Blind Stealing Profitablity (inspired by HUSH post)

[Borodog]

Used my last 36k $2-$4 hands. My ASB is only 24.4%. [img]/images/graemlins/confused.gif[/img] Yuck. Need to work on that.

Raised 314 times netting $739.33. Called 20 times netting $63. Folded 952 times. Net $802.33 = 200.5 BB / 1286 Opp = 0.156 BB/Opp.

One thing that I'm noticing is, with a few exceptions, the really tight bunching of people's numbers around 0.15BB/Opp, regardless of their ASB. Am I imagining that, and if not, can someone explain it to me?

One good note is that my table selection seems good. I'm only getting a steal attempt every 28 hands. That's 3.57 Opp/100 HP, so 0.156 BB/Opp works out to . . . Holy [censored]. Almost 0.6 BB/100 HP. That's freaking AMAZING. That's 1/3 to 1/4 of my profit.

This is a damn good thread.

[crunchy1]

Thought you guys might be interested in a couple samples...
(These are all from Party Poker $2/4)

First sample is from my original post in this thread
(This is my play from about Jan 2005 - Aug 5, 2005):

63K hands, ASB% = 37%

Chance to Steal and Raised = $1751.94 on 684 times
Chance to Steal and Called = $110.50 on 23 times
Chance to Steal and Folded = $0.00 on 1160 times

Totals = ($1862.44 on 1867 times)/4 = .2494BB/opportunity


[/ QUOTE ]
Second sample is from my play since Aug 5, 2005. August was kind of a bad month coupled with me progressing through a "LAG-Phase". Interestingly enough however, my winrate for this period is actually .12BB/100 higher than in the first sample. This sample also occurs after reading and participating in several discussions regarding blind stealing here on 2+2.
(This is my play from Aug 5, 2005 - Oct 26, 2005):

31K Hands, ASB% = 43.06%

Chance to Steal and Raised = $402.09 on 507 times
Chance to Steal and Called = $17.50 on 6 times
Chance to Steal and Folded = $0.00 on 720 times

Totals = ($419.59 on 1233 times)/4 = .0851BB/opportunity


[/ QUOTE ]
The last sample is from my total play at $2/4:

95K hands, ASB% = 39.45%

Chance to Steal and Raised = $2118.03 on 1221 times
Chance to Steal and Called = $128 on 29 times
Chance to Steal and Folded = $0.00 on 1880 times

Totals = ($2146.03 on 3130 times)/4 = .1714BB/opportunity


[/ QUOTE ]

I think it's difficult, and likely fallible, to draw too many conclusions from these numbers - but at the same time I find them interesting. However, given the considerable dip in my winrate/steal attempt over the last 30K I think that some aspirations I had about getting my ASB% up in the 40's are probably incorrect. Looking at this data seems to suggest - at least for me personally - that my optimal ASB is somewhere in the 37-38% range. I think that I'd need to take a close look at the individual situations here. I'm guessing that my recent steal attempts have taken on a renegade approach which probably disregards reads and other information ANY time that it's folded to me in the CO or Button.

Enjoy the data! [img]/images/graemlins/smile.gif[/img]

[SeaEagle]

[ QUOTE ]
However, given the considerable dip in my winrate/steal attempt over the last 30K I think that some aspirations I had about getting my ASB% up in the 40's are probably incorrect.

[/ QUOTE ]
Why do you draw this conclusion? If I'm reading correctly, it looks like you have a lower ASB% in the second series, and your overall BB/hand is larger than what most people in this thread are showing.

I'd love to see your numbers with the big hands removed. On the surface, it just looks like you were running hot in the first sample ($2.50/raise) and cold in the 2nd ($.80/raise).

[mack848]

[ QUOTE ]
[ QUOTE ]
If I'm reading correctly, it looks like you have a lower ASB% in the second series

[/ QUOTE ]

I read it as 37% in the first group, 43% in the second.

[crunchy1]

[ QUOTE ]
[ QUOTE ]
However, given the considerable dip in my winrate/steal attempt over the last 30K I think that some aspirations I had about getting my ASB% up in the 40's are probably incorrect.

[/ QUOTE ]
Why do you draw this conclusion? If I'm reading correctly, it looks like you have a lower ASB% in the second series, and your overall BB/hand is larger than what most people in this thread are showing.

[/ QUOTE ]
I think you're reading wrong. From Jan-Aug this year my ASB% was 37% and my steal winrate was about .25BB/hand ($1/hand). From Aug-today my ASB% is 43% and my steal winrate has dropped to .08BB/hand ($.32/hand).

[ QUOTE ]
I'd love to see your numbers with the big hands removed. On the surface, it just looks like you were running hot in the first sample ($2.50/raise) and cold in the 2nd ($.80/raise).

[/ QUOTE ]
I ran the sample and excluded AA-QQ, AKs, and AKo:

95K hands, ASB% = 39.45%

Chance to Steal and Raised = $1603 on 1145 times
Chance to Steal and Called = $29 on 22 times
Chance to Steal and Folded = $0.00 on 1880 times

Totals = ($1632 on 3047 times)/4 = .1339BB/opportunity


[/ QUOTE ]

These numbers don't really surprise me - nor do I think they're particulary significant. I removed the hands that account for a large percentage of my profit in non-steal situations - I would expect that my BB/hand would drop a big as well. I didn't look - but I would suspect that the ratio of those hand's bb/hand vs. my total, average bb/hand would be pretty close if we looked at in in steal vs. non-steal situations.

[SeaEagle]

Yeah, I read backwards. Nonetheless, I can't believe that you lose so much with the extra 4% in the 2nd set that your win/raise drops from 2.50 to .80.

It looks to me like were just dealing with really small sample sizes, and we probably need to be careful of drawing conclusions. (For instance, my stats show that I shouldn't be stealing w/ AKo. I think that was my biggest single losing steal-hand.)

[crunchy1]

Sea- [ QUOTE ]
It looks to me like were just dealing with really small sample sizes, and we probably need to be careful of drawing conclusions. (For instance, my stats show that I shouldn't be stealing w/ AKo. I think that was my biggest single losing steal-hand.)

[/ QUOTE ]

Me- [ QUOTE ]
I think it's difficult, and likely fallible, to draw too many conclusions from these numbers - but at the same time I find them interesting. However, given the considerable dip in my winrate/steal attempt over the last 30K I think that some aspirations I had about getting my ASB% up in the 40's are probably incorrect.

[/ QUOTE ]

I should have said "possibly incorrect". I agree with you that sample size is a definite issue here. I also think that there must be some critical point where profitability will start to decline from "over-stealing". I don't have the data to determine what that number is - but I'm working on it!
[img]/images/graemlins/grin.gif[/img]

[MaxPower]

I have a bit of a problem with this analysis because a good number of these hands are not really "steals". This includes times when I had a premium hand and raised.

Anyway...

With 188920 hands at 15/30, my BB/opportunity is .16 and my ASB is 34.

I would expect it to be higher at this limit because of the 2/3 blind structure.

But I realy think this is misleading. We should really filter for certain hands.

[Borodog]

[ QUOTE ]
I would expect it to be higher at this limit because of the 2/3 blind structure.

[/ QUOTE ]

There is an offsetting effect between the tendency of the small blind to defend less in a 1-3 unit structure and the fact that when you successfully steal his blind it is only 2/3 as much as it would be in a 1-2 unit structure.

[MaxPower]

[ QUOTE ]
[ QUOTE ]
I would expect it to be higher at this limit because of the 2/3 blind structure.

[/ QUOTE ]

There is an offsetting effect between the tendency of the small blind to defend less in a 1-3 unit structure and the fact that when you successfully steal his blind it is only 2/3 as much as it would be in a 1-2 unit structure.

[/ QUOTE ]

Sure, but these stats we are looking at also include a lot of hands where we didn't really "Steal" the blinds. For instance in my sample there was a flop 77% of the time that I raised from a steal position.