Systemactic Self Evaluation

[NYplayer]

This is a question for people who play online a lot. i.e. astro, James282, Peter_Rus, URmeowed, Bicyclekick, 1800Gambler, Steve G., obi_one etc. (sorry if i left you out)

Do you have a systemactic way of evaluating your play?
For example, do you go through your databases once a month and look at which hands you are making the most money with or losing the most with position by position.
Or after a session look over every had that you lost a lot of money on?
Go over all of the AK hands you play after a session.
I guess i'm just asking if you have a method of evaluating your play on a periodic basis. kind of a monday morning quarterbacking.

Thanks

[skp]

I have a crude self study.

After every poker night, I spend about 10 minutes playing back the 20 hands that cost me the most money that night. Of course, some of these hands tell me nothing: my AA could have been outflopped by 55. C'est la vie.

But others might shed some light particularly because when playing back the hand, I can see the other guy's hand and can get into his head as to what he may have been thinking when he made whatever play he made. Occasionally, I will then make a note about that player's thought process. More often, I will just make a mental note of where I may have gone wrong on the play.

I think that it's also useful to review your play from the bb every now and then.

Another is to take hands like JJ,TT and see if you are holding on to them too long postflop.

But I too would be interested in hearing thoughts on a systematic review. Having said that, I have been doing all of this only because PT is still a new toy to me. In a couple of months, I might get too lazy to even invest the 10 minutes post session that I do now to try and spot leaks.

Good post and I hope there are some helpful replies.

[Peter_rus]

I look through my database every day after session. I don't look through hands. I examine dynamics of running certain experiments, i make currently, like limping first, limp-raising, SB-first in, marginal 3-betting PF, and blind-defence HU, as i change my game a lot last time and want to compare if these are good changes opposing to my earlier play.

[YoureToast]

This is a great thread....but I don't understand half of what you just said Peter -- "dynamics of certain experiments?" "SB-first in?" Not sure I know what these mean. Are you say you change your game frequently and want to see if your changes have been successful? If so, how exactly do you go about doing this.

I personally don't do much self evaluating other than stuff in my head, but thats only because I have not been able to figure out a convenient way to actually do it. I like skp's technique and I may try that, but I, too, am curious about other ideas.

[Peter_rus]

[ QUOTE ]
"SB-first in?"

[/ QUOTE ]

If all folded - you're 'first in on a __ position'. If you're on SB and all folded - you're SB-first in. You can raise first in or limp first in or just muck.

[ QUOTE ]
Are you say you change your game frequently and want to see if your changes have been successful?

[/ QUOTE ]

Not frequently. It concerned to my analysis. If i make some researches - they lead to change some spots of my game, if i don't make any and im lazy im playing my latest style.

[ QUOTE ]
If so, how exactly do you go about doing this.

[/ QUOTE ]

To analize i need to collect quite many hands, but if im running very bad with something in a short term i exclude it from my game and trying to find fault in my researches/game and also analyze if i was enough lucky to make some conclusions.

Let's say im running pretty bad after 150 times with limping low pairs. And i know that 150 hands isn't enough hands to analyze - then i count how much sets i've got on flop - if this number is lower than actual probability to catch one i proceed to collect hands, - if no, - then i look through hands to see how much money these sets give to me - if this cypher is lower than my sets brings me in average - i proceed collecting, if no - i exclude low pairs and don't limp with them.

[stoxtrader]

you can sort pokertracker by your biggest losing individual hands and then play them back. but you will probably just see a lot of suckouts.

I think statistically examining specific areas and comparing them to others is a good way, problem always will be sample size, and defining the sample in a productive manner....

another way is to get enough data on a limit to know how you do vs the limit avgs....

[pfkaok]

[ QUOTE ]
I examine dynamics of running certain experiments, i make currently, like limping first, limp-raising, SB-first in, marginal 3-betting PF, and blind-defence HU, as i change my game a lot last time and want to compare if these are good changes opposing to my earlier play.


[/ QUOTE ]

How many hands do you think you need in order for these experiments to have a lot of merit? I know you're alot better with access than most of us, so are you able to run tests that can also take into account how you're running. like the fold equity you're getting with certain plays? or say the actual equity you have in calling down vs. the range of hands that opponents would play in a certain way. B/c i know that sometimes over thousands of hands you can make seemingly good calldowns, but be unlucky and lose more than your share. or you can also be making changes that are hurting your true EV, but you've been getting lucky, so it looks like the changes are working.

[lil feller]

PokerTracker is a great tool for self-evaluating your online play, and I review it every session. I don't so much look for hands that cost me money, as I do hands that I remember. Often times the hands that were loosers that day just got drawn out on (and here I want to be sure that I lost the max...you get it?). I'm interested in value bets lost with my winners too, as well as pots lost by being too passive, or by making the wrong decision during the hand.

I also play a lot of B&M and I keep a journal. I remember interesting hands, and then write them out after I get home from my session. I write about how I played the hand, what I put other people on (and how accurate I was), and then I grade the hand and count up lost bets (either bets I should have collected and didn't, or bets I should have saved). I do this after every live session, and then I study my journal at the end of every week.

Hope this helps

lf

[Peter_rus]

I don't want to know exactly if my test is right. I just want to know it with some good probability. The most problems appear with close situations, when ev is so small that you need very large amount of hands to know if it's good to do this way or no. But in case you pick up new action where true ev is around 0.15-0.20bb/h (which i search for) your ev evaluation quickly enough will be in positive area after 500-700 samples. But if during first 200 hands you're below -0.25 -0.30bb/h than probability that you play right is very small and you need to reconsider.

[bicyclekick]

Nothing systematic. I just look over hands and different groups of hands from time to time to get a feel for things. Also look at position and how i'm fairing etc. I guess I don't really look all that in depth.

I spend more time thinking and trying to understand why the good players are doing what they're doing based on hands they post or situations that arrise.