Poker and Finance II

[The DaveR]

The article draws a poor analogy to finance. The obvious problem is that, playing at a poker table, each hand is independent of the next. Sharpe Ratios, as intended for portfolios of correlated assets, is a poor metric in such a circumstance. Further, overall variance decreases dramatically as one plays more hands. What is your variance this hand, this session, this year, over your life? We know this, hence the often repeated maximum that over the long term, skill prevails at poker.

Even more important, particularly at a limit table, the variance of each decision is between 1-4 BBs (1-5 BBs in Vegas) whereas the EV decisions are fractions of the eventual pot. These decisions are straightforward and almost never require a second set of metrics that incorporate one's risk of ruin. For example, calling with a 4-flush on the flop doesn't need one to ponder my bankroll, variance, or whatever. Just call.

Finally, I would argue that variance is, for the most part, a function game texture, not one's decisions. Casual observation of LA vs. Vegas vs. East Coast mid limit games shows this. In that context, altering your decisions to try to alter variance is futile.

I suggest the author, and the editor of the magazine, read this post by the departed NPA (http://archiveserver.twoplustwo.com/showthreaded.php?Cat=0&Board=holdem&Number=1122424 &Searchpage=1&Main=1122424&Words=rareness+Ed+Mille r&topic=&Search=true#Post1122424) and this old RGP thread written by my friend. (http://groups.google.com/group/rec.gambling.poker/tree/browse_frm/thread/e42ec7b0d9c032db/8fc4b7ecdc1e5178?rnum=1&_done=%2Fgroup%2Frec.gambl ing.poker%2Fbrowse_frm%2Fthread%2Fe42ec7b0d9c032db %2F65ca2ebceb825d89%3Flnk%3Dst%26q%3D%26rnum%3D6%2 6#doc_1e551626997e1d4b)

[gergery]

Yeah, I like the idea of trying to apply some finance theory ideas to poker, but I think Portfolio Theory is the wrong area. There are a whole bunch of things that are very different.

First, portfolio theory and sharpe ratios allow you to select a number of investment choices at one time (ie. i'll buy IBM, and some bonds and some puts, etc.), whereas poker hands come sequentially. you can't say, 'i'm going to hedge my risk in playing 87s by playing 99 too'

The article also doesn't really consider the next best alternative for your money either. But in poker, you are locked in to your money sitting on the table while you play. So that money can't go be put to work in another investment vehicle. Since most of your decisions at a table are not for your whole stack, that means you essentially have a bunch of extra volatility capacity that is not being used, making the EV side of the equation more relevant.

Variance in poker can be a GOOD thing. For the same reason that good players at high stakes games don't like to "run it twice", increasing your variance can put your opponents on tilt and make them play bad. There is no equivalent "make other investment managers pick bad investments" in finance.

Finance is done exlusively for money. But poker is a game for money AND enjoyment, and often the enjoyment can be more important. For example, Limit Omaha 8/b has MUCH less variance than Limit Holdem, and comparable winrates. And online play takes away many of "yes, but..." reasons (ie. you can find game good game selection on line, you can play for stakes of your choice, etc.). Yet there has not been a mass exodus of people to O8. You hear poker players say, "I'd rather play Limit than NL today", but you don't hear finance pro's saying, "yeah, the returns are a little lower but its just more fun to invest in XX" -- the utility of playing one game in poker is much more relevant than the utility of investing in one vehicle.

Risk of Ruin is often irrelevant at the table, and even irrelevant over a selected number of hands. If I have $10k to play poker and sit in a $10-20 game, then what is my risk of ruin at that limit? It's exactly 0.00% Because if I ever lose more than 25% of my bankroll I'll move down in stakes. What you REALLY should be calculating related to this is the impact of losses on my future action (ie. the % chance i need to move down * the winrate differential between my old limit and my new one). Or you should calculate what I'll lose in income if I lose some at poker and have to withdraw money from my mutual fund as ppker working capital, and now don't get the mutual fund appreciation.

Personally, I think there may be more application of finance theory to poker in option valuing. for example, "what is the value of the Bellagio poker room manager calling you to tell you there is a huge fish playing NL right now at $10-20 blinds with $3k on the table?" ie. how much would you be willing to pay for that info? Now change the amount -- the fish has $10k on the table --how much is that info worth to you now?" Or in tourneys, what is the value of not getting knocked out -- NOW, how much does that change if your table is strong vs. weak? Black-Scholes models should have some insights for figuring out the answer to these questions.

-g