I think my biggest leak is...

[Ed Miller]

I don't lay aces down enough when they are beaten...
I play medium pairs too aggressively after the flop...
I play medium pairs too often before the flop...
I misplay ace-king when I don't hit the flop...
I chase flushes and straights too often...
I don't chase often enough...

Have you said something like this about your game? If you have, you are wrong. In fact, not only are you wrong - that isn't your biggest leak - but I know what your biggest leak is!

Your biggest leak is that you have no idea how to tell a big leak from a small one.

I'm not trying to be cute; it's true. You guys can't tell the big mistakes from the small, the important concepts from the irrelevant. An inability to do this is a HUGE problem. People who can't decide what's important "fix" things that aren't broken, tinker with their games in ways that have almost no impact on their winrate, and sometimes overthink themselves into a tendency to make huge errors (i.e., folding too often in large pots, playing too passively in multiway pots, etc.).

How do I know that none of the "leaks" I mentioned at the top (and many more I didn't mention) aren't your biggest leak? Because they are all small leaks at worst! If any of these was your biggest leak, you would not be reading this post. You would be cleaning up in the Commerce $40-$80 to the tune of $100/hr or more.

So how do you decide what's important? Well, there's a logical process. A "leak" is a tendency to make a particular error systematically every time the opportunity to make it arises.

Folding every time you flopped a straight flush would be a leak (albeit a very bizarre one). But while it's clearly a ludicrous leak to have, it's NOT a big one. You almost never flop a straight flush, so you never get the chance to make your mistake!

The importance of a leak is determined by the product of two quantities:

1. The average magnitude of the error (in expectation) when you make it
2. The FREQUENCY with which you are presented the opportunity to make the error.

The magnitude of an error is defined by your AVERAGE LOSS every time you make the mistake. This is a place where people go wrong all the time when they evaluate mistakes:

First, make sure you are thinking about your average loss... you cannot conclude that you are chasing flushes too often simply because you missed the last four times and lost big pots. Think logically! How often do you expect to make the flush ON AVERAGE, how big is the pot when you win, what is your AVERAGE win or loss from playing the hand the way you are playing it?

Second, realize that there are certain types of errors that are almost never large in magnitude. Calling preflop when you should fold is an example. Say two people limp, and I have J6s on the button. Should I call? Probably not. But how big an error is it if I were to call?

It's small. Like tiny. A fraction of a fraction of a small bet. For instance, calling with J8s is probably correct. J7s might even be right. That's about where the line is. If J7s is just about break-even, how bad could J6s be?

Is J7s a monster compared to J6s? Do you fall off some magic profitability cliff between those two hands? Will calling with J6s ruin you? NO! That's absurd. If J7s is break-even, then J6s can fare no worse than -2% or so of a bet. Like a two cents at most in a $1-$2 game. Probably not even that much.

If you want to learn more about estimating the size of certain errors, you can read this thread from Matthew Hilger's forum.

Ok... so if playing J6s is a tiny error, how big could playing T4s be in the same spot? Well, it's a bigger error, but it's still small... maybe -5% of a bet. How about playing T4s from UTG? That surely must be a big error. Well, it's bigger, but still it's probably no bigger than -15% of a bet.

So if playing all these weak hands is not a big error, then how come every beginner text emphasizes playing tightly so much? Why do I have to fold 80% of my hands?

Frequency. Each loose call you make is a small error. But you get the OPPORTUNITY to make that error on every hand. If you make the error a hundred times a session, a small error becomes a big leak.

On the other hand, if you never get the chance to make the error (e.g., folding the flopped straight flushes... a HUGE error compared to the loose calls we examined... make sure you understand why it's a huge error), then even a big error does not translate to a big leak.

So a leak can be big only if

1. The error is somewhat costly
2. You get the opportunity to make it a lot

In the loose calls example, each individual error is somewhat small, but you make it SO OFTEN that it turns into a big leak.

Other examples of big leaks are folding too often in big pots (a HUGE error that you get to make with rare to occasional frequency), or playing too passively in multiway pots (a substantial error that you get to make frequently).

This morning, someone asked what to do with A2s after five people have folded from middle position.

Does it matter what you do?

Not really. Why? Because no matter what you choose to do, it could NEVER be a big leak. First, it's a borderline situation. If it weren't, he wouldn't be asking about it. Sometimes it's right to call, sometimes fold, and sometimes raise. It depends on the game.

But almost any time the correct answer "depends," they are all relatively close in value no matter the situation. Now getting the "depends situations" correct IN GENERAL distinguishes experts from average winners. That's because "depends situations" IN GENERAL occur relatively frequently. But any INDIVIDUAL "depends situation" doesn't occur very often.

Also, getting the "depends situations" right should be the priority of an already solidly-winning player. So many people ask trivial preflop questions while huge postflop leaks fester in their games.

So back to the A2s after five people fold question... if you make an error in that spot, it can't be a big one. Now how often does it come up? Well, the situation isn't applicable just to A2s... obviously A4s would put you in the same boat, as would 33 and possibly a hand like K7s. But 77 or A8s doesn't really put you in that spot. So the range of "weird" hands in middle position after five people fold is maybe ten or twelve hands: 66-22, A7s-A2s, K8s-K7s... maybe one or two more.

How often are you dealt one of these borderline hands exactly four or five seats from the button and have everyone fold to you? Once a week? Probably not even that often.

So he was asking about a situation where it doesn't really matter much what you do that arises only once every week or so.

Put things into perspective. That's why my book hammers on the postflop stuff so much. Because THAT is where people's big leaks are! That's where people are making LARGE mistakes OFTEN.

I gave a LOT of thought to which ideas I included and which ones I left out. Sometimes I intentionally left out advice that would be helpful because I felt it would DISTRACT people from the much more important stuff. So if I mention something like four times in the book, chances are it's important.

If I didn't mention something, or if I left it vague and didn't devote much space to it, chances are it's not important. I am not perfect, so perhaps I missed an important point. But at least THINK about how important what you are asking about is before you ask. Just doing that thinking will give you insight into the game that knowing what to do with A2s after five folds could NEVER give you.