Oh Danny Boy

Years ago when I lived in NYC, I played at a weekly poker game with a bunch of improvisors and comedians. At its height, our silly little $10 freezout and $25 cash game regularly boasted about 30 players in attendance.

Somewhere along the way, the organizer of the tournament got the idea to start tournament stat-tracking. I believe we used a points formula based on finishing position and number of players; I know for sure we tracked money in and money out.

What I observed at the Above Malibu game -- and what I have observed with every public stat-tracking of its kind since then -- is that most of the players were happy to show up and lose before the stat-tracking started. They didn't think much about it. If they made the money one week, they were very happy to do so. But after the stat-tracking was implemented, those players stopped showing up after being presented with the cold hard data of their (net losing) performance over time.

So it was with some interest that I read a recent blog post by Daniel Negreanu in which he performs some analysis on WPT results using the WPT stats database. Nothing about Negreanu's analysis strikes me as terribly wrong. I think he makes some excellent points. But I also think he mis-stepped by naming names of many of the WPT's net losers.

Rule Number One in the winning poker player's handbook is "Don't tap the glass." Negreanu's point about the WPT's need for a satellite system could have just as easily been made without naming actual names of net losers. By doing so he's put those people on direct, unavoidable notice that they are net losers on the WPT over 30 events. Will that encourage those people to play more events? If they stop playing, doesn't that make the fields tougher for the people who already are net winners?

Some might say "but tournament results databases have existed on the internet for a long time." Of course they have. But the fallacy of sites like The Hendon Mob is that you don't see a player's net results, only their cashes. You might see that Doyle Brunson has cashes of $135,203 in 2009, but without some work it's difficult to determine: (1) what his net on those cashes was (about $97,000); (2) how much money he spent on buy-ins for other tournaments where he didn't cash; and therefore (3) whether he is a net overall winner or loser in tournaments in 2009. Those omissions make results databases mostly harmless -- perhaps even helpful, because they tend to paint a rosier picture of a player's results than is actually the case.

On the other hand, comprehensive, net-result-based public stat-tracking in poker is a terrible, terrible idea. It only serves to dry up the pond. I have always been struck by a quote Nat Arem, a founder of PokerDB, gave to PokerNews in early 2009. Nat was asked, "If you could, what one thing that could be attributed to poker's "boom" would you prevent or change?" His reply:

I wish that all of the things that made the poker world less fishy would've never developed. That would include things like datamining stuff, like what we do at the PokerDB... It would also include CardRunners... [and] StoxPoker [and OPR and PXF}... [and] things like rakeback. ... The reason why is because it turns poker into this business that essentially exists entirely for the good players to extract money as quickly as possible from the bad players...

Tapping the glass doesn't just take the form of berating a bad player for sucking out with a bad play. Like tilt and many other elements of poker, it can often take more insidious forms. Scare the fish away and after long enough they'll stay away.