I think everyone forgot JMO. I only played once,but he had true game. I would add him to my list

Zultar's a pussy and never plays. #Zultarsapussy.

He pretends to be busy. :()

Looking for only win percentage does not tell the whole story. Wins+draw ration give more accurate parameter.

In general sense player with higher win+draw percentage is better player as he has the highest success ration.

When you analyze the ratios of wins and draws you get insight how aggressive player is.

For example THM has 23+48=71 which means in 71% of the games he got favorable result. Ratio of wins and draw show balanced player.

Some other player with stats 25+30

show success rate of 55%, and probably more aggressive player.

My own stats: 23+57=80% show that success rate is rather high, and with 23% wins and 57% draw I would conclude that playing style is more defensive that THM.

martinck has 16+50=66% success which is very high but 16%wins and much larger percent of draws indicate cautious player.

My stats (win 23%, draw 40%)...show that I play a lot of gunboat.

My stats in the more limited (25) press games I've played (56% win, 40% draw) show that there's a good reason I'm a gunboat player.

@jmo
I didn't remember that. I was watching the game as it was being played though.

I rather like the Win/Draw ratio as an indicator of skill.
Won: 38%
Drawn: 36%
Combined: 74% favorable outcome. Ratio of Wins to Draws, nails me being a more aggressive player pretty accurately.

If you are going to assess people by these standards, i think you need to break press/gunboat apart

38% won and 36% drawn is a 38% favorable outcome, not 74%.

@CS, entirely different debate. That said, it's still one of the higher ones I've seen.

Win/ draw is not a very good metric- it's much more about game selection. I've won like 10 of my 12 live games or something which is a win rate most people in these lists could achieve. I used to literally just play live games whenever my Win rate dropped below 33%.

@Uclabb, never played live. Though I do agree with you. I tend to go with a combination of Ghost Rating, Win/Draw combined, Messages Sent Per Press Game Average, and Points as a pretty accurate determination of how good someone is.

All of those systems are pretty flawed. They massively benefit a player who just goes into games against bad players and sends mountains of press. It's usually better to just examine somebody's actual games.

@ghug, those metrics combined are actually pretty damn accurate for a good estimate of how talented someone is as a player. And examining someone's games is still not a good way of measuring their skills since you can't see the press or their neighbors press.

VI and pW have plenty of those types of wins. Ghug

Who has the highest combined win/draw rate for non-live press games?

Ghost Rankings are pretty hard to game actually. I think Zultar and I were really the only 2 people to take serious advantage of the 1 hole that let people artificially sore up them. Though it obviously isn't a perfect system when you add in win/draw plus messages sent it gets more accurate. Points is the least accurate of those stats, but still isn't a bad overall measure. I think if we queried the DB to pull (total points ever bet in games +existing points /total games player) then that would serve as a better indicator since it removes a lot of the data problems with just total points.

When I was an admin I also pulled some stats from the DB related to average length of messages in a game by players which was really interesting to take a look at since it correlated pretty highly to more length equaling a higher win percentage.

"Who has the highest combined win/draw rate for non-live press games?"

Probably some lucky new player who won their first full press game.

Ghostratings are very easy to game.

Yeah. Jmo's way is the easiest (and is also inflates win/draw at the same time and gives you credit for messages you didn't send), but the fact that you don't lose more from elimination against bad players means you can overshoot your "true rating" by playing worse players.

Someone's gonna point out that you lose in draws if players are worse enough. This is true, but it's intended behavior for a rating system, not a counterweight to that flaw, so it doesn't nullify it.

@ghug, no taking over a CD does not give you credit for messages you didn't send. The code checks the userID of the message when making that calc.

Really? I assumed because it processes it all at game end that it would just add all the messages from the game.

Anyway, that's largely unimportant to what we're actually discussing. The point is that there are other ways to game GR provided you have some skill. Taking over the best CDs is probably the only way if you don't though.

If i wish to size up the credibility of a player i'll look at the following:

1) Totally win percent (Drawn + Solo), given extra credit for how many of those were solos
2) Number of games completed. The more completed the more credible the win percent is as it gets harder the more of a reputation you have.
3) Review their completed games to see who they've won over. A string of solo wins over nobodies isn't nothing, but solo wins (or even solid draw wins) over known quality players is going to say even more about those players.

I'll look at tournament performance as well, but depending on the nature of the tournament there may be a bit of apples to oranges there when it comes to the skill set useful to succeed. Ditto for gunboat or variant heavy players.

I can usually find out who the best players are pretty quickly because they'll tell me. That's how I know Zmaj is the best player on webDiplomacy.

Does anyone know if there's a reason that the GR by categories doesn't give your details and record anymore? I quite liked that part of it.

"Who has the highest combined win/draw rate for non-live press games?"

Probably some lucky new player who won their first full press game."

Wasn't that Yellowjacket? Didn't he get off to a crazy start?

Didn't YJ only ever play like 10 games?

MadMarx is the one with the crazy start to end all crazy starts...