Running it twice can change the EV of future hands
Hi everybody.
As we all know, the EV of a hand is not changed by the number of times we run it out. However, expected stack sizes are changed. And since stack size alters our positional advantage (with larger effective stacks favoring the player in position), we can leverage this effect to maximize our positional advantage in later hands.
Consider this example:
Suppose that the player on our left has a short stack, which we will define as less than half of the minimum buy-in. Then running it once is guaranteed to increase his stack size: A win doubles him up, and a loss busts him out, forcing him or another player to buy in again as more than twice his current stack. But if we run it twice and get a split pot, his stack size is unchanged. Likewise, the probability of growing his stack through a win or rebuy is reduced. Keeping the effective stacks small reduces our positional disadvantage against this player in future hands, reducing EV loss. Conversely (and by analogous logic) we should prefer to run it once against a short stack on our right, as this will increase his expected stack size and thereby increase our positional advantage and EV gain against him in future hands.
As I understand it, the problem of determining our EV advantage/disadvantage against a table with varying stack sizes is quite difficult in general. We might have to run simulations to determine how stack size affects our positional advantage against the table as a whole, which is why I have chosen to focus on a single opponent. With this in mind, I think we often maximize EV by running it twice against short stacks on our left, and once against short stacks on our right. But these rules are only guaranteed to maximize our expected positional advantage in future hands against one opponent, and not against the table as a whole.
The number of runouts is often considered a personal decision with no right or wrong answer, but I think that I have made a strong argument that there is a mathematically correct solution that maximizes EV against optimal players. As such, it seems clear to me that the number of runouts affects win rates, and not just volatility. This is (to the best of my knowledge) a novel observation, but please correct me if I am wrong.
The logic in this post could probably be extended to produce additional heuristics about the optimal number of times to run it out in other situations with the goal of maximizing EV in future hands. Such heuristics could produce real (though likely marginal) gains in a player's win rate. Any thoughts?
Sam
7 Replies
For sure, another point to consider is that players might be more aggressive with their draws in future if they know you run it twice (since they know they will be less likely to bust). Which is definitely something that hurts your EV.
Hi everybody.
As we all know, the EV of a hand is not changed by the number of times we run it out. However, expected stack sizes are changed. And since stack size alters our positional advantage (with larger effective stacks favoring the player in position), we can leverage this effect to maximize our positional advantage in later hands.
Consider this example:
Suppose that the player on our left has a short stack, which we will define as less than half of the minimum buy-in. Then running it once i
running it twice helps people split pots instead of losing them. This keeps them from going on as much of tilt as they would had they of lost
its better to run it once, and tilt them twice; tilt them when they lose, and tilt them cause you dont run it twice.
Interesting take Sam. Personally I usually prefer not having the short stacks on my left in live cash. The reason is that I'm often playing exploitatively by raising large with a relatively wide range to isolate other deep stack players, with a plan to outplay them later in the hand. If a short stack player on my left plays a TAG strategy against my opens they will destroy me, and force me to raise smaller or tighten up.
What I find interesting about running it twice is that almost every player, good or bad, gets it wrong on whether it’s beneficial to them. Most of those who want to RIT are bad players. Most who don’t want to RIT are good players. This is the complete opposite of what is beneficial to each.
The reason is variance. A losing player stands little chance of ever showing a win long term. He can though expect some short term winning sessions. Maximizing variance is the best chance the losing player has to show some winning sessions. RIT reduces variance, so it makes it less likely that the losing player will have a good session.
For a winning player, the opposite is true. If the winning player reduces variance it’s beneficial. RIT obviously reduces variance so it would be beneficial for a good player.
I would not argue, though with any good player who does not RIT. A good player should be able to handle the variance in poker and should manage his bankroll well enough that losing a stack has no significant impact on it. The fish, though, is just killing himself by RIT.
Interesting proposition and takes in the thread. I like to run it twice. The logic being, if I bust the fish they will often leave. Sometimes when they double up they leave. If it splits they are more likely to stay and my future EV is increased, because the half of the chips that they won from me are more likely to come back in my direction in future hands.
If I end up with a large stack turns out the good players at the table are usually going to have large stacks as well. Most of my money is made from fish not other deep stacked regs.
I also enjoy reduced variance.
Interesting proposition and takes in the thread. I like to run it twice. The logic being, if I bust the fish they will often leave. Sometimes when they double up they leave. If it splits they are more likely to stay and my future EV is increased, because the half of the chips that they won from me are more likely to come back in my direction in future hands.
Future hands where you'll also run it twice?
This logic seems backwards to me. If the point is to maximize the chance that the fish loses their money to you, you should run it once; that way you avoid a chopped pot where the fish is more likely to dump their money later to someone else who only runs it once.
For me, I'm swayed by the "repeated game" logic that I don't want to be known as someone who will give you a chance at a split pot if you get it in with me. My own personal rule is that I never run it an even number of times. This way, when you get all-in with me, there is a guarantee someone will win money and someone will lose it--no chops (unless it's an actual hand-value chop). But since no one ever wants to run it more than twice, it means in practice I always run it once.
The fact that the fish can lose the money to other players is definitely a good point to consider which does lower our future EV, but I don't think that factor is enough to weigh me into wanting to run it once. Regardless, with the odds of the fish losing his money to another reg our average future EV is still higher than it would be if he left. I don't really believe that fish are even paying all that much attention to who is running it once or twice and there is a decent chance that you'll never play them again since they usually aren't regulars. Fish love to overfold and play passive regardless of whether you run it once or twice imo.
If you change the game type to a heads up cash game my point is further accentuated. Say you get it all-in with 75% equity, you get the following outcomes:
Run it once:
W: 75% + 75ev
L: 25% -25ev
EV: 50
Run it twice:
WW: 56.25% +56.25ev
WL: 37.5% +20ev (would be 0ev, but since fish stays, we gain a bit)
LL: 6.25% -6.25ev
EV: 70
You can add more players to dilute this, but at the end of the day they're just a bit of noise and on some occasions, can be a deterrent from doubling up.
