31.7% of games end in a draw while 68.3% of games end in a solo win. i posted a thread on the subject a while back.

then the chance to win is 9.7%

and it varies absed on country and player.

how many people have on average a higher than 9.7% win rate?
what is the standard deviation? (in fact it'd be nice to look at the distribution of player's win rates, and the distribution of win rates per country... though i imagine they are the same) still you could work out he normal, and see what it means to be abnormal (and thus super-normal if you're a good player, and sub-normal if you're a bad player)

Ahhh, but what's the ppsc divide, rather than wta. Because surely the chances of you getting a win in ppsc are much higher than wta.

i think the chances of a draw in ppsc are much higher... but then i'm ignoring the willingness of people to lose while holding many supply centres...

StevenC. and Tru Ninja: Thanks

Centurian: Maybe, but PPSC sure has its share of draws. I agree that the statistics must be different, though. And I bet no-press ('gunboat') games nearly never draw.

hellalt: It's actually closer to 9.8%. 68.3 divided by seven is 9.7571428 with '571428' repeating.

orathaic: Disregarding games joined after the pre-game phase, I've won a third of my games. There's one.

if i can i will take a look at the database....

does anyone have/know how to find the link which kestas posted recently to download a copy of the database?

Well my personal experience definitely suggests WTA are MUCH more likely to draw. Or to phrase it differently, people are more paranoid about someone winning in WTA.

Does being over 9.8 really suggest you are above average? I mean, there are plenty of cds and flop players that contribute to keeping that average pretty far down.

i believe this is what youre looking for orathaic:

http://webdiplomacy.net/webDiplomacy.net-SQLdata.zip

additionally, i think that in order to find the mean win pct, youll have to take an average of the win pct's of a variety of players when playing the 7 given countries. each country has their own pct's of soloing so soloing with russia is far easier than soloing with austria or italy.

the average win pct is likely closer to about 15-20% for the avg player without taking different countries into account. additionslly, to find this value, youll have to determine a baseline. do you take any player even those that play 1 or 2 games and then quit or do you make a minimum requirement of finished games before you take their stats into account?

Considering this is a game of skill and negotiation, all these percentages don't mean squat. This isn't dice rolls or a card game where there is a random chance. Skill should never be viewed on percentages.

Centurian and Tru Ninja: The 9.8% we calculated has nothing to do with individual players or individual countries. And I don't see mathematically how the average player can win more than 1 ⁄ 7 of the time. That doesn't make any sense.

Draugnar: Yes it should.

How in the world do you open a .sql file?

.SQL files are just text files. Use notepad.

No, it doesn't because a skilled player is less likely to play an unskilled player and more likely to play against someone who challenges them. Therefore the truly skilled players record may not be that much different from the lousy player who plays noobs all the time, thereby skewing your stats.

never-the-less no matter the skill of the player individuals will deviate from the averages.

if i am great at playing austria but shit on average then that owuld be shown in my performances. (that said i know i don't have enough games under my belt to get a statistically useful win/loss/draw percentage) Still interesting to look at some of these differences... Also Draug, it is not neccesary for skilled players to only play unskilled players.

Further, i've found it much easier to convince a good player that he was making a mistake by attacking me (or that wiping me out would let someone else win) anecdotally the sucess of my diplomatic skill seem to be proportional to my opponents strategic skill.

You can't measure skill in a game like Diplomacy using percentages because there are far too many deviations. Every single game of diplomacy is different and might require a different approach and there are just so many different possibilities that you can't boil it down to mathematics.

Uh, hello, this only works if everyone playing a game is at about the same level. As for a long-term measurement, it should balance out in the end. People will eventually start playing better opponents and the sample size will increase. You can say you start off playing easy folk in any system. As for openning the file in notepad, try it. Notepad tells me: "The . . . webDiplomacy.net-SQLdata\sqlStrippedDB.sql file is too large for Notepad. Use another editorto edit this file.". I tried wordpad and nearly froze my computer.

That was to Draugnar

i tried openoffice - which is capable of handling databases - and it slowly used up all my memory...

i need a way of querying the database without loading the whole thing into memory...

All I want to know is the average chance of one player winning a game before a country is even chosen. Of course people will deviate from the averages and of course there are far to many deviations.

@ JECE as i stated in the tail end of my post, you have to draw the line somewhere on who to tally and who not to tally. if you tally every player ever, then you can likely simply take the 60.8% and multiply by 1/7. if you wanted to tally every player who plays at least 10 games, then its likely higher as i listed above.

i also had the same problems with opening the database and i didnt know jack about SQL to do anything there.

I feel like I suck but I play the highest pot games I can find. Maybe if I played 5 point games my success would improve as would my win ratio!

i'll get an SQL/mySQL server and do some queries on the database shortly...

Tru Ninja: For my purposes, I'm counting games, not people. I just need to know the percentage of all games that end in a draw. Then I can divide that percentage by seven, as you said.

the % of games ending in a draw is approximately 31.7%. games ending in a solo are 68.3% take the 68.3% and multiply by (1/7) and youll have what youre looking for. the percentages are what i tallied over the course of 1001 games on this site.

Yeah, we did that and got 9.8%.