Sort of a general thought about stealing blinds

[Evan]

Well, for starters, I used to suck at this a lot and I feel like it's pretty important.

For various reasons, I have played a few thousand hands of 3/6 in the last couple days. Anyway, I have been stealing a TON of blinds. I've been running sorta bad (big hands are barely winning/losing) but my win rate is still okay (no, this post is not about a win rate after a few thousand hands). This is definitely due in large part to how much I've made in blind steal situations. These guys defend their blinds soooooo bad that it is a crime not to steal them.

So I decided that I'd like to compare my rate to some other people. I looked up CallMeIshmael's post where he compiled a bunch of stats (and did a very nice job of it) and I found this:
[ QUOTE ]

Attempted to steal blinds:

Mean: 27.19
Standard Deviation: 4.70
Min: 19.29
Max: 33.33
A one-half standard deviation range about the mean: 24.84-29.54


[/ QUOTE ]
My att. to steal is 49 so far. I don't really have any great conclusion to draw from this, just that I think it's something that those of you who have been playing 3/6 for a while and are kinda stagnant should be thinking about (I wish I had sooner than I did).

Um, I guess that's all. Discuss.

[brazilio]

Stealing blinds that high I think is problematic. 1, because we can assume we're starting to semisteal from MP1 on, and stealing with half of our hands means we have quite a large chance of being beaten badly preflop, even though HU we might have an edge. 2., although our postflop play we can assume is superior, eventually I think it just begins to hurt us, ala the LAG period of skill advancement where no matter how good our postflop skills are, we're playing dogshit hands. My attempts to steal blinds have never been over 25-30 or so, and I think I'm rather aggressive when it comes to stealing as it is, although I definitely concur that postflop defense by most 3/6ers is pretty terrible.

[Evan]

[ QUOTE ]
we can assume we're starting to semisteal from MP1 on

[/ QUOTE ]
Yae, definitely, and I think that is a pretty good plan. Seriously, these people cannot wait to get out of your way once you raise. I feel like you're cheating yourself out of a lot of money by not taking advantage of this. Everyone seems to be complainging about 3/6 being difficult because the players are very tight, this is how you beat very tight players, you run over them.

[admiralfluff]

I think having a LAG image is great, and aggressive stealing is a great way to get one. That being said, I can't imagine having an ASB of 50%. does pt keep track of steraling winrate?

[jason_t]

[ QUOTE ]
I think having a LAG image is great, and aggressive stealing is a great way to get one. That being said, I can't imagine having an ASB of 50%. does pt keep track of steraling winrate?

[/ QUOTE ]

You can filter by chance to steal blinds & raised.

[pokerkai]

[ QUOTE ]
[ QUOTE ]
we can assume we're starting to semisteal from MP1 on

[/ QUOTE ]
Yae, definitely, and I think that is a pretty good plan. Seriously, these people cannot wait to get out of your way once you raise. I feel like you're cheating yourself out of a lot of money by not taking advantage of this. Everyone seems to be complainging about 3/6 being difficult because the players are very tight, this is how you beat very tight players, you run over them.

[/ QUOTE ]

I dropped down to play 3/6 the other day and I came to this very same conclusion.

[Evan]

You can filter for "Raised first in" and "Less than or equal to 1 off button"

Given those condtions my rate is .32 BB/hand (73 hands, so I didn't include that intentionally).

[Chairman Wood]

What do you think is ideal? What range of hands are you stealing with against a typical or unknown at 3/6? Same question for Evan.

[Evan]

[ QUOTE ]

You can filter by chance to steal blinds & raised.

[/ QUOTE ]

Hmm, totally missed that apparently. Given that my rate is .12 BB/hand in 92 hands. I thought that was essentially what I was doing with the other filter, apparently I was wrong. /images/graemlins/blush.gif

[SinCityGuy]

Sample sizes are obviously very important. In the long run, you want to be in the low to mid 30's on your blind steals as part of a balanced strategy. When you get to the upper 40's, you're going to get played back at more often when you have nothing. (Yes, even at 3/6 online, a lot of players notice this.)

If you're raising Ax, Kx, 75s, etc., this will get you in the mid 30's.

Of course, if the blinds are overly weak-tight, then it's +EV to raise with any two, as a flop bet will take down the pot a large percentage of the time. If they play back at you, it's easy to muck.

[Evan]

It sort of depends on how I'm running, but here's a list of approximations (partially stolen from a conversation w/ sthief09 the other day):

any ace, K5s/K7+, Q5s/Q7+, J6s/J7+, T7s/T8+, 97s/98, 76s+, any pair

[Evan]

[ QUOTE ]
you want to be in the low to mid 30's on your blind steals

[/ QUOTE ]
What? Why do I want to do that? That sounds a lot like something you just made up. When people are not observant and grossly underdefend, why do I want to only steal in the low-mid 30's?

[ QUOTE ]
When you get to the upper 40's, you're going to get played back at more often when you have nothing. (Yes, even at 3/6 online, a lot of players notice this.)

[/ QUOTE ]

In 25 3/6 sessions, which means I've been stealing against ~75 different blinds (probably more), one person seems to have caught on and 3 bet me more than I expected or played tougher against me postflop.


I'm completely open to arguments that I am over-stealing, but I am not open to arbitrary numbers that are presented on their own without evidence.

[admiralfluff]

[ QUOTE ]
one person seems to have caught on and 3 bet me more than I expected or played tougher against me postflop

[/ QUOTE ]

This number is far lower than my what I have experienced, proportionally, over ~500 steals. However, playing *tough* postflop means 2 different things to us.

I would expect you would be able to steal far more often than I, due to your bad-assedness, but 50% is more than I would expect, and likely more than your longterm rate would be.

when and why did you start playing 3/6?

[Evan]

[ QUOTE ]
playing *tough* postflop means 2 different things to us

[/ QUOTE ]
To me, it means something like not immediately folding or check/folding a lot. Instead they seem to lead a disproportionally high number of flops of check raise too often for my comfort.

[ QUOTE ]
I would expect you would be able to steal far more often than I, due to your bad-assedness

[/ QUOTE ]
This made me laugh. Can I no longer be known as short and from here on out be 'bad-ass'?

[ QUOTE ]
but 50% is more than I would expect, and likely more than your longterm rate would be.

[/ QUOTE ]
Maybe, I have no idea how long it should take something like this to converge. I'll leave that to people who are better at math than I to figure out.

[ QUOTE ]

when and why did you start playing 3/6?

[/ QUOTE ]
I have been running very shitty lately and not wanting to kill myself/hate playing was creeping up there on my list of priorities.

[RacersEdge]

Dumb question: how does PT define BS (that would be Blind Steal)? First in raise from button or CO?

[Evan]

[ QUOTE ]
First in raise from button or CO?

[/ QUOTE ]

[Catt]

[ QUOTE ]
Sample sizes are obviously very important. In the long run, you want to be in the low to mid 30's on your blind steals as part of a balanced strategy.

[/ QUOTE ]

I've posted several times on blind stealing in the past few days, and I'll just repeat here what I've posted elsewhere: at 3/6 (or in a 1/3 blind structure, especially with lots of weak-tighties such as are found at 3/6), I think anything less than 35% is leaving money on the table; I suspect that ~40% is pretty solid; and I wouldn't be surprised to find that >40% is better still. Even when played back at, we are good players post-flop, and we have position throughout the hand.

[PokerBob]

If I open raise in MP, am I stealing or do I just like my hand? At what point does an open raise become a steal? The Hi-Jack? The CO? Does anyone know when PT starts counting it as a steal?

I need to do more of this, I just need some clarification of what a "steal" is. Great thread.

[Evan]

PT defines a steal as an open raise from the CO or button. Read the whole tread, n00b! /images/graemlins/tongue.gif

[chesspain]

[ QUOTE ]
Does anyone know when PT starts counting it as a steal?


[/ QUOTE ]

The CO.

[PokerBob]

[ QUOTE ]
Read the whole tread, n00b! /images/graemlins/tongue.gif

[/ QUOTE ]

That would make too much sense. /images/graemlins/grin.gif

[brazilio]

I have to agree with you now, I didn't realize pt was only CO and beyond for open raises in their steal categories. I think 49 might be pushing it with steals from the CO, as even in the CO we have three left to act, and two with more of an incentive to call, but raising on the button with 40% of hands I can see as productive.

[einbert]

[ QUOTE ]
[ QUOTE ]
we can assume we're starting to semisteal from MP1 on

[/ QUOTE ]
Yae, definitely, and I think that is a pretty good plan. Seriously, these people cannot wait to get out of your way once you raise. I feel like you're cheating yourself out of a lot of money by not taking advantage of this. Everyone seems to be complainging about 3/6 being difficult because the players are very tight, this is how you beat very tight players, you run over them.

[/ QUOTE ]

Yes, this is totally dependent on game texture. At 3/6 my ASB would be way higher than it is at 5/10 6max.

That said, my inital reaction is that 49% is overdoing it. I believe at the Party 3/6 in particular 40%-45% might be close to optimal.

EDIT: If you're not semistealing from MP1 and up, you're missing out on a ton of money. For example, it's folded to you in MP1 and you have KJo, you should be raising every time regardless of who is left to act IMO.

[CallMeIshmael]

I would like to point out (and Chief and I discussed this like 2 days ago in a thread) that the blind stealing % in the stats post is the stat with which I have the biggest diagreement vs my own game.

I defintely think it should be higher than that average indicates.

[Josh.]

clearly you shouldn't be stealing at a 50% rate. something like 40-45 is probably best

and is the same Evan that I saw fold A2o on the button after it was folded around to him?

[Josh.]

[ QUOTE ]
It sort of depends on how I'm running, but here's a list of approximations (partially stolen from a conversation w/ sthief09 the other day):

any ace, K5s/K7+, Q5s/Q7+, J6s/J7+, T7s/T8+, 97s/98, 76s+, any pair

[/ QUOTE ]


I threw the hands I mentioned into a spreadsheet to see what % it was.

<font class="small">Code:</font><hr /><pre>
44+ 11 6 66
Ax 12 16 192
K5s+ 8 4 32
K7o+ 6 12 72
Q8s+ 4 4 16
Q8o+ 4 12 48
J7s+ 4 4 16
J9o+ 2 12 24
T8s+ 2 4 8
T9o 1 12 12
97s+ 2 4 8
98o+ 1 12 12
87s 1 4 4
76s 1 4 4

38.8%

</pre><hr />

[Evan]

I already credited you w/ my renaissance of blind stealing.

I don't know what makes 45 materially different from 49 in your mind, but okay. I said I have no idea when this stat convereges, so who knows what my 'true att. to steal' is. My only point is that when people defend so poorly, it should be high.

[Josh.]

in 6-max I steal at 35-40. you should steal more in full games because the players are generally tighter and adjust less. but you shouldn't steal too much or they're going to start assuming you're full of [censored] and playing back at you. that's why I like 40-45

[Fnord]

[ QUOTE ]

and is the same Evan that I saw fold A2o on the button after it was folded around to him?

[/ QUOTE ]

I will often fold A2o, A3o and A4o because I will often take down Axx boards anyway after a steal raise. A pair of 2s, 3s, or 4s is pretty slim and if I get action with those hands on an Axx board I'm probably not going to like it anyway.

[Dane S]

This will sound obvious to some but everyone should keep in mind that the rationale for stealing blinds isn't always to steal. When you openraise from the hijack, CO, or Button, it will generally be folded to the BB who will generally defend with a fairly crappy hand. He will then proceed to always be out of position and usually make plenty of postflop mistakes the rest of the hand. Needless to say this is a very profitable scenario regardless of what the cards happen to be and whether or not you can actually steal, and the chance to be on the right side of this duel is much of what you're paying for when you attempt to steal.

[Trix]

Or SB, nooblar /images/graemlins/tongue.gif

[bobbyi]

[ QUOTE ]
I looked up CallMeIshmael's post where he compiled a bunch of stats (and did a very nice job of it) and I found this:
[ QUOTE ]

Attempted to steal blinds:

Mean: 27.19
Standard Deviation: 4.70
Min: 19.29
Max: 33.33
A one-half standard deviation range about the mean: 24.84-29.54


[/ QUOTE ]

[/ QUOTE ]
I agree that those stats are ridiculuous. When I played 3/6, my asb was 36 and I consider myself to be a conservative blind stealer (if it matters, that was over 16k hands with a 3 bb/100 win rate).

[ QUOTE ]
My att. to steal is 49 so far.

[/ QUOTE ]
I do think that's too high. Evan, what you need to remember is that at 3/6, the rake is really brutal. Playing heads up is unreasonably expensive. So even if you are better than your opponent and have position, the rake turns a lot of profitable hands into unprofitable ones.

If you are interested in expanding your blind stealing range, you should look at some of Abdul's old stuff. He cared deeply about this topic and spent a lot of time trying to divine the game theoretically correct opening standards for the button. Even though he later tightened up a bit, he originally said "I need to run more simulations, but right now my best guess for the minimum hands needed to raise on the button is as follows: 22 A2s K2s 54s 53s 63s 92s A2 K4 87 97."

I'm sure if someone posted a hand here where they had raised with something like 92s, they would get flamed. But these hands were for a game with 2/3 blinds (bellagio 30/60) and with no rake (time charge instead), so they would not be directly applicable to an online 3/6 game anyway. To show how big a difference the rake makes, in his legendary preflop guide (http://poker.betfirms.com/holdempreflop-abdul.shtml) (note how his button raising standards there no longer include 92s), he says:
[ QUOTE ]
A rake seriously reduces the number of hands with which you can steal, as you will be paying a lot for a crapshoot against the big blind. With a Draconian rake, like where the big blind gets dropped once the flop comes, you would need about JJ or better to open on the button! Even with a modest rake, JT and 76s should be folded.

[/ QUOTE ]

[Trix]

I think you are stealing too much from the SB if you have a number this high, considering the 1/3 structure, unless your opponents are folding their blinds too much or playing fit or fold after the flop.

[Evan]

[ QUOTE ]
I think you are stealing too much from the SB

[/ QUOTE ]
According to the mouse-over definition that is displayed on the bottom of the PT window, SB raises are nto accoutned for in Att. to Steal.

[ QUOTE ]
unless your opponents are folding their blinds too much or playing fit or fold after the flop.

[/ QUOTE ]
Welcome to Party 3/6.

[Trix]

[ QUOTE ]
According to the mouse-over definition that is displayed on the bottom of the PT window, SB raises are nto accoutned for in Att. to Steal.

[/ QUOTE ]

[ QUOTE ]
It does include stealing from the SB. I changed it after I wrote the help file and forgot to update the help file to include this. So if everyone folds to the SB and he raises, that's considered a steal attempt.

Thanks for pointing this out.

Pat

[/ QUOTE ]

[Nick C]

I was thinking I was trying to steal too much and was planning to cut back. But now that I've read this thread, I'm not so sure.

I mostly play during the evening, night, and late-night, when things aren't (this is my impression, anyway) always so weak-tight.

So, again, I'm not sure. But my number for 3/6 is around 40.

[Evan]

Your floppy-eared avatar belies your wisdom. I stand corrected.

[NT!]

[ QUOTE ]
clearly you shouldn't be stealing at a 50% rate. something like 40-45 is probably best

[/ QUOTE ]

41.7% here

I think I steal goot.

NT

[bobbyi]

Evan, I think you will also find this interesting:
http://archiveserver.twoplustwo.com/showflat.php?Cat=&amp;Board=headsup&amp;Number=1162671

[chief444]

49 seems high. But I think it may be more profitable than low 20's. I agree that the average is too low.

[meep_42]

[ QUOTE ]
Evan, I think you will also find this interesting:
http://archiveserver.twoplustwo.com/showflat.php?Cat=&amp;Board=headsup&amp;Number=1162671

[/ QUOTE ]

I'm bumping because this got mentioned yesterday ... and this linked post is very good.

-d

[crunchy1]

I'm surprised that nothing has really been mentioned regarding game selection here.

Obviously if you are playing in games this tight then your stealing % must be theoretically correct to be higher than if you were playing in loose games. How do you think that this increased stealing % concept affects those of us who are really taking the time to select games that are loose/juicy where most of our steals are going to be played past the flop. (Of course I'd need you to quantify this against how you select games)

[mdob]

[ QUOTE ]
I have no idea how long it should take something like this to converge.

[/ QUOTE ]

Using a fairly liberal definition of converge, I'd say about 150 hands where you had an opportunity to steal. This gives you an 85% chance that your 'true' steal rate is within 5% of your observed win rate. For example, if you tried to steal 60 times in 150 opportunites (40%), you could say there's an 85% chance that over infinite hands you'd steal between 35% and 45% of the time.

If you want a 90% chance you're within 3%, you'd need almost 700 blind steal opportunities. (The number of opportunities you need is far more dependant on the '5% number' than the '85% number,' if you're interested.)

Evan said he stole 49% of the time over 73 hands. That's about an 85% chance his 'true' steal percentage is in the low 40s to mid 50s.

The above, of course, is based solely on statistics. If you want to know your exact true steal percentage and can say I steal with any ace, any suited king, etc., just do the math.

By the way, can anyone reccommend a good resource on blind defense? Any book that deals with this?

[Evan]

When I wrote this post it was semi-specifically in reference to Party 3/6, which I found to have more weak-tight games than super juicy ones.

I didn't do any game selection, I went on the waiting lists for first table with 7+ players. Accross a 10k hand sample, I found 2 blinds that I altered this stealing method against. Obivously you have to put some thought into what you're doing, but I do believe that this is a good guideline to work from.

[meep_42]

[ QUOTE ]
By the way, can anyone reccommend a good resource on blind defense? Any book that deals with this?

[/ QUOTE ]

I could really use a blind defense primer, too.

-d

[bobbyi]

[ QUOTE ]
By the way, can anyone reccommend a good resource on blind defense?

[/ QUOTE ]
The HUSH forum.

[SackUp]

[ QUOTE ]
Evan, I think you will also find this interesting:
http://archiveserver.twoplustwo.com/showflat.php?Cat=&amp;Board=headsup&amp;Number=1162671

[/ QUOTE ]

wow awesome. this is almost exactly what I have bumped up my blind stealing hands to.

I've also been increasing my semi-steals from mp2 and mp3 when folded to me or a limper who is 45+ vpip.

I would love to see some stats run on the semi steal from earlier positions.

Also how many of the "steal" hands from CO and BT do you still raise to isolate a limper? And how many are you now mucking?

[The DaveR]

Here was a standard blind steal scenario the last time I played 3/6.

I raise 2 cards in the CO. BB calls. Flop is xxx. BB bets, I raise, BB folds.

[The DaveR]

[ QUOTE ]
When I wrote this post it was semi-specifically in reference to Party 3/6, which I found to have more weak-tight games than super juicy ones.

[/ QUOTE ]

Emphasis mine.

[BlackRain]

Great thread. Just wondering what people thought about ASB stats at lower levels, particularly 1/2 and 2/4.

[Hobbs.]

bump for the nits the don't want to raise J9o OTB!

[five4suited]

[ QUOTE ]
[ QUOTE ]
When I wrote this post it was semi-specifically in reference to Party 3/6, which I found to have more weak-tight games than super juicy ones.

[/ QUOTE ]

Emphasis mine.

[/ QUOTE ]

You chose poorly.