The Cost of Mistakes in Poker

[W. Deranged]

Okay,

This post is largely inspired by the "Devastating 30 BB Hand" post where the OP folded AQ unimproved which would have ended up winning a three-way showdown against a bunch of monkeys. Discussion arose as to the general concept of "never fold for one bet in a big pot."

My point is this: In poker, what seems like a mistake of size x is almost never a mistake of size x . That is the nature of a probabilistic and incomplete-information enterprise: the uncertainty that surrounds the cards to come and our opponents possible holdings mean that mistakes that we makes are usually not "full" mistakes, because that would only be the case if, given all of the information available at your disposal at the time of the decision, the hand would play out the same way every time . That's almost never true.

Here are a few examples:

1. From the "Devastating 30 BB Hand": How big a mistake was OP's fold on the river? The answer is not 30 BB! If you review the hand, and the fact that multiple bets went in on every street, the pot was very multi-way, and so on, hero's hand is holding up unimproved there a very, very small percentage of the time. For his mistake to be a "30 BB mistake," his hand would have to be good in 100% of scenarios where like the one villain faced. That's a patently absurd idea. The OP's hand needs to be good about 3 1/3% of the time to make a call correct, true, but think about it this way:

-If hero's hand is actually good 10% of the time (which is ridiculously high), his mistake is not a 30 BB mistake, but about a 2 BB mistake (for every 9 BB he invests and loses, he makes 30 1 time; so his expectation on a call is like 2.1BB). A big mistake, certainly, but far different from the catastrophic 30 BB some people were suggesting.

-If hero's hand was good 5% of the time, his mistake is about a 1/2 BB mistake or so; still big, but again, not monumental.

-Now it's very hard to estimate numbers down in the 0-5% range, but let's assume his hand is only good 2% of the time. That means that for ever 49 times he invests a BB and loses, he wins 30 BB once. In other words, his expectation is like -19/50 BB, or about -2/5 of a big bet.

Observe, that calling on the river when you only have a 2% chance of winning is almost as big a mistake as folding when you have a 5% chance of winning !

2. Another example: Let's consider making a bad turn call. Let's say that in a 6 BB pot (including villain's bet to you), you find yourself with presumably 5 outs. You call even though you shouldn't. How big a mistake is this? The answer is far less than 1 bb . With five outs, you'll improve a little less than 1 in 9 times. For ease, let's say you'll improve exactly 1 in 9 times. Then you'll win 6 bets one time for every 8 times you lose, and so your expectation is -2/9 BB.

3. Let's consider a bad turn fold: let's say that you have those same 5 outs, but in an 10 BB pot. You mistakenly fold. How bad is your mistake? Well, for every 8 times you fail to improve, you'll improve once and make 10 BB. So your expectation is 2/9 BB on a call. So folding is a 2/9 BB mistake.

My basic point is this: Since poker is a probabilistic and incomplete information game, we have to quantify "mistakes" in terms of expectations. There are many situations (folding hands on the river is the most obvious, but others including pre-flop decisions, chasing longshot draws, and so on) where poker players (of all levels) tend to put too much emotional weight on specific kinds of decisions. This ranges from silly ideas like "never draw to an inside straight," to "avoid folding the winner in a huge pot" and "poker's more about the bets you save than the bets you win." Often times, as I think the "Devestating 30 BB Hand" thread demonstrated, the emotional weight players place on certain kinds of decisions obscures the underlying poker principles going on. So, start thinking in terms not of "devastating" or "catastrophic" or "fishy" mistakes. Think instead of expectation, and in conceiving of "mistakes," think about "how big a mistake is this?"

That is all.

[einbert]

Good post. And while this kind of case is relatively easy to analyze, if you can come up with a ballpark "HeroIsGood%", it's much more difficult to analyze other mistakes. Especially on the flop and turn, where figuring up our overall equity, and then estimating our equity if our opponent reraises us/calls us takes quite a bit of work. And problems like these are why LHE is incredibly difficult to learn, especially to learn from experience. And it's why many players "tilt" by making bad folds on the river, because psychologically they hate giving up that 1BB most of the time. It is extremely hard to wrap our minds and emotions around all the counterintuitive stuff that goes on in LHE.

[damaniac]

Awesome. I first saw Bobbyi expound on this re: river calls, and I and others have brought it up from time to time, but you just did it in a detailed yet very clear way. This should be bumped periodically.

[jason_t]

bobbyi was the first on this board to explicate this important point about six months ago but your post is as usual very nice.

[W. Deranged]

[ QUOTE ]
bobbyi was the first to explicate this important point about six months ago but your post is as usual very nice.

[/ QUOTE ]

I remember reading some kind of post about this but I have no idea where and so I figured it would be worth reminding everyone.

[Hobbs.]

To add on to your post einbert, some of the major factors that make flop and turn decisions much more difficult to quantify stems from the fact that LHE IS a game of incomplete information and on the flop and turn we have less information at our disposal to make correct choices. Furthermore a decision can be swayed one way by the lure of things like implied odds to make up for an apparent -EV situation.

When people say things like preflop doesn't matter all that much and other such stuff, I like to think about it in terms of error bars around some true expectation and as the hand progresses these error bars start to shrink with the information that becomes apparent to us. Basically, all streets have important decisions in terms of your expectation and as you improve as a player your 'error bars' start to decrease and you have a better grasp on what the true expectation is for a given decision.

Basic stuff that is not always apparent and worth repeating...

[jason_t]

[ QUOTE ]
[ QUOTE ]
bobbyi was the first to explicate this important point about six months ago but your post is as usual very nice.

[/ QUOTE ]

I remember reading some kind of post about this but I have no idea where and so I figured it would be worth reminding everyone.

[/ QUOTE ]

The basic reason that most fail to grasp this point is that a lot of people don't think in terms of EV and weighted decisions. Thinking in these terms clarifies almost every poker concept.

[W. Deranged]

[ QUOTE ]
[ QUOTE ]
[ QUOTE ]
bobbyi was the first to explicate this important point about six months ago but your post is as usual very nice.

[/ QUOTE ]

I remember reading some kind of post about this but I have no idea where and so I figured it would be worth reminding everyone.

[/ QUOTE ]

The basic reason that most fail to grasp this point is that a lot of people don't think in terms of EV and weighted decisions. Thinking in these terms clarifies almost every poker concept.

[/ QUOTE ]

/images/graemlins/heart.gif Indeed /images/graemlins/heart.gif

[W. Deranged]

[ QUOTE ]
To add on to your post einbert, some of the major factors that make flop and turn decisions much more difficult to quantify stems from the fact that LHE IS a game of incomplete information and on the flop and turn we have less information at our disposal to make correct choices. Furthermore a decision can be swayed one way by the lure of things like implied odds to make up for an apparent -EV situation.

When people say things like preflop doesn't matter all that much and other such stuff, I like to think about it in terms of error bars around some true expectation and as the hand progresses these error bars start to shrink with the information that becomes apparent to us. Basically, all streets have important decisions in terms of your expectation and as you improve as a player your 'error bars' start to decrease and you have a better grasp on what the true expectation is for a given decision.

Basic stuff that is not always apparent and worth repeating...

[/ QUOTE ]

Nerd.

[jason_t]

[ QUOTE ]
[ QUOTE ]
To add on to your post einbert, some of the major factors that make flop and turn decisions much more difficult to quantify stems from the fact that LHE IS a game of incomplete information and on the flop and turn we have less information at our disposal to make correct choices. Furthermore a decision can be swayed one way by the lure of things like implied odds to make up for an apparent -EV situation.

When people say things like preflop doesn't matter all that much and other such stuff, I like to think about it in terms of error bars around some true expectation and as the hand progresses these error bars start to shrink with the information that becomes apparent to us. Basically, all streets have important decisions in terms of your expectation and as you improve as a player your 'error bars' start to decrease and you have a better grasp on what the true expectation is for a given decision.

Basic stuff that is not always apparent and worth repeating...

[/ QUOTE ]

Nerd.

[/ QUOTE ]

When are you two getting married?

[oxymoron]

This is a great post. My only add on is you also have to consider how many times you have mistakes to make on each street. If you make a "loose" call on a 30BB hand once in 500 hands it is a mistake but one you do not make often. If you make a smaller mistake but make it more often it is a bigger mistake due to the multiplication of your mistakes.

[W. Deranged]

[ QUOTE ]
[ QUOTE ]
[ QUOTE ]
To add on to your post einbert, some of the major factors that make flop and turn decisions much more difficult to quantify stems from the fact that LHE IS a game of incomplete information and on the flop and turn we have less information at our disposal to make correct choices. Furthermore a decision can be swayed one way by the lure of things like implied odds to make up for an apparent -EV situation.

When people say things like preflop doesn't matter all that much and other such stuff, I like to think about it in terms of error bars around some true expectation and as the hand progresses these error bars start to shrink with the information that becomes apparent to us. Basically, all streets have important decisions in terms of your expectation and as you improve as a player your 'error bars' start to decrease and you have a better grasp on what the true expectation is for a given decision.

Basic stuff that is not always apparent and worth repeating...

[/ QUOTE ]

Nerd.

[/ QUOTE ]

When are you two getting married?

[/ QUOTE ]

Two weeks after you and Shant.

[Elevens]

I haven't even read it yet, but...

Good post.

EDIT: I've read it now, and I'l change my response to:

Great post.

[James.]

nicely done.

[QTip]

I was thinking chief had started this about 6 months ago.

At any rate, at that time we talked about doing this exercise in posts...quantifiying mistakes. (btw, I like hte section in top on page 252 and also sklanksy's "The 8 mistakes in poker"). So, I think doing so lasted about 10 posts and then it died. I like the idea though.

W.D. Somewhere in your 1000 word reply essays, you should start putting mistake analysis thingies.

Nice post.

[belloc]

I played in a hand live last night that had similar issues as those discussed here in the "Devastating 30BB Hand", except that I think I was good here more often than the AQ would have been in that hand (mine was an AA, and I ended up overcalling on the river instead of folding).

I'll post it in a separate thread as soon as private joker helps me remember the details of the turn. There is one important action I'm unsure about that might affect the decision.

After the hand, Zach and I talked about it; his first reaction was that MHIG there like never, and my continuing with the hand was just a matter of being in love with AA, and that I shouldn't have put in any more bets; I thought it would have been a mistake to fold. Pot was big, opponents were looooose aggro. Details forthcoming.

Good post, Will.

[newhizzle]

[ QUOTE ]
nicely done.

[/ QUOTE ]

[QTip]

[ QUOTE ]
bobbyi was the first on this board to explicate this important point about six months ago but your post is as usual very nice.

[/ QUOTE ]

AH...I remember which post you're talking about now. I've searched for that a couple times now without any luck.

It seems like sthief (only God knows which name he was using in that thread, but I think it was 09), was in that post. bobby made his post, and sthief said something like "you're quickly becoming one of my favorite posters".

Anyway have the link at all? I'd like to read that again.

[W. Deranged]

[ QUOTE ]
W.D. Somewhere in your 1000 word reply essays, you should start putting mistake analysis thingies.



[/ QUOTE ]

Will do...

[Haggis McHaggis]

I like this post a lot. I think it serves to illustrate nicely that you don't have to play perfectly to be a winner, you only have to play better than your opponents (hence the importance of table selection).

Thanks for taking the time to make this post /images/graemlins/smile.gif

[damaniac]

I couldn't find it either, but fortunately I took the liberty of putting some of my favorite posts on hard copy, and seeing as how the alternative is studying for Contracts, I believe I'll repost it (I presume this is it, there is another very similar post in the same thread, all in response to Tommy Angelo's river fold in a big pot with AA I think.)

(Hopefully this is permissible)

-----
In response to: Folding the winner in this massive pot for just 1 more bet closing the action would take me weeks, maybe months to get over.

Bobbyi:
This is very results-oriented thinking. If you folded before the flop with 53o and it turned out that you would have flopped a full house and won the biggest pot in history, would you lose any sleep over it for months? I hope not. Because even though it only costs one bet to call and you would have won a huge pot for that one bet, it was a good fold. The actual results are irrelevant. The same is sometimes true on the river. There are situations where you are getting 15-1 on a river call and you know that your hand is only good 1 in 25, given the opponent. So you fold. And your opponent shows you this his hand revealing that this was one of the very occassional times that you were best. Was this a disaster? NO. Given the information you had, you made a good fold. The fact that it cost you a big pot is irrelevant. Of course, if your assessment of the situation was wrong and your hand was really good one time in five then you made a terrible fold. We all understand that. At least I hope that anyone posting on the mid and high stakes board understands that. But poker is a judgment game, and there are cases where you really can be sure enough to fold in a big pot, though your hand isn't being beat quite 100% of the time, and when you call in those spots, you are making a small mistake. But small mistakes add up. And winning poker consists largely of small edges that add up. So rationalizing a bad call as only a small mistake is a dangerous path to take. Being a skilled experienced player is about having the judgment to recognize these situations. And if you do, you shouldn't be upset when you occasionally lose a pot as a result. If you don't, then it is true that until you acquire that judgment, it is safest to just always call in big pots, as everyone in the small stakes forum seems to do since they treat SSH as some sort of bible. But this was a good fold, at least given the information Tommy provided us. Well played, Tommy.

[QTip]

Thx damaniac, but this isn't the one i was thinking of. The OP in the other one is a hand from stheif.

[duckman]

Excellent and discussion provoking. (most of your posts are)
Even enticed jason t to talk some trash.

[bobbyi]

[ QUOTE ]
Thx damaniac, but this isn't the one i was thinking of. The OP in the other one is a hand from stheif.

[/ QUOTE ]
link (http://archiveserver.twoplustwo.com/showthreaded.php?Number=1813536)

Similar post from astroglide (http://archiveserver.twoplustwo.com/showflat.php?Number=1629302), which actually predated mine but I never saw that post until we had the same discussion again here (http://forumserver.twoplustwo.com/showflat.php?Number=3708995).

[TripleH68]

[ QUOTE ]
My basic point is this: Since poker is a probabilistic and incomplete information game, we have to quantify "mistakes" in terms of expectations. There are many situations (folding hands on the river is the most obvious, but others including pre-flop decisions, chasing longshot draws, and so on) where poker players (of all levels) tend to put too much emotional weight on specific kinds of decisions.

[/ QUOTE ]

Good post. Always appreciate your insight.

On folding for one more bet on the river...

Hero is MP with Q /images/graemlins/diamond.gifT /images/graemlins/diamond.gif.

Preflop: UTG limps, 2 folds, Hero raises, 3 folds, button 3-bets, SB folds, BB calls, UTG calls, Hero calls.

Flop(12.5 sb): K /images/graemlins/diamond.gif9 /images/graemlins/spade.gif8 /images/graemlins/diamond.gif.
BB checks, UTG checks, Hero bets, button calls, BB raises, UTG folds, Hero 3-bets, button folds, BB caps, Hero calls.

Turn(10.2 bb): 7 /images/graemlins/club.gif.
BB bets, Hero calls.

River(12.2 bb): Q /images/graemlins/heart.gif.
BB bets, Hero ?

Okay so this is an exaggerated example on purpose.
The question is how big of a mistake is folding here for one more bet?
-Let's say you are good one in ten times.
-You are getting around 13:1(minus rake).
-You have no idea whether you are good/no solid read.

The point of my post is that you would need to call this every time would you not? If you call 60% of the time and fold 40% of the time there is a chance you are folding the one or two times you are good. Then the result is a very big mistake.

This is my understanding of not folding for one on the river at least. In a big pot if you are going to call for one more you need to do it and ingnore the results.

Any comments? Is my post a waste of space?

[jason_t]

[ QUOTE ]
[ QUOTE ]
[ QUOTE ]
[ QUOTE ]
To add on to your post einbert, some of the major factors that make flop and turn decisions much more difficult to quantify stems from the fact that LHE IS a game of incomplete information and on the flop and turn we have less information at our disposal to make correct choices. Furthermore a decision can be swayed one way by the lure of things like implied odds to make up for an apparent -EV situation.

When people say things like preflop doesn't matter all that much and other such stuff, I like to think about it in terms of error bars around some true expectation and as the hand progresses these error bars start to shrink with the information that becomes apparent to us. Basically, all streets have important decisions in terms of your expectation and as you improve as a player your 'error bars' start to decrease and you have a better grasp on what the true expectation is for a given decision.

Basic stuff that is not always apparent and worth repeating...

[/ QUOTE ]

Nerd.

[/ QUOTE ]

When are you two getting married?

[/ QUOTE ]

Two weeks after you and Shant.

[/ QUOTE ]

I'm sorry to break the news but I had to call off my engagement with shant for someone else. He understands, of course, and I still /images/graemlins/heart.gif him.

[belloc]

Thanks for the link, bobby. This discussion makes me wonder exactly what Ed meant in that canonical Micro post way back when when he said that folding for one more bet in a big pot is a terrible mistake. He must have meant that habitually folding there is terrible. Is that how you read him?

[ESKiMO-SiCKNE5S]

id heard and understood this concept before, but it was well written, so nh

[QTip]

[ QUOTE ]
[ QUOTE ]
Thx damaniac, but this isn't the one i was thinking of. The OP in the other one is a hand from stheif.

[/ QUOTE ]
link (http://archiveserver.twoplustwo.com/showthreaded.php?Number=1813536)

Similar post from astroglide (http://archiveserver.twoplustwo.com/showflat.php?Number=1629302), which actually predated mine but I never saw that post until we had the same discussion again here (http://forumserver.twoplustwo.com/showflat.php?Number=3708995).

[/ QUOTE ]

There it is. Thanks.

[Gregatron]

[ QUOTE ]
[ QUOTE ]
Thx damaniac, but this isn't the one i was thinking of. The OP in the other one is a hand from stheif.

[/ QUOTE ]
link (http://archiveserver.twoplustwo.com/showthreaded.php?Number=1813536)

Similar post from astroglide (http://archiveserver.twoplustwo.com/showflat.php?Number=1629302), which actually predated mine but I never saw that post until we had the same discussion again here (http://forumserver.twoplustwo.com/showflat.php?Number=3708995).

[/ QUOTE ]
You know, I think Clarkmeister wrote something along these lines way back in the day iirc. And on it goes....

[W. Deranged]

[ QUOTE ]
[ QUOTE ]
[ QUOTE ]
Thx damaniac, but this isn't the one i was thinking of. The OP in the other one is a hand from stheif.

[/ QUOTE ]
link (http://archiveserver.twoplustwo.com/showthreaded.php?Number=1813536)

Similar post from astroglide (http://archiveserver.twoplustwo.com/showflat.php?Number=1629302), which actually predated mine but I never saw that post until we had the same discussion again here (http://forumserver.twoplustwo.com/showflat.php?Number=3708995).

[/ QUOTE ]
You know, I think Clarkmeister wrote something along these lines way back in the day iirc. And on it goes....

[/ QUOTE ]

They say that all modern philosophy is all just footnotes to Aristotle...

Similarly all recent 2+2 forum poker discussion is just footnotes to Clarkmeister... :P

[oreopimp]

[ QUOTE ]

2. Another example: Let's consider making a bad turn call. Let's say that in a 6 BB pot (including villain's bet to you), you find yourself with presumably 5 outs. You call even though you shouldn't. How big a mistake is this? The answer is far less than 1 bb . With five outs, you'll improve a little less than 1 in 9 times. For ease, let's say you'll improve exactly 1 in 9 times. Then you'll win 6 bets one time for every 8 times you lose, and so your expectation is -2/9 BB.

[/ QUOTE ]

Good post. Just wanted to mention something obvious here (or obvious to most Id assume). On turn calls like this the situation is even better that what u describe if u take into account implied odds.

Taking your example. Say you have a hand u know if it improves will be better than opponents, so u plan to raise if u improve. You figure opponent will always bet, and probably call your raise 50 percent of the time. so 1.5 bets in implied odds. You will also fold to a river bet, so u only invest one bet on the turn. (.11*7.5)+(.89*-1) = a .065 bb mistake. Which is pretty damn small.

Obviously the situation is a little worse if its a situation where u have no implied odds (board is very dangerous for your opponent) or opponent will bet, but not call your raise...or even if your hand improves it may not give u the best hand. But the mistakes are small, although if u keep making them over and over they will add up.

[ALurker]

It depends a lot on your read here, What % of the time do you think you are good here. If you are beaten %100 of the time then this is an obvious fold. If you think there is an off the wall chance that the BB will cap the flop with air then a call is best here.

You have to be good here somewhere around %9 to make this call correct. But I'm not that great with the math for these things. I'm guessing in this situation your good less than %5 so its a fold.

[James.]

bump for the noobs.