PT att to steal stat in MTTs

[lestic]

There's a stat called "attempted to steal" in Poker Tracker.

I'm thinking this could be very useful when considering restealing later in tourneys.
Do you use this stat? What kind of values should I be looking at here?

[drmcboy]

I'm interested to hear replies.

I have never bothered using it because it seems like you would need a HUGE sample size for it to mean anything. Not just to get a good distribution of hands themselves but you also need to have them with lots of different stack sizes, both villain and the folks around him, you're going to steal a lot more with a big stack and three middles to your left.

[baudib]

I use a different program, but yeah, I don't look at ATT to steal stats too often, because you rarely have a large enough sample for it to be useful in tournaments. Occasionally, you see instances where something really sticks out as unusual....Let's say you have someone who is a 20/8, and his attempt to steal from button is only like 8%...you know he usually has a decent hand.

Or if you have someone who steals from CO like 80%...even in small samples, you can occasionally find significant results.

[Trons]

I use relative stack sizes more frequently to determine my play then attempt to steal %. The main reason is because nothing can hurt you more then to play back at somebody with a 80% attempt and they wake up with a hand.

I generally ask myself, are they making this play because they half too (usually a much wider range). Can they afford my play back or are they trying to keep the blinds away from me? What are the stack sizes of the people between PFRer and me and is it possible he was making a move on one of them and if they folded, can I capitalize?

Even with those questions answered, I use those answers to put him on a range then I'll call/raise based on that information. If I think I'm ahead of his range, then here we go.

[baudib]

If someone is stealing 80% you don't worry about the 1 in 10 times they actually have a hand.

[chardrian]

ummm - I use this stat a shitton.

the stat is simply the percentage that someone openraises from the CO or button (and maybe the SB) when it is folded to them.

it does not take a phd in stats to realize that someone who overall is like a 20/15 but has a steal percentage of 40% is full of it purty damn often. So yeah I make them eat it pretty frequently as well.

[Sprayed]

I use this stat as well. This was from a $12 180 on the bubble. This guy had a 50% steal and 50% fold to 3bet. Thought that I was money on this play. Didn't notice if he had been 4xing when he raised my blinds previously. I guess I was just happy to have a hand to shove in his face. Good or bad?

PokerStars Game #27207563564: Tournament #156716382, $11+$1 Hold'em No Limit - Level XII (500/1000) - 2009/04/17 21:15:31 ET
Table '156716382 20' 9-max Seat #8 is the button
Seat 1: Sprayed (14046 in chips)
Seat 2: cirus Ger (10730 in chips)
Seat 3: gerrytallin3 (8154 in chips)
Seat 4: royallfllush (5960 in chips)
Seat 5: padgettgirl (10190 in chips)
Seat 7: FreddyLBS (19370 in chips)
Seat 8: BLUEDREWS (15355 in chips)
Seat 9: LuckyKuhn (23523 in chips)
Sprayed: posts the ante 100
cirus Ger: posts the ante 100
gerrytallin3: posts the ante 100
royallfllush: posts the ante 100
padgettgirl: posts the ante 100
FreddyLBS: posts the ante 100
BLUEDREWS: posts the ante 100
LuckyKuhn: posts the ante 100
LuckyKuhn: posts small blind 500
Sprayed: posts big blind 1000
*** HOLE CARDS ***
Dealt to Sprayed [9h Th]
cirus Ger: folds
gerrytallin3: folds
royallfllush: folds
padgettgirl: folds
FreddyLBS: folds
BLUEDREWS: raises 3000 to 4000
LuckyKuhn: folds
Sprayed: raises 9946 to 13946 and is all-in

Edit: sample is 45 hands.

[Muzzard]

Bad IMO.

You probably don't have a big enough sample to rely on these stats. I'm not sure what a good number of hands will be, but In 45man turbo's I doubt there will be enough hands, unless you've played in more than a couple together.

Also, I dont think he's 4x-ing and putting in 1/4 of his stack to fold to your shove.

I duno if you've noticed, but I've played in a fair few 45 mans with you.

[Muzzard]

Edit: sample is 45 hands.

Okay so that is 5 orbits on a 9 handed.

He can only be stealing in 15 of those hands maximum and it's doubtful that it folded to him everytime.

Let's say he was able to do it 1/3 of the time, when there was no raise before him - that's only 5 hands, then 50% of this is a nothing sample.

[Sprayed]

Edit: sample is 45 hands.

Okay so that is 5 orbits on a 9 handed.

He can only be stealing in 15 of those hands maximum and it's doubtful that it folded to him everytime.

Let's say he was able to do it 1/3 of the time, when there was no raise before him - that's only 5 hands, then 50% of this is a nothing sample.

I agree, however it was the bubble and I was under the impression that he was trying to abuse it. I normally play several at a time and his stealing stood out. I guess I was being reactionary. I use these stats more so when I play FRNL.

What's your name on stars? Sorry if I haven't said hi at the tables if you said something. I use TableNinja and it auto refreshes the info tab so that I know where I'm at in the tourney so I don't see the chat.

[fjuanl]

I wish I could remember some things from my stats class but it seems like you wouldn't need much of a sample to find the people who steal a ton. I mean, how likely is it someone who has raised 7/9 pots as steals has ONLY 55+ A9+ KJ+ in their range?

I use this on my HUD and kinda use it a lot when considering resteals and 3bets. I'm generally ok with using the stat to justify my play even if I only have ~10 steal possibilites if its someone who has stolen like 6+ times

[drmcboy]

it does not take a phd in stats to realize that someone who overall is like a 20/15 but has a steal percentage of 40%

how you would play differently if att to steal was 30 %? if someone is running 20/15 and they raise OTB you know their range is much wider than 20% with or without the stat

[chardrian]

I use stats extensively.

So yeah it might seem like a 20/15 should be stealing 30% but when an 11/7 is stealing 30% I'm gonna make them eat it whereas someone who is just looking at vpip/pfr will be much more hesitant.

[drmcboy]

based on what sample?

Can you define "eat it"?

[mcatdog]

based on what sample?

I somewhat agree; stats are helpful but I think the best plan is to just watch people play and see how well they're playing in general. If someone is playing well you can assume they will be stealing a lot on the button even if they've attempted to steal 0/8 so far. If they suck then they probably don't know they're supposed to raise the button.

Can you define "eat it"?

I think it means shoving your stack in their face?

[chardrian]

no matter what, poker is a game of limited info. I understand that a 5/5 over 40 hands might actually be raising 68s from UTG+1 this time and that a 40/30 might actually have AA from the button this time as well.

I will make stats based reads off of just 20 hands but that doesn't mean I'm surprised when a move blows up.

And yeah eat it means you raise and I make you eat it when i shove.

[lestic]

Okay.. my thoughts on sample size. Let's consider using reads for doing resteals. I think most people would say that if someone raises from say the CO to our BB 3 times in a row, he's stealing and it seems a good move to shove it in their face.
Now, that read is based on 3 orbits, maybe only 3 hands, he might have been folding everything else?
Umm, so what's the point, I don't really see much difference between that or a stat from 40-50 hands?
Okay, I'm exaggarating, I know good players read a lot more than just that, but I'm not that good

[baudib]

Just like everything else, it depends. A lot depends on your own image and also if villain is capable of folding or if he's going to snap off our shove with QJ.

[mcatdog]

no matter what, poker is a game of limited info.

I know that, I'm just saying that as you get better at poker you pick up on more info that other people might overlook. And I think you can make a rough estimate of what someone's steal % might be just based on how they play. You don't have go solely on their steal % so far.

[drmcboy]

And yeah eat it means you raise and I make you eat it when i shove.

ATC?

[chardrian]

against the right opp 4 sho.

making people get their pillows sopping wet while crying themselves to sleep because my J3o beats their top of their range hand is allows fun.

and yes - I think we are on the same page mcat.