Is there any merit to playing in a way which attempts to increase your chances for a winning session, as opposed to just trying to maximize EV.

I don't mean "quitting when you're ahead," but things like "waiting for a better spot", or playing hands in a lower variance way.

Specifically I'm asking in a live play context where you don't see many hands, you can't be truly sure of winrate, and minimizing risk of ruin is paramount.

https://forumserver.twoplustwo.com/3...-play-1725357/

Check out this thread. There has been some stuff written about this topic.

Long story short: Depends on the game. Vs. nits you'll bleed out, vs. donkeys or maniacs it's very good to play that way.

It's all one long session, thinking about doing anything differently to what makes the most money on any given position just because you're up or down for the day is bad

If you want to maximise your chances of a winning session, sit down, open jam, hope nobody has aces and immediately leave

It really isn't one long session at all. Kelly critertion and all that.

On top of that do you even know what expectation means? If you randomly open jam a hand it has an EV and that EV is very low. Playing optimally gives you the highest chance of having a winning session.

How's it not (assuming it's a cash game anyway)? This just goes against common sense.

Yes, this can be done..it's called overbetting.

IS it ideal poker? No. Does it Maximize EV? No. But do you want to win a small amount of money and limit variance? Overbetting your big hands will accomplish this.

Suppose you are on the button, you pickup pocket Aces. Under the gun raises to $10 and there are 3 callers ahead of you. You can reraise to 100 or go all in. You will lock in the 40 or so profit. You may, may get called by 1 person. They could beat you, nothing you can do. But for the most part you will double up + 30 extra. It's not a terrible spot to be in, but it's far from ideal.

Suppose you decide to play J-10 suited on the button after a table full of limpers. An aggressive player would raise and build the pot from position. But lets say you call and the flop comes Q-8-9. The Queen and 8 are hearts and you are holding spades. There's a big bet in front of you. It's probably a hand like Q-J or A-X of hearts. You can pop here and take down this pot. Only a really loose player would call with just top pair and you're huge. You might get sucked out by a flush draw but there's nothing you can do. By overbetting you probably take down the pot with limited risk. Alternatively, you can just check-call it down to see if your nuts hold up and just make a big bet on the river.

Kelly Criterion is perfectly compatible with "one long session". You use KC when you're making a wager repeatedly forever. "Forever" is as long a session as it gets.

This is wrong. If your opponents are playing exploitably, your chance of winning the session is higher if you exploit them.

Sounds like OP is playing in games he can't really afford to play and is going to play them anyway so avoiding high variance low EV spots is probably good. That being said if players who ask these type of questions actually knew enough to be able to make those adjustments then they wouldn't be in those situations.

You're confusing optimally and game theory optimal. Playing optimally is exploiting villains mistakes.

optimally
adverb
in the best or most favourable way.

This doesn't change the fact that Poker is one long session. You have yet to explain how it's not.

Great look condescendingly pulling out the dictionary. On a poker forum, the word "optimally" will almost always be interpreted to mean "GTO". If you're using the layman connotation "best", you should say so.

Why are you so mad? It's ok you got two words confused, understandable even when people throw around words they don't get all the time. The definition was just so you could see that it was actually a really good choice of a word.

It's to do with the fact that in poker we usually don't have to care about anything more than playing as optimally* as we can against villains because all this type of stuff is accounted for in things like BRM. If however you are playing in games that you are a winner in but it's too large an amount of your bankroll** then some of the bets that we would be making if we were rolled properly now become bad because our risk of ruin is too high.

Whether you could actually be a winner in games where you were avoiding all of these spots (which in reality you never know what they look like so see **) is a different question, I'd imagine so in weak games.

*See my previous post for the definition of optimally.

** Well really you probably shouldn't be playing in them but people do.

I'm not mad; I was just pointing out that you made an unprovoked douche of yourself. Do what you will with that info (it seems you're electing to double down, which is fine, double down on 18 for all I care).

Anyway, back on topic: My disagreement isn't with anything you're saying, but with the idea that "Poker is one session" implies it's wise to buck the Kelly Criterion. We simply have different ideas of what that phrase means, and debating it is unlikely to be interesting. I'm pretty sure ~every good poker player simultaneously 1) understands what you're saying and 2) views Poker as one big session, so imo you're in the minority with your belief of what the phrase means.

That's the whole point. If you sit down and open jam one hand and get up, your chance of winning the session will be very high, but your EV will be very negative.

That's not really what OP was thinking, though. Of course sometimes one should sacrifice EV to reduce variance, but it should happen so rarely if one is properly rolled that I think a beginner would be better off not worrying about it.

I don’t want to get into your argument or ‘pick sides’, but when somebody says “optimally” instead of “GTO” or “game theory optimally” you shouldn’t assume he means “GTO”. Otherwise he probably would have said GTO.

I also think the discussion if (cash game) poker is one long session is pretty pointless. Even if we could agree that it’s just one long session, you could argue that somebody who played 1/2 for 3 hours in 1988 and another 3 hours at 50/100 in 2018, basically played a 6 hour session in two different “sports”.

If we turned that “shove the first hand” example around and assume our villain is open shoving 500BB 100% of the time, we get to a situation where we certainly want to fold a hand that has 55% equity against said 100% range and sacrifice some EV and wait for a hand that has a lot more equity.

Not certainly. It depends on the size of our bankroll. A 10% edge is huge if we're rolled for it. I'll admit many players are not rolled for this particular spot, though.

Sorry, I forgot to add that our opponent won’t rebuy if he loses his stack.

Is it okay to trade some EV for a decrease in variance? Sure. It depends on what's important to you. It's okay to trade some EV to play a style that is more fun if you care more about fun than making money.

The most obvious example is if you are on a short bankroll. I generally don't try to maximize EV when am new to a game. I play a more passive, sub-optimal style while getting comfortable and gathering data about new opponents. But I know that I am trading EV for decreased variance.

For an inexperienced player who is learning the game, it makes sense to considering passing up situations which might be +EV but are higher variance with a greater degree of difficulty.

Ignoring the fun argument, which is perfectly valid if you're a rec on an entertainment budget, then if you're playing with a proper bankroll then this should never come up outside of running like the sun in a huge donkament that you've satted in to

Not everyone is playing with a proper bankroll. You might also be in a situation where you are taking a shot at a bigger game because it is juicy and you are not properly bankrolled for that level (but have enough of a bankroll for your current level if you lose) and you might want to play a more conservative style that doesn't maximize EV.