![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
What bits matter
May. 7th, 2009 11:46 amI think that one of the most common errors made by people trying to improve their poker skills is to overestimate the importance of "big river decisions" and to underestimate the importance of small pre-flop decisions".
The reasons for such an error are not hard to find. First, because it's easier to analyze the "final" decision in a hand (be that a jam pre-flop or an all-in bet on the river), the temptation to look at solvable rather than fuzzy problems is irresistible to most minds. Secondly, the river decisions usually involve more money. You might be facing a decision that will win or lose you 50 big blinds on the river, whereas the preflop decision is only for three big blinds. Thirdly, people make note of and remember such "big" hands, whereas they tend to forget the smaller, more frequent decisions preflop. The very rarity of the "big" hand makes it more memorable.
My training in limit has long led me to look at the more frequent marginal decisions preflop than the marginal decisions on the river, and I think that the principal still holds in No Limit, although the difference in significance is obviously less. Because bet sizes increase, you can get away with a range of styles pre-flop that would be suicidal in limit.
However, I have come to the conclusion that I really need to focus on my river play, solely because, even though the situations are relatively uncommon compared with pre-flop decisions, they are large enough (and frequent enough) to make a significant (but not overwhelming) contribution to my overall win rate.
For example, I see about 15% of Full Ring Flops and get to a showdown about 20% of the time. That means I am making some kind of river bet decision 3% of the time, less those times when I am all-in pre-river. Given the number of short-stackers playing, let's assume that this shaves off a third of all river hands. That leaves 2% of hands, or roughly four hands an hour when I am multi-tabling. If I can increase my average return on those hands by just a big blind per hand, I'm increasing my overall rate by four big blinds an hour, or (at present rates), anything between 10% and 30% of my expected win per hour.
In other words, I need to look at river play more carefully. (Although, note that this still leaves between 70% and 90% of win in other parts of the game. I may well try to do a breakdown of where that money is actually coming from, because by so doing I can see where I need to focus more).
So, what can I do to increase my river win rate? There are a few possibilities that spring to mind.
1) Throw in a higher percentage of triple-barrels when I have no showdown equity. If my bet is pot-sized, I need to generate folds more than 50% of the time.
2) Bet more often when I have some showdown equity. This is the area where I have been really trying to improve and I think I have achieved some success.
3) Fold more often when I am beaten.
4) Call (or raise) more often when I am winning.
5) Let go of hands on the turn that will lose on the river.
6) Stay in with hands on the turn that are going to win on the river.
The last two are really "turn decisions" and part of the overall rate, so I'll worry about those on another day.
Before you can decide how to improve, you have to decide what you are aiming for. What's an optimum win rate at showdown? I'd been looking for about 56% to 58%, but 6-max winners seem to be getting by on 49%. The big difference, of course, is that they make a lot of their profit (well, all of it) from eliciting folds. Their showdown performance is a minor loss. I, meanwhile, have a high showdown win, but am losing money on non-showdown hands.
This is exactly how Harrington thinks it should be. Indeed he states that you have to win enough at showdowns to make up for the inevitable dribbling away in non-showdown hands.
But good 6-max players don't see it that way. They plan to make their money from that "inevitable dribbling away" by other players, while not losing too much when they get called.
Looking at my own performance, it seems clear to me that I should be winning a lower proportion of showdowns and generating a higher proportion of folds. This would dictate a laggier style (even at full ring) that would inevitably generate higher volatility. My actual target is for the "flat red line" in Hold Em Manager (i.e., breaking even in non-showdown hands) with a more modest upward slope in the blue showdown line. My gut feeling here is that I would be aiming for a 52% to 54% showdown win and (and this is the crucial point) my lower proportion of showdown wins would enable me to win a higher average amount of money when I bet the river. Clearly if you never bluff, you arfen't going to be called that often once your opponents get used to your style.
What does all this entail? Does it shift back as far as pre-flop?
I don't think so. I don't think I need to loosen up pre-flop at all. What it looks to me that I need to do is to be less willing to release on the flop and the turn. I may be putting in money which is marginally negative Ev (although even that isn't certain) but it should generate a considerably higher number of folds (if I jam first) and a better payoff for my showdowns when I win.
___________
The reasons for such an error are not hard to find. First, because it's easier to analyze the "final" decision in a hand (be that a jam pre-flop or an all-in bet on the river), the temptation to look at solvable rather than fuzzy problems is irresistible to most minds. Secondly, the river decisions usually involve more money. You might be facing a decision that will win or lose you 50 big blinds on the river, whereas the preflop decision is only for three big blinds. Thirdly, people make note of and remember such "big" hands, whereas they tend to forget the smaller, more frequent decisions preflop. The very rarity of the "big" hand makes it more memorable.
My training in limit has long led me to look at the more frequent marginal decisions preflop than the marginal decisions on the river, and I think that the principal still holds in No Limit, although the difference in significance is obviously less. Because bet sizes increase, you can get away with a range of styles pre-flop that would be suicidal in limit.
However, I have come to the conclusion that I really need to focus on my river play, solely because, even though the situations are relatively uncommon compared with pre-flop decisions, they are large enough (and frequent enough) to make a significant (but not overwhelming) contribution to my overall win rate.
For example, I see about 15% of Full Ring Flops and get to a showdown about 20% of the time. That means I am making some kind of river bet decision 3% of the time, less those times when I am all-in pre-river. Given the number of short-stackers playing, let's assume that this shaves off a third of all river hands. That leaves 2% of hands, or roughly four hands an hour when I am multi-tabling. If I can increase my average return on those hands by just a big blind per hand, I'm increasing my overall rate by four big blinds an hour, or (at present rates), anything between 10% and 30% of my expected win per hour.
In other words, I need to look at river play more carefully. (Although, note that this still leaves between 70% and 90% of win in other parts of the game. I may well try to do a breakdown of where that money is actually coming from, because by so doing I can see where I need to focus more).
So, what can I do to increase my river win rate? There are a few possibilities that spring to mind.
1) Throw in a higher percentage of triple-barrels when I have no showdown equity. If my bet is pot-sized, I need to generate folds more than 50% of the time.
2) Bet more often when I have some showdown equity. This is the area where I have been really trying to improve and I think I have achieved some success.
3) Fold more often when I am beaten.
4) Call (or raise) more often when I am winning.
5) Let go of hands on the turn that will lose on the river.
6) Stay in with hands on the turn that are going to win on the river.
The last two are really "turn decisions" and part of the overall rate, so I'll worry about those on another day.
Before you can decide how to improve, you have to decide what you are aiming for. What's an optimum win rate at showdown? I'd been looking for about 56% to 58%, but 6-max winners seem to be getting by on 49%. The big difference, of course, is that they make a lot of their profit (well, all of it) from eliciting folds. Their showdown performance is a minor loss. I, meanwhile, have a high showdown win, but am losing money on non-showdown hands.
This is exactly how Harrington thinks it should be. Indeed he states that you have to win enough at showdowns to make up for the inevitable dribbling away in non-showdown hands.
But good 6-max players don't see it that way. They plan to make their money from that "inevitable dribbling away" by other players, while not losing too much when they get called.
Looking at my own performance, it seems clear to me that I should be winning a lower proportion of showdowns and generating a higher proportion of folds. This would dictate a laggier style (even at full ring) that would inevitably generate higher volatility. My actual target is for the "flat red line" in Hold Em Manager (i.e., breaking even in non-showdown hands) with a more modest upward slope in the blue showdown line. My gut feeling here is that I would be aiming for a 52% to 54% showdown win and (and this is the crucial point) my lower proportion of showdown wins would enable me to win a higher average amount of money when I bet the river. Clearly if you never bluff, you arfen't going to be called that often once your opponents get used to your style.
What does all this entail? Does it shift back as far as pre-flop?
I don't think so. I don't think I need to loosen up pre-flop at all. What it looks to me that I need to do is to be less willing to release on the flop and the turn. I may be putting in money which is marginally negative Ev (although even that isn't certain) but it should generate a considerably higher number of folds (if I jam first) and a better payoff for my showdowns when I win.
___________
