What are good situations to isolate for in a data base?

[kdmeteor]

Presumably a lot of players, even winning players, make similar mistakes when playing online cash. So there should be common things you can isolate for in a data base -- e.g., what happens when you open small pairs UTG, what happens when you defend suited connectors from the SB, what happens when you flat 2bets from middle position, etc. -- where many players will find that they do things that underperform a fold on average.

So I expected there to be lots of articles about this, something like, "here's the 50 MOST COMMON situations where people LOSE MONEY, fix them NOW to BOOST your winrate" -- but somehow I couldn't really find anything. So I'm instead I'm posting the question here: what are some situations where you've found leaks in your own game by isolating for them in your data base?

If it helps, leaks I've found in my data base so far are (a) flatting from the SB (flatting pairs makes profit, everything else is a disaster), (b) opening suited Aces from early positions, and (c) defending too wide from the BB (this one was non-obvious for me bc I play the format on GGpoker with the 0.5BB ante, so you have to defend more than in normal games, but the stats showed that I was taking it too far). I've also looked for various other things where I suspected leaks and didn't find them. But surely there are others, so -- any ideas?

 

[TeUnit]

I like to look for things the villans do too much or too little. Do they play too many hands? do they fold to too many cbets, do they defend their blinds to much or too little, etc. etc.

Also I think its good to look at how they play by position.

You need to make sure you have enough hands to make the stats significant.

 

[seiya1989]

Me gusta observar si los villanos hacen demasiado o muy poco. ¿Juegan demasiadas manos? ¿Se retiran ante demasiadas apuestas de continuación? ¿Defienden sus ciegas demasiado o muy poco?, etc.

También creo que es bueno ver cómo juegan por posición.

Debes asegurarte de tener suficientes manos para que las estadísticas sean significativas.

I noticed that you raise almost all the time with all the chips. Xd

 

[Station_Master]

Are you talking about your stats or the populations? Suited aces should be opened from most positions so perhaps they are just being played badly post flop?

Flatting from SB is in theory OK, but only v small raise sizes from EP, otherwise 3betting is preferred (or folding).

I think alot of these database reviews are reserved for pay-for coaching videos, which is why there are few articles on it.

 

[Sulfuros001]

Identifying and addressing common leaks in your poker game through database analysis is a valuable approach to improving overall profitability.

You need to make sure you have enough hands to make the stats significant.

 

[BetterThanAvgButNotByMuch]

Yeah, I think you can pay Jon Little or one of those "teaching sites" a ton of money a month to tell you info that's been around for like 20 years and the answer is in an old poker book or some crazy price for a solver where the answer will be for players you're probably not playing against or you just won't understand how the solver reached that answer.

But I would go back to the basics and plan your hand out before you actually play it and if your understanding of basic poker is ok then you can rationalize why or why you don't do something.

 

[kdmeteor]

Are you talking about your stats or the populations? Suited aces should be opened from most positions so perhaps they are just being played badly post flop?

I mean, I only have access to mine, and all I know is that if I open them, I lose more than folding (this is in a game that I beat overall). Maybe other people can make it work ¯\_(ツ)_/¯

One thing that it's harder to identify post flop mistakes. I wouldn't know what to look for, whereas it's easy to see "what happens on average when you do this". The other thing is the 6% rake; if a hand is only barely profitable in a tournament, you'll lose money from it in this format.

Have you ever checked for this in your data base?

I think alot of these database reviews are reserved for pay-for coaching videos, which is why there are few articles on it.

Yeah that would make sense

 

[kdmeteor]

You need to make sure you have enough hands to make the stats significant.

Yeah, as long as I'm only messing around with stats myself, I'm fine looking at cases where I have inconclusive samples, but I'm not gonna hire a coach until I'm positive I have enough hands to get a lot out of it.

It also depends on what you're looking at; the above example of raising suited Aces from early position is sth you'll get data on pretty quickly since it happens a lot. Postflop spots are gonna be much trickier and require more hands.