edit: - the "or not"

Skrum has never been a good judge of skill no matter what his math says ;)

It is misleading to call it a pge, as much as just certain players that try to play a table game like a tourney tend to annoy people who like table games. Getting killed for a shit attitude isn't a pge, getting killed for having 0 logical foresight isn't a pge it is just being dumb and bad.

awww, I actually set that up so Vermont could comment ;)

but yeah, if you don't have the skills to play at the 2k/5k tables, you might as well donate your points before sitting a game

The regular play is a better measure of skill because there is (was) only a net ten points per player per game built-in bonus. So PPG of +10 is the average. Whether a negative PPG can exist in the presence of skill depends on what the player's total buy-in is, and Ryan does not provide that statistic *cough*. If a member like Mouloue uses his 2500 points to buy into big tables he has much more room for a negative PPG than a member who uses his 2500 points to buy into tournaments.

Tourney play has more elements or dimensions of skill to master and the payout is likely to reflect that number of dimensions. The number of dimensions is also dependent upon the number of players who compete against each other at a time and their degree of interactiveness. In pro golf, there is great degree of inequality in the tourney payoff but virtually no player interaction. In NASCAR racing all the drivers are in it together and a screwup by one can have major impact on the others. The payoff is much more nearly even. If Tiger Woods muffs a chip shot, that does not result in the death or injury of half a dozen other golfers.

I hope to publish a paper in the not too distant future on the relationship between the payoff structure and the dimensions of skill in a game.

Jurg: Check out who awarded my my contributor cross.

Oh hey skrum, good to see you.

Ryan badges are the coolest ones! well deserved one for you :)

Oh and now that I have your attention while talking stats, what are your thoughts for good tournament stats?

I think "points gained per tournament" (taking into account buyins so you could have a negative ppt) would be cool to add

also, I have a stat on my wish list that reflects where you end on average in a tournament. I'd call it something like average tournament performance %.

It would take into account your placing + the number of people who joined. The idea behind it that getting 2nd overall in a 10 player tournament is far less impressive than a 2nd place in a 30 player tournament.

So I would attribute a % for every tournament you enter. The stat would take the average of all your results.

Let's say there is a tournament with 10 players. 1st would get a 100% , 2nd would get 90%, last (10th) would get 10%. You could make the scale 100%-0% but I dunno, dividing 100% by the number of players feels better intuitively.

My self-criticism tells me that maybe a linear way of awarding a % isn't best to reflect skill. Someone who is very conservative could always end up in the top quarter of a tournament without ever getting 1sts. But his average could be better than someone very skilled who plays to win each tournament and will end up with a lot more top 2 finishes but also with some early exits.

so maybe there should be a bigger weight for the top spots? 100% for 1st, 80% for 2nd, 60% for 3rd and something linear from 50% to 0% for 4th to last?

Anyway, if you think this has potential for a good tournament stat, feedback would be appreciated

Jurg:

An approximate tourney stat can be generated by taking a player's total point score and subtracting games x PPG. What is left is tourney winnings (or tourney winnings plus 2500 for members) subject to some error because of hitting the bottom zero constraint, or now, scoring kills. The tourney winnings are according to the payout schedule, not games in the tourney x PPG.

The percentage payout for a particular place of finish decreases with the number of entrants, but, except for freerolls, the point payout is larger because the buyins are included in the pot. And in the case of freerolls, where you get something for nothing, would it not make sense that you get less something if there are more nothings?

The point payout is likely the best measure of skill regardless of your place of finish and two players' skill difference is the difference in their point payout. An article by Arnold Snyder,

http://www.blackjackforumonline.com/content/correcting_poker_tournament_chip_value.htm

shows that a chip does not change its value during the course of a tournament.