what poker players can learn from Blackjack pros

[BillC]

Certainty equivalence (CE) a standard concept in utility theory, and is used in blackjack teams to account for variance.
I often hear poker analyses that say to do something since the EV is positive. Well, unless your bankroll is large enough that you can ignore variance, you should take the variance into account. For instance, if you are getting 31 to 1, is it OK to draw to a runner-runner flush? Answer depends on the CE, which depends on your the EV, the variance (which is huge), risk-aversion and your bankroll. I might run some numbers for some common situations if there is any interest in this.
in this. Fyi, There is a write-up about CE in the rec.gambling.blackjack faq and Ithink also on the bjmath.com site.

[Lou Krieger]

Though I've never heard of poker players talk about CE in terms as cogent and succinctly as you did, I do believe their actions show a distinct awareness -- tacit though it may be -- of this concept.

I've heard numerous players tell me they prefer to avoid games with a couple of maniacs in them, precisely because they want to minimize the variance associated with them. Some players, pratricularly those on short bankrolls, also change their playing style in an attempt to minimize variance, although this cannot completely be managed by one's own playing style, since vairance is also dependent on how others in the game play.

I, for one, would like to see you run some numbers, as long as you can prsent them in a format that even those of us who are mathematically unsophisticated can understand and think about.
_____
Lou Krieger
Raise your game with Lou Krieger, author of "Poke rFor Dummies," at Royal Vegas Poker.
http://www.royalvegaspoker.com/lou

[Bozeman]

I think that CE is less important (it is closely approximated by EV) in poker because edges are larger, so that many (most?) winning players adequately account for variance by sufficient bankroll size. If this is off base, perhaps you can clarify,
Craig

[Dynasty]

Does royalvegaspoker advertise on 2+2?

Why do I have this supspicion that Mason won't like links to online sites that don't pay him for it?

[Nottom]

To me the variance on any 1 particular play in a poker game is probably not as big a deal as your play overall. Whether or not you draw to runner runner flushes when getting huge pot odds isn't going to have a huge affect on your bankroll in the long run. If you are playing with a reasonable bankroll to begin with I just don't see how the risk on any one hand can be enough to not warrent making a play if you confidently believe it is +EV.

[Zele]

An important expecption would be in tournament play, where players are purposely underbankrolled.

I think the problem with applying CE/Kelly Criterion type thought to poker is that there is no "house" that will accept all bets of whatever size; that the amount of action you can get is dependent on your edge as your opponent sees it.

[Nottom]

Very true, I was going to make a note of that in my post but for somereason I didn't get around to it. Proper tourney play is much more conservative than ring game play for a lot of the reasons noted in this thread.

[BillC]

If you have a close decision, and big variance, the CE depends on your bankroll. In fact, the CE alsways depends on the BR.

The point is that decisions should factor in variance and bankroll and utility (risk aversion, acceptable risk of ruin)

A lot of analyses of decisions in poker come out to be very close, so CE should play a role. There is a rule of thumb here--if it is borderline EV, chose the lower variance route. This goes for all decisions, not just cards.

In poker, you can talk about CE for micro-decisions within a hand, or macro ones, like whether you can afford to attack that juicy but wild 20-40 game.

How big must your bankroll be? That's what poker can learn form bj theory. The usual 3 (or 2) sigma EV model is incorrect because of the absorbing barrier at zero (i.e. you might go have gone broke before your EV is 3 sigma)

I'll eventually do some numbers.

[J.A.Sucker]

[ QUOTE ]
Proper tourney play is much more conservative than ring game play for a lot of the reasons noted in this thread

[/ QUOTE ]

This isn't totally true. A better way of stating it is that your response to OTHER players' actions is much more conservative in tourney play than in ring games, but when nobody has done anything, it's often better to be looser, since the other players will be acting more conservative. Changing speeds in a tourney is what makes the good players win; they pick up so many antes/blinds when people are tightening up. The best time for this is after a rebuy period has finished - nobody wants to be that first person busted. Also, near the bubble is a good time, since you can often steal blinds, which are usually very large compared to stack sizes at this juncture of the tourney.

[J.A.Sucker]

I don't agree with your statements very much, and here's why.

1. In poker, if you are confronted with a neutral EV situation (or close to it), then other factors will swing your decision. Sometimes, it's better to fold and wait, while other times, it may be well worth it to draw to a runner-runner hand, a gutter, or other longshot if you know that you will secure other advantages. A primary example is trying to draw out in a close spot headsup against someone who is likely to go on tilt if you suck out. Another example is if you have a small pocket pair and may not be getting quite correct odds preflop, but decide to play it, with the hope that if you hit a set, your image and or attitude will be improved. This isn't really tilting in the proper sense of the word, since the mistake is small, but the payoff is huge. I will frequently make this "mistake."

2. The overall standard deviation increase of always going with these decisions, versus always going against these decisions is very small. A good friend of mine, who posts frequently on this forum, usually plays these situations, while I usually fold (he will agree with this), and our hourly earns aren't significantly different. However, my SD is about 10 BB/hr, while his is about 12 or so. This isn't a very large increase. He may need a few more bets in the bankroll, but it's not terrible.

3. The number of high-variance/low EV situations doens't come up very often in poker. Therefore, these decisions matter little. See example 2 for details. Most games don't require you taking one off to see if you spike your set on the turn, since the pots don't usually get that big preflop, for example.