Jimbo,

the number of players involved in a variant plays a significant role in the amount of Vpoints moved at the endgame.
So, ceteribus paribus, gameID=12675 moves more points than gameID=13435 does.
Large variants weight more.

Then, the other non-negligible factor is the rating level of people involved.
Winning (or drawing) against a low-rated opposition gives you few Vpoints.
Losing against a low-rated opposition takes out many points.

Basically you have to combine those two factors.

So the more opponents the more I have to bet? Again this is not making sense to me.

If you think of a quadrant with four boxes:
- The upper right has the biggest maps and the highest quality opponents. There's the most points to be had there.
- The lower left has the smallest maps and the lowest quality opponents. There's the most points to be lost there.
- The upper left would be small maps but good opponents. Moderate point loss/gain
- The lower right would be large maps but lower opponents. Moderate point loss/gain

Does this clarify at all?

I think Mapu's quadrants are slightly off...

Basically, the more people you have in a map the more points you will win if you win, and the more points you will lose if you lose. This is because of the way draws are sorted out - each person in the draw scores against every other player, so the bigger the draw you are left out of, the more points you lose. Since bigger maps have bigger draws you will tend to lose more points when losing them.
This isn't too much of a problem - if you draw or win a large map you will gain much, much more than when you lose one (provided you don't have a large number of vPoints already) so provided you win/draw a couple it should outweigh the results of many games lost.

So is a 2 way in a five players map the same as 4 way in ten players msp? Or is latter better? Seems backwards

A 4 way draw in a ten players map would give you more points, as you gain points from 6 players rather than 3.

This is all from memory by the way - I may be completely incorrect. I think that's how it works though.

The large map/small map consideration only matters if the number of people involved in a draw are the same.

4 people drawing on a 10 person map is the same as 40 drawing on a 100 person map.

3 people drawing on a giant map of 30 people gain more than the 27 lose. The 3 are gaining from lots of people and the lots of people are only losing to three.

Unless there is some bizarre math involved that my pea brain cannot understand. If I am wrong, let me know.

“So is a 2 way in a five players map the same as 4 way in ten players msp?”

Considering users all equally rated, cause this is the only way to make cross-game decent comparisons, the latter is slightly better.

But is a 4 way draw on a ten player map a bteer result than a 2 way draw on a five player map? I would argue no, not at all, it is definitely a worse result.

Well, Jimbo, it's not up to me to give that kind of judgment… I just answered your initial question

No.
A ten player map always requires some more effort, so it awards more points.
From the wiki:
The game-value (gV) is set to: ((numberCountries -1 ) / numberCountries )^2.7
This gives small games a very low gV and the gV gets more, the more players are in the game.

OK, thanks for the explanation.

I suppose I should just conclude in that I think there are major flaws in this rating system. I do not mean offense by this.

I think that the current rating system penalizes players who enjoy small-games while it also rewards players who join the big map games. I have played only a few times on the big maps, however, it seems to me from this experience that it is VERY easy for a decent player to survive until an end-game. I do not enjoy these games as it is fairly boring as they move slowly, entering orders for sometimes 40 units is NOT my idea of fun and they tend to suffer from constant CDs due to the size of the pool.

Meanwhile, playing in smaller maps, 10 people or so, to me seems a lot harder to succeed. In these games they tend to play to their logical conclusion, all players decisions affect the outcomes for all other players (I would like to see someone argue that perspective on the big maps) and generally we find more competitive games.

You talk about "defeating more players" in the big maps. But really, as the mpas get bigger you interact with a smaller portion of your competitors. In the classic map, one has to interact with all 6 other players. Now, say Haven, I would say that you really only interact with about the same many. So really, you're no defeating more players you are defeating the same amount while 2 other quadrants have the same thing happening. I recognize the dynamics are different, but to say that for instance a 3 way draw on a 7 player maps is a worse result than a 6 way draw on a 14 player maps seems BACKWARDS to me.

I could understand that a three player draw on a 14 player map is better than a 3 on a 7 player map, but even a 2 way draw on a 7 player map is better than a 3 way draw on a 14 player map. Do players on this site truly believe that a 12 way draw on a 48 person map. Is better than a 2 way draw on an 8 player map? its easier to lose on the big maps, not harder to win. I think that these maps should pay out the same via the "buy-ins" changing.

Even looking at the HoF all the top players are "big map" players. Do those of us who play on small maps really deserve no chance to be compared on this website?

The problem is not the player-size, but in gameID=12675 a clear underdog has won the game. SO everybody lost a large ammount of points.

Why is losing to an underdog worse than losing to anyone else? Shouldn't the buy-in be fixed based on your skill level and not the skills of the others.

Should it not be the surviving players penalized for a solo? Defeated players whould be penalized when the game ends in a draw. (I mean additional penalties as Oli describes),

ANother example from "kaug's" profile:

13800 1897gunboat Classic1897 Won 38 1700
13673 There Can Only Be One Victor ClassicFog Survived -48 1652

A solo earns him less than a loss on the same sized map. How is thiss possible?

I'm at #34 with what might be (don't know how to check to be certain outside of manually checking all profiles) the highest win total on the site (172) and almost certainly one of the highest win% (65%). I've won 6.5x more games than the guy at #30 has *played* and won 43x as many. I understand that I have a high volume of 1v1 games, but this nonetheless strikes me as a little odd. How is the ranking system calculated?

The interesting thing about elo-like-systems is that you will reach your final score really quick, and than you just gain or loose a few points and your rating ever hardly changes.
PS: You are #34 in the new HoF and #43 in the old HoF.
PPS2: I moved the HoF back to it's original space in the help-page. The thing is working as expected, and I think we fixed most issues that where discussed in the last 5 month. I will add the gametype and pottype and variant filters, but after so many math and a real busy RL I'm concentrating on the interactive-map code now for some relaxing... :-)

The example with kaug:
He lost 38DPoints, because he was totally favored against Jimbozig (73% vs. 27%).
In game13800 he had to share his win with 6 (of 7) other players (only one defeated) and again was clearly the strongest player on the board (expected win % = 80%).

I guess that is the PPSC/WTA thing again. I usually don't check the pot type and play all games the same.

I end to think that kaug and I are pretty comparable gunboat players (no offense intended if he disagrees). We usually have good comeptition and each win some of the time. That is partly why I used him as a reference point.

I guess at the end of the day I will just argue that DPoints are the only rating system that matters ;-). Makes me seem better.