Shireroth

Statistic Explanations

Games played: for every game that you are in from start to finish, you get a games played point, even if the game was abandoned by the host. If you took over for someone on the very second turn, sorry, doesn't count. A little leeway is given for people that quit like the turn before they are eliminated, but only a little and on a case-by-case basis. India in Antiplomacy is a good example of this and will get a games played point regardless of their surrender.

Games participated: if you played for even a turn you get a participation point.

Total game points: The meat of the scoring. Points are awarded based on your finishing rank in all games you played.

Points are awarded as follows in standard variant:
1st: 20
2nd: 16
3rd: 12
4th: 9
5th: 7
6th: 5
7th: 4

And as follows in ten-player Youngstown:
1st: 20
2nd: 17
3rd: 14
4th: 12
5th: 10
6th: 8
7th: 7
8th: 6
9th: 5
10th: 4

Even losing outright gets you 4 points, under the idea that it's better to play games and lose than not play at all.

If a country was controlled by multiple people or abandoned due to ragequit / flat out disapeared and not taken over by any other players then the points will be divided among controlling players based on the number of years they controlled the country. So in a 7-year game, if Al controlled Germany for 2 years and Bob controlled for the remaining 5 and it ended up in 3rd place (worth 12 points), 3 points would go to Al and 9 of the points would go to Bob. In the same game, if Charlie controlled Italy and just stopped submitting orders in year 5 and Italty ended up in 5th (7 points), then Charlie would get 5 points and the remaining 2 points wouldn't be awarded to anyone.

Games that end with most of the players still in the game are only worth half points, with rankings being determined by total supply center control at the time of game abandonment. Ties for certain ranks are possible and likely.

Note that even if you get eliminated early on in the game, you still get full points for your country having participated in all of the years of the game. It's better to come in 5th and be eliminated by force than to come in 4th but quit right before you lose and only get a fraction of your points. So if you're going to quit, at least get your country conquered first!

Average Points Per game: calculated by diving the total number of points by the total number of games played. Games abandoned count half as much as games fully played. Games with leader swaps are unaffected (not playing a full game will always hurt your average)

Examples: David 18 points after completing two games. His average points per game is 9. (18 points/2 games fully played). Ethan 22 points after playing 2 full games and 1 game that was abandoned. His average points is 8.8. (22 points / (2 games fully played + 0.5 for a game not fully played)) = (22/3.5)

Average Ranked Position: the average finishing position your country came in throughout your career with the highest being 1st and the lowest being 7th (actually it's possible to go lower than 1 in big player games, but that's more of a math quirk). Abandoned games count half for this statistic but leader swapping counts full.

In non-seven player games (like Youngstown), your finishing rank is multiplied by the number of players and divided by 7. So finishing 5th in 10-player Youngstown is the equivalent of finishing 3.5th in a normal game (5/10*7).

As example calculations, if Freddie plays two games and gets a 2nd place and a 5th place, his ranked position is 3.5 ((2+5)/2) If George plays 3 games and gets a 1st place, a 4th place, and a 2nd place in a game that was abandoned, his ranked position is 2.4 ((1+1+4+4+2)/5).

Gold Star: awarded for coming in first place in a game.
Silver Star: awarded for coming in second place in a game.
Bronze Star: awarded for coming in third place in a game.
Administrative Chevron: awarded for administrating a game.

Stars are not awarded for abandoned games but admin chevrons are.

Player Rankings

Games Played / Games Participated / Total Game Points - Average Points per Game / Ranked Position (Any Awards)

The list is currently sorted by total game points in descending order, but may be sorted based on some sort of average in the future.

Official League Games

No official league games have been completed yet.

Unofficial Diplomacy Rankings

Rankings from the preceding games: Ari's Game, ICEBREAKER's Game, Harvey's Game, and Antiplomacy. League Games are also added in here.

Benjamin Kern: 2/2/32 - 16/1.85th (Gold Star) (Bronze Star)
Malliki: 3/2/31 - 15.5/1.96th (Silver Star x2)
Chrimingules: 2/2/30 - 20/1st (Gold Star) (Administrative Chevron)
Scott: 2/2/18 - 12/3.33th
Moshe Goltz: 1/1/14 - 14/3.1th (Bronze Star)
Evans: 1/1/12 - 12/2.8th
Ardashir: 1/2/11 - 7.33/4.53th
Andreas: 1/1/10 - 20/1st
Ruth: 1/1/10 - 10/3.5th
Harvey: 0/2/9 - 6/2.46th (Administrative Chevron)
Artz: 1/1/8 - 16/2nd
Corey: 1/1/8 - 16/2nd
Leo: 1/1/8 - 16/2nd
HisMajestyBOB: 1/1/8 - 8/4.2th
Allot: 1/2/7 - 4.66/4.33th
Erik: 0/1/6 - 6/4th
Aryeztur: 1/1/6 - 6/4.2th
CJ Miller: 1/2/6 - 3/4.1th
Nick Foghorn Leghorn: 0/1/5 - 5/6th
Braden: 1/1/4 - 8/5th
Max: 1/1/4 - 8/5th
Ryan: 1/1/4 - 8/5th
Elliot: 1/1/4 - 4/7th
Maksym Hadjimehmetov: 0/1/3 - 6/5th
ICEBREAKER: 1/1/2 - 2/7th (Administrative Chevron)
Ari: 0/0/0 - 0/unranked (Administrative Chevron)

If any person wishes to dissociate themselves with the league, they may say so and their ranking will permanently be erased from the board.

We were supposed to use the system I proposed, that is, the league points you get are your game points adjusted for the number of games you've played.

http://shireroth.org/forum/viewtopic.ph ... 31#p150931

I thought there was a general consensus on this. Furthermore, you are administrator of the first game, not the entire league. I handle the league table and would appreciate being included before you propose something like this.

Now, I like most of the things you propose, but there still is the problem of number of games played. Therefore, I would like my original proposal to remain, that is, your points in the league are multiplied by the fraction of games you've played. That removes the problem of people playing different numbers of games.

No offense but your system is... well, irredeemably bad. I'm going to do my own rankings. If people want to follow yours or follow mine they can make up their own minds.

Mike, in your system I get 0 points because I simply took over from someone else. Thus didn't actually play a game. So when you multiply.. 0. I like Harvey's. I get points in it.

I said I like most things in your system, Harvey. Do read everything I write please. The problem is that there is no solution to the problem of people playing a different number of games in your system. That's what I wanted to keep from my proposal.

Erik, I have no idea what you mean.

Then just add another aggregate stat.

Games Played / Games Participated / Total Game Points /Average points per game/ Any Awards

There both systems in one!

Harvey's system looks good, because it's already done and is here.

Erik, that was the entire point of my first post.

Well, I haven't figured out how to make them score in the overall player rankings but I do count games played and participated. So a player that is 1/1/20 (played one game, came in first) would be a lot better than a player that is 1/3/20 (played three games but only finished one and had a cumulative total of 20 points for all three games). In this example, player 1 is clearly a better player on average but player 2 puts himself out there more and plays more games even if he doesn't do all that well. Which is the "better" player - the one that one once but only once or the player that plays a lot and wins some, loses some? Hard to say.

I can't really think of a statically correct way to make those adjust the scorings in the same way I don't feel that just dividing points by the fraction of games you play is also valid. Maybe I should look up some RL sports stats and see how they deal with situations like this.

I can add an average stat if you want. I could do it as a straight average or as a "rank average" of what place you came in. So Chris would have a rank average of 1st because he came in first in both games he played and, say, Malliki would have a rank average of 2.5 since he came in second in one game and was tied for 3rd place in another.

Edit: guess it makes sense for abandoned games to have half the effect on your rank average as normal games.

Average would be good. Could you add points average (average points per game) and rank average? Would be awesome for the stats nerd in me.

In general, I think that players that play more games should be ranked higher. That is, if you play three games and win one, you should be ranked higher than if you play one game and win it. My proposal didn't do that, which I just realised. But you might be a better player if you win the only game you've played in, or it might be a fluke. Tricky...

I'll get to work on those. I'll probably also post the raw data at some point too since there's becoming a lot of it and I'll need it when new games are completed to recalculate averages.

Added averages. Abandoned games make the calculations annoying but everything seems to be right as far as the rankings go.

Want a database?

Eh, I should be able to handle it by hand.

Rankings updated with the results from Antiplomacy. Benkern is our new point leader and Malliki is close behind him and Chrimingles a hair behind them both. Both new champions have an overall ranking of less than 2nd, which is really amazing, especially given that Malliki has never actually won a game to date.

Brother Harvz, you can just call me 'Ruth' for the purposes of the table.

Harvey, here's for you:







I'm trying to remove the white around the stars, but for some reason Paint.NET is messing with me.

Use the Magic Wand, then baleet/

Why thank you. Next time I want completely useless advice, I'll ask you.

Eh, technically that's completely accurate in photoshop.

The useless part was that I of course already know about it.

Antagonizing comment!

Stars are absolutely excellent and I love them. Chevron could use a bit of style, but I guess lazy gits can't be choosers.

I didn't know what to use for the chevron, so I just took something.

At least you're using a Chevron and not a BP Sunflower.

"It's such a small oil spill compared to all the water in the Mexican Gulf"...

A Ragequit award would be a neat idea.

Yeah, that would be awesome and fitting.

I approve.