and Spain
4. -.379 Britain and Germany
5. -.378 Poland and Russia
6. -.339 France and Spain
7. -.324 Spain and Ukraine
8. -.320 Germany and Russia
9. -.302 Poland and Ukraine
10.-.281 Germany and Spain
11.-.261 France and Poland
12.-.238 Italy and Turkey
13.-.236 Britain and France
14.-.225 Poland and Turkey
15.-.219 Britain and Spain
16.-.211 France and Italy

Games in sample
gameID=125453, gameID=129025, gameID=135951, gameID=137626, gameID=141582, gameID=147911, gameID=149103, gameID=151900,
gameID=154760, gameID=157258, gameID=166546, gameID=167301,
gameID=169561, gameID=172461, gameID=178123, gameID=180408,
gameID=181231, gameID=184488, gameID=197714, gameID=218378,
gameID=234646, gameID=237777, gameID=246091

[dargorygel]

What was your expectation before you analyzed? Were you surprised by any of this?

[naked]

i think the results mainly show that a low sample size has huge variance. if you look at:

https://www.vdiplomacy.com/variants.php?variantID=19

italy seems more like a under average country, which is in line with my (limited) experience. italy is a middle power with reasonable possibilities to expand at the beginning, but it has no easy way to a corner.
i like the correlation numbers. first six are covered with my experience, then it gets shacky.

[Doug7878]

@naked - Thanks for the link and the comments. I agree with you that the limited sample is .
I didn't know that data was available elsewhere.

@dargorygel -
I have played in 3 games of Modern2 gunboat - as Britain, Italy and Poland (through 2006), plus have looked through these sample games and others for guidance on strategy.

I expected Egypt to be even more dominant than the data reflects.
In my limited experience, it seemed to me that the hardest four powers to play were France, Turkey, Ukraine and also Russia (because of the long border and two fleets that cannot coordinate unless either SWE or NOR is captured).

I expected some of the negative correlations, as @naked also pointed out.
I recently finished a game as Britain. Prior to the game, I searched finished games between good players to see how Britain fared under various strategies. The negative correlations between Britain vs. Germany (.-379), and also France and Spain were no surprise.

Regarding Egypt - I had not thought through, and found the strong correlation of results of Spain and Turkey in the sample to be very interesting from a Spanish or Turkish strategy perspective.

[Doug7878]

@naked - I agree that the limited sample is a problem. For instance, Italy's win stat and average EOG SCs, in particular, is simply not believable as an expected result.

Regarding the VDip data, I would argue, though, that statistics from all games creates the opposite problem. It would seem to me that variation in player skill levels will be the primary determinant of game results across all games. If this is true, it may cause a smooth of the results by country, if one makes the simple assumption that the more highly skilled players will tend to be evenly distributed across countries.

I guess my suspicion is that play-balance of the Modern2 map is somewhere between my small sample and the VDip data.

[Doug7878]

FYI

I have looked at an expanded game selection of 56 high quality games, including 10 low-bet games in a webDip round-robin tournament and 15 games on vDip. Italy comes back to reality, and Poland now comes out on top, slightly above Egypt. France and Ukraine are still in the dog house.

Turkey is an odd case. I don't see Turkey doing very well in recent games, but it has done reasonably well historically.

Average EOG SC's by country:
1. 9.110 Poland
2. 8.789 Egypt
3. 8.022 Spain
4. 7.416 Italy
5. 7.398 Germany
6. 6.757 Turkey
7. 4.974 Britain
8. 4.386 Russia
9. 3.762 France
10. 3.387 Ukraine

Significant Country Correlations in EOG SC's:
Positive:
0.263 France and Ukraine (they both lose frequently?)
0.239 Egypt and Germany (anti-Italy?)
0.212 Spain and Turkey (anti-Egypt?)
Negative:
-0.451 Egypt and Turkey
-0.386 Britain and Germany
-0.365 Poland and Russia
-0.334 Britain and Spain
-0.319 Egypt and Italy
-0.310 Poland and Ukraine
-0.301 France and Spain
-0.283 Germany and Russia
-0.261 Italy and Turkey

FYI and FWIW

[naked]

Some general staff. These numbers (you are giving) can give some inside, but not more. It makes a big difference if a good or a bad player is playing a country. Most nations have a clear strategic direction. If you dont recognize it you lower your odds, most of the time by a lot. As you dont know who is a good player (and there are few) your numbers get screwed a lot.

Poland is in my book the best country. It is a middle country (which is bad), but it has easy gains at the start, a good chance to live in peace with some neighbours (you cant fight them all) and a good shot for a corner.
Ukraine is in my book at least an average country. It is also a middle country, but has good chances to grow, there are clear incentives to make peace with Poland (for both sides) and the rest of the borders are manageable. Ukraine is weaker compared to Poland because it has a worse shot for a corner.

As a gernal rule in modern diplomacy 2 i would say the west countries are under average, the east countries are above average. I guess this is because countries in the east have a much easier way to get rolling and if you start rolling there is a really good chance for a solo.

[Doug7878]

Thanks @naked. You are right. Its a complex board, and skill plays a big part.
There are solo's by every country. You, yourself, had a beautiful solo as Ukraine.