Tricky incentives in tournament poker

Big poker tournaments are a zero-sum game. The competitors pay to enter and those entry fees are returned as prize money. It is common practice for a player to be sponsored by someone else, i.e. a third party pays a player’s entry fee in return for a share of the profits earned. This is true in professional golf as well. Indeed, I have a good friend who even sponsored a bowler on the PBA tour, but the guy turned out to be a drunken skirt-chaser and he never cashed a single dollar the entire season.

The big difference in poker vs golf is that it is often the other poker players who pay the entry fees. As a consequence, I might be playing at a table with some people who make me money if they win. If I’m running low on chips, maybe I want to make sure I lose my last hands to the guy I own a piece of. Or maybe he and I can gang up on another player at the table or collude in other ways.

Steve Rosenbloom at espn.com has written a really interesting article about the current debate going on in the poker world over this issue:

(Thanks to John List for pointing this article out to me.)


scunning

Paul Phillips has some excellent discussion about this, too, at his blog. At the 2005 World Poker Tour CHampionship tournament, it came down to three players: Tuan Le, Paul Maxfield and Hasan Habib. It was rumored at the time that Tuan Le and Hasan Habib both had a piece of one another, and had miraculously ended up at the final table. In one controversial hand, Habib lays down KJs against an all-in raise from Tuan Le when Le was shortstacked (Le goes on eventually to win). Phillips' analysis of the hand provides some justification that collusion - albeit tacit collusion - may have existed, as Habib should want Le to at least make it to second, over third, thus providing the incentive for a dumping move.

JohnMcG

I guess it depends on what type of sport poker wants to be.

In NASCAR and cycling envents like the Tour de France, there are "competitors" whose main concern is helping someone else on their team win, and nobody considers that unethical.

Poker seems like it ought to be more individualized, but is that neccesarily the case? If it's impossible to govern out this collusion, why not make it explicit?

Dr. Funk

As a fairly competitive poker player, I'm not too concerned about this. First off, I'm not convinced there are that many strategies that make team play effective. While it seems that at different times it might make sense for teammates to avoid confrontation (as in the Tuan Le case) or to seek it in order to dump chips from a weaker player to a stronger one (like in Tilt), I'm not convinced either of these plays have a significant effect on the expectation of the other players.

My second consideration is that unless two players in a tournament actually own an equal amount of each other (it doesn't sound like this is the case), there will still be strong incentives for them to beat each other. Say Le and Habib each own 75% of themselves and 25% of each other and the difference between 2nd and 3rd place prize money is 800 grand. Cooperation is a risky and expensive option.

Bruce Hayden

Proviso - I don't play poker, but have clients who do, and also who back other players. And one of them claims to make a lot more money that way than on his own play - he claims to be better at picking players to back than playing himself.

My solution would be a combination of disclosure and prohibiting two players at the same table if either one had a piece of the other. Maybe even extend it to keep players in the same tournament from having a piece of other players, since that would eliminate the problem of what to do if two ended up at the same finals.

Yes, two could end up at the same table being backed by the same person, but that risk would seem to be significantly less than if they had a direct financial interest themselves in some other player at the table.

scunning

Paul Phillips has some excellent discussion about this, too, at his blog. At the 2005 World Poker Tour CHampionship tournament, it came down to three players: Tuan Le, Paul Maxfield and Hasan Habib. It was rumored at the time that Tuan Le and Hasan Habib both had a piece of one another, and had miraculously ended up at the final table. In one controversial hand, Habib lays down KJs against an all-in raise from Tuan Le when Le was shortstacked (Le goes on eventually to win). Phillips' analysis of the hand provides some justification that collusion - albeit tacit collusion - may have existed, as Habib should want Le to at least make it to second, over third, thus providing the incentive for a dumping move.

JohnMcG

I guess it depends on what type of sport poker wants to be.

In NASCAR and cycling envents like the Tour de France, there are "competitors" whose main concern is helping someone else on their team win, and nobody considers that unethical.

Poker seems like it ought to be more individualized, but is that neccesarily the case? If it's impossible to govern out this collusion, why not make it explicit?

Dr. Funk

As a fairly competitive poker player, I'm not too concerned about this. First off, I'm not convinced there are that many strategies that make team play effective. While it seems that at different times it might make sense for teammates to avoid confrontation (as in the Tuan Le case) or to seek it in order to dump chips from a weaker player to a stronger one (like in Tilt), I'm not convinced either of these plays have a significant effect on the expectation of the other players.

My second consideration is that unless two players in a tournament actually own an equal amount of each other (it doesn't sound like this is the case), there will still be strong incentives for them to beat each other. Say Le and Habib each own 75% of themselves and 25% of each other and the difference between 2nd and 3rd place prize money is 800 grand. Cooperation is a risky and expensive option.

Bruce Hayden

Proviso - I don't play poker, but have clients who do, and also who back other players. And one of them claims to make a lot more money that way than on his own play - he claims to be better at picking players to back than playing himself.

My solution would be a combination of disclosure and prohibiting two players at the same table if either one had a piece of the other. Maybe even extend it to keep players in the same tournament from having a piece of other players, since that would eliminate the problem of what to do if two ended up at the same finals.

Yes, two could end up at the same table being backed by the same person, but that risk would seem to be significantly less than if they had a direct financial interest themselves in some other player at the table.