By that logic, why not have the same penalty for losing to them as well?

I'm not saying it's wrong or inferior the way it is - I'm just surprised that the expectation value only matters in a loss.

As MM says, when you win, you are getting the points that the others contribute, regardless of whether you are good or not.

But the other thing is you are ignoring draws and are focused only on wins and losses.

Let's take MM's example and look at some draws. MM puts in 50 while the other six put in 10 each for a total of 110. In a 5 way draw, each player gets 22 which means MM lost 28 while the others gained 12. MM is penalized for a big draw against such lesser competition. In a 3 way draw, each player gets 37, with MM losing 13 and the others gaining 27. MM has to get to a 2 way draw to make 10% on his risk. That's what you get for playing players that are on average that much worse than you.

Now for a game without MM, the pot is 70 and a 5 way draw gets 14 for a net of 4 while a 3 way draw gets 23 for a net of 13.

So the expectation value does matter in a draw very much. The system certain rewards you for playing tougher competition which is good.

It is a self correcting system in that if you simply played lots of games against bad players, the inevitable elimination or even large draw will cancel out any gains you make. You've got to play players of close or better skill to advance.

I explain it the way do because
(a) It only simplifies like that in the WTA case
(b) To compare it to things such as Elo rating, you need to think of it that way.

@YellowJacket, abgemacht

The reason the expectation value only matters in a loss is because of the value of "V", which denotes volatility, essentially. I picked this as it is to ensure convergence as fast as possible, and that happened to cause the cancelling.

I rather like the win-loss ratio as a metric. I am afraid that by this measure I have an awful lot to learn. I wonder who would come out on top, given a threshold of say 50 games?

i just realised that i went down from 2565 to 338 in one month

Went up, you mean :)

Thank you, Alderian, TGM, that makes sense. I've actually never drawn, so I keep forgetting what a large percentage of upper-echelon games end that way.

Upper-echelon games are typically WTA, that's probably the bigger difference.

It will be interesting to see how far I fall next month. For the six ABI games not included so far, I have (2) three-way draws, (2) four-way draws and (2) defeats. I also potentially have one more three-way draw coming this month. I'm guessing I'll drop to the low 900's or upper 800's in January and that I'll never see four digits again. :*-(

Though you clearly are the best (GR and points) player on site and will likely remain that for the foreseeable future. So be happy about that. (until I take it from you :-) )

@MM, even if you do fall a bit you will almost certainly go past 1000 again even if only due to GR points inflation. I say "if" because I would have thought the one Victory you had will offset your defeats and draws with low points players (like me). What will be interesting to see is how long you can remain the only player to have passed a thousand points.

Looking back at Ghost Ratings from the past couple of years and the rate of points growth think it will take at least a year before any other player can reach you - maybe two years.

It is interesting how far you have pulled away from the pack though. There has been some points inflation, but your growth has far exceed that. One year ago the second ranked player (BigZombieDude) had 545 , while you had 815 . Now Babak is second with 612 . So even given inflation, the second position has not increased that much.

Looking at the number of games played by top players (as it takes time to accumulate points) I would say Dejan0707 looks like he could be a real contender. He is ranked third with 601 after just 31 games.

... actually after taking a closer look at the history. Maybe no one will ever catch MM The Master Gamer has come the closest to achieving as many points as MM, his highest ever was 781 , but he achieved that level back in February 2010.

The trick, if at a GR of 1050, is that your six opponents need to have an average GR of 350 if you are to keep the same GR for a three-way draw, any worse and you lose points for a three-way draw. So, you either need to solo or two-way draw to increase your GR, which means that almost every game you will lose GR. I'd need to solo a couple WTA right now to get back to 1050, and if I keep slipping, it will take three solos.

That said, I play to have fun and not to keep my GR at a certain level, so it's not a big deal, just an interesting statistic to watch each month.

But HWMNBM, is not winning fun?

And realitically, you contribute 60 GR points per game, so the lost one game could lose would be 60 . Assuming a solid win with opponents who average 175 (you probably play games with averages higher than that), you would recover from the worst case of a complete defeat.

MM you probably knew your GR might take a hit when you started the games. Which I thank you for. I enjoyed it greatly. None of the players in our game had high GRs (except for you) but I think we still gave you a reasonably challenging game. I think it is testament to how far ahead of everyone else you are that you can afford to play seven games simultaneously with players of varying GR, without worrying to much about your rating. There are players who are ranked well below you who probably wouldn't play someone like me because of my rank. But you let me play, thank you.

spyman, that is very nice of you to say. I didn't start the ABI based on what my GR might or might not do, I started it because I had an idea I wanted to try out and I thought others might find my idea interesting. I feel it was a great success and I had fun playing with so many different people, 42 different people in fact! It actually turned out to be a bit more than I could handle very well, *starting* seven games within a matter of a couple days, so next time I may limit it to four games, we'll see.

Anyway, I agree, all of the games were very competitive regardless of GR or any other imperfect way to rate people, which is one of the great things about diplomacy, the "little" guys can often take down the "less little" with just a bit of cooperation, it's always an exciting challenge no matter what your skill level (as long as everyone is engaged in the game, which I often find to be the case).

Draug, winning is fun, but in the end it doesn't matter whether you win or lose, I'm more interested in how the game is played. One of my favorite ABI games was when I was Austria and got eliminated, things were very exciting in the beginning and I was doing well for a while, then unknowingly I stepped on someone's toes and got my ass kicked, live and learn!

Draug, if I solo a game with 6 players that average 175 GR, then I only gain 6/17.5 * 175, which is a gain of 60, so I need to solo like that for every defeat just to stay even.

spyman, what the hell are you talking about?! Your GR is 225 and your win/draw percentage is 63%, in what way are you not a high caliber player?!?! The reason our game was a challenge is because it had very talented players. You don't have to be rated #1 or top 10 or even top 100 to be an excellent player.

That's what I said, HWMNBM... And I'm sure you rarely suffer a defeat/survive and that few of your games are against opponents only averaging 175.