Assessing Errors

[barongreenback]

It’s become clear to me that hoping to get any info about your play from your results is pretty pointless but I would like to record a (rough) idea of how I played. The only way I can see of doing this is to quantify your errors.

Obviously there are lots of errors I don’t know about but I am going to start on the ones I can see (from studying the hands after play). It’s easy to get this wrong by saying stuff like, ‘I made a bad call on the river so that’s 1BB’. It's unlikely the play would never work out so it’ll be less than 1BB. I also think it’s more important that the error figure is accurate relative to other sessions than being accurate in an absolute sense.

I know the math of analyzing hands but I don’t want to spend ages on it and I’m only looking for an approximation. Has anyone got a system for doing this? How long do people spend on this sort of thing? Are there other alternatives?

Thanks in advance for any help,
BGB

[Dov]

Your best bet is to post hands that you felt had something wrong with them, and hands that you thought you played perfectly in the appropriate forum.

Then check other peoples analysis against your own.

This is the fastest way.

Good Luck

Dov

[johnnybeef]

one thing i like to do is assess every big pot i play in.

[mosquito]

The key is recognizing when you have made a mistake in
play. After that, the analysis is the easy part!!!

Take advantage of the forum, and spend time on your
own thinking about troublesom hands.

[barongreenback]

[ QUOTE ]
The key is recognizing when you have made a mistake in play.

[/ QUOTE ]
I agree. However, this is not just about improving my play but getting a sense of how far off my 'A' game I was in the session of play. This is important for a variety of reasons.

The main issue I think is the size of errors. We all say, 'I should have bet' or 'I made a bad call' but the error needs to be quantified in order to decide its importance. You shouldn't be spending equal time tackling a 0.2BB error as a 1BB error.

Some examples of very different errors (in limit holdem) are:
- Most preflop errors are small but serious because they come up frequently.
- Errors of tightness or passivity in big pots can be huge.
- Flop looseness may not be too big but can lead to compounded errors later on.
- River looseness can be a 'pure' error(no outs) but odds are usually good so tend to only be part of a BB.
- The turn. Big bets, compounding errors, missed opportunities to take the pot, poor odds for calls. It seems the turn is where the game is at.

Critcism of results based thinking is common but I don't see it going away until something is put in its place.

BGB

[mosquito]

You seem to have a fairly good handle. When posting
a hand, only post details up thru critical(?) decisions.
Not posting results tends to get better responses.

Also, knowing in your own mind that there are two classes
of errors: Those that cost a bet (or two) and those that
cost the pot. Once you can classify the error, it is easier
to assign a fractional bet or fractional pot 'error size'
to it.

Thirdly, and more important IMHO, is the realization that
your "A" game will include a certain number of errors.

To digress, I am an expert Bridge player. I have long set
my standard at 3 errors per session (27 hands) as being
the maximal acceptable level. TO BE SURE, the errors I make
now as opposed to several years ago are much different (though no less painful). The fact that you are improving
the quality of your error, as it were, is also an important
factor.

I believe you will be making errors for the rest of your life, in fact.

Good Luck.

[CMonkey]

I use results-based criteria to try to find errors myself. As you pointed out, it's a flawed method but I don't really see an alternative. I'm not going to analyze every single hand I play; I'd never have time for anything else.

The problem with your idea of quantifying errors is that they become more difficult to quantify the earlier you get from showdown. Figuring out the cost of raising on the river instead of calling is pretty easy; assigning a value to raising instead of calling in MP1 preflop is difficult since a change in action there makes the whole rest of the hand play out differently. I prefer to try to find errors I make often and eradicate those with the idea that even if each individual error is small, making them often is costly. Kind of hand waving idea, but the best I can do given that it's hard to quantify the magnitude of each error.

My method is generally to sort my hands by the most amount of money lost and assume that my biggest losing hands had errors in them. I do wind up with a decent number of false-positives this way (suckouts, having odds to draw to the nuts and not catching, etc...) but in many cases these situations are pretty easy to identify. Once I've identified a big losing hand where there's some question as to whether I played it well, I start at the flop looking up what stats I have on each player that saw the flop and make a rough estimate of their playing styles. I then simply step through each action I made. Fortunately, there are at most three actions at any given moment in limit holdem, so stepping through alternatives is not too painful even if it is time-consuming.

Preflop is more difficult to assess because there you have to take into account the texture of the table as a whole. Fine if you just got up from the table, but much more difficult if you're analyzing a hand you played several days ago. Unless something truly awful stands out (wait, I limped JTs UTG+1 and there were 5 folds behind me?), I generally just look at my starting cards and ask, "Ideally, how many other players did I want to see a flop with?" And then I look to see if my preflop actions achieved this result.

I’d be nice if we could just post our problem hands in the forums for scrutiny, but I think the moderators would start to get upset if everyone posted 20 hands a day.

[TStoneMBD]

if you are playing limit games, write a complex math problem to calculate the profitability of your play. i know that you said you dont want to do complex math, but these routine problems will have you thinking on new levels and you will become skilled at calculating math-related problems at the poker table at a faster rate. every dime you win or lose at poker is an expensive loss, especially when you get to higher levels. if you lose 1/4 a BB at 20/40, and dont learn from that lesson, causing you to make the same mistake 20 times, you will have lost $200 simply because you didnt feel like taking the time assess the problem. complex math is vital in understanding complex situations. many of my recent posts have had complicated math analysis of related hands, as ive been trying dilligently to perfect my technical analysis capabilities. attacking the math of poker will allow you to approach the game from a more logical standpoint, making you much wiser in the process.

[barongreenback]

[ QUOTE ]
My method is generally to sort my hands by the most amount of money lost and assume that my biggest losing hands had errors in them

[/ QUOTE ]
Any analysis is better than none and this is what I did to start with. There is a danger with just looking at hands you lost that only see money you could have saved by folding or not betting. This can make you weak tight.

I tended to be tight to start with and I found the most error prone hands were the ones where I held big hands and failed to maximise the money going into the pot because of slowplaying. Failing to value bet the river was also serious. You tend not to be punished for these mistakes because you're happy to have won and so don't notice them.

[ QUOTE ]
The problem with your idea of quantifying errors is that they become more difficult to quantify the earlier you get from showdown

[/ QUOTE ]
Yes, it takes a lot of experience to know what to do preflop but its easier to learn from others because there are fewer options. You can go a long way on a basic strategy. I'd say that once you've got this in place your biggest preflop leaks are those that distract you from your postflop play.

[ QUOTE ]
but I think the moderators would start to get upset if everyone posted 20 hands a day.

[/ QUOTE ]
Definately true. Also most would be ignored.

[barongreenback]

[ QUOTE ]
attacking the math of poker will allow you to approach the game from a more logical standpoint, making you much wiser in the process.

[/ QUOTE ]
It may be strange but I quite enjoy working on the theory for its own sake sometimes. Actually, in the past I've spent too long working on improving my win rate when I'd have been better taking advantage of the one I already had.

It comes down to finding a balance between improving my game and improving my bankroll (and enjoying myself). This is something I'll have to find myself. It's correct to play even if I'm not at my best but there comes a point where I'm better off taking a break. Finding this point efficiently and quickly is what I'm after. Of course, at higher limits, there'll be less room for error and this will be a luxury I don't have.

The bottom line is that I'll get out what I put in and I'll just have to get down to work if I want results. Everything will get quicker and easier with practice. Still, no harm in looking for shortcuts. I like to live life in the slipstream of others' hard work, wherever possible. [img]/images/graemlins/smile.gif[/img]

BG