the conditions for stealing

[kdotsky]

In past posts it has been said that you don't want to steal from short stacks where you would have to call a push by them due to pot odds and it would put a nice dent in your stack. I find that later in tournaments there are almost always stacks like this to my left, which inhibits my stealing. There is almost never the "correct" conditions for stealing. It seems there's always bigstacks or shortstacks to my left, never stacks I have decently covered that I can fold to a re-raise.

My main question is: How much do you disregard these conditions? And in what scenarios?

Also one method I've tried is min-raising. For example I'm on the button with 20BB, the big blind and small blind have 9BB. I minraise because a) it's not correct for them to call off that much of their stack b) I can fold to a reraise.

Thoughts?

[Unparagoned]

I like this question and would like to see some replies from more experienced players. My thoughts are essentially that this is a pretty read-dependent situation. But obviously, if the big blind has 4x BB, it is not a great idea to raise him with T7o, regardless of position. By the same token, I tend to make my normal raise here any time I have a decent hand (so, I'm on button with 20xBB or so and the BB has 4xBB, I will probably raise QT+). Now, I do this in part because I have found that lots of these small stacks will fold and this may be the result of the level I play (5-30).

The min-raise thing I am just not sure about. I think that too often it just looks weak and like you want to take down the blinds with minimal risk. In order to use it effectively I think you would have to use it often and make sure you show down some quality hands with it, but then I could be way off-base here. So much of this all depends at the players at your table and what you think they're capable of doing if they suspect you have nothing.

[hurlyburly]

I like to put in a bet that has them covered with any Ace or 2 cards above 8, maybe 67s, 56s. At this stage of the tourny, you want to try to collect as many of these desperate stacks as you can, because they disappear fast. If you are under 10xBB, then it's even more important to loosen up against these stacks.

Big stacks are tougher, I just wait for a hand there, and push. By "hand", I mean any group 1-4 and some group 5.

Read all the Sossman and MLG posts, they are artistes when it comes to aggression.

[Mez]

Blind stealing is one of those things that separates the cash player and the tourney player. And doing this effectively is very important. I think most steal attempts are situation dependant. But I usually lean towards making a steal in your scenario, especially if the blinds are signficant.

I dislike min-raising because you're giving the BB odds to call (and I generally dislike any min-raise). If the BB calls, the pot is 5BB. If he checks, you're likely to make a continuation bet of at least 2.5BB whether you hit or not. I'd rather force them to pick a hand preflop. Of course, watch your opponents closely to see how loose/tight, passive/aggressive they're playing and make sure you know their stacks. All of that factors play into deciding if you're going to steal.

[kdotsky]

I agree that situations are very read-dependent. So say you have a decent idea of how often the small stacks will fold and what hands they will push with. Where does that get you? I decided to try combining the read with the math...

Say the blinds are 300/600, you have 12000 on the button, the small blind and big blind both have 5000 after posting. I'll just pay attention to the big blind to start.

Let x = chance the big blind folds
- Assume you raise to 1800
- Assume if he doesn't fold he pushes
- Assume if he pushes you're a 2:1 dog and you (correctly) call

So from the standpoint of before making your preflop action your EV is (we find x so EV>0):

x*900 + (1-x)(5900*.33 - 5600*.66) > 0

900 is what we get when we win the blinds, 5900 is what we get when we beat his all in (5000 + blinds), and 5600 is what we loose when he wins all in.

It comes out to x > 66%, so we break even if the BB pushes a third of the time - which seems like more than expected for almost anyone. Take into account the small blind may do the same thing - you need at least a 66% chance that both of them fold. x(SB)*x(BB) > .66. If they both have the same chance of folding, x > 81%, meaning you need to think either one of them will only push 20% of the time.

The paradox is that if they push more often, you have a better chance of winning since they're not pushing with as good of cards. So as x decreases your EV on the call increases.

I don't know how insightful/useful any of this is, I just thought it was sort of interesting.

[bmxreed36]

When the big blind still has more than 9XBB, I sometimes try the miniraise to see if it is the type of player who will fold to it (at lower buy-ins, it is very likely). If it works, I'll start mixing up the size of my raise. Mini-raise with a great hand and if it's gets shown down, they will keep respecting it. I mostly raise a little more with legitimate hands and although this looks very suspicious, if I have a deep stack, I may make a larger raise with trash and show if it's not called, this way I may get action with my bigger hands with the bigger raises and they will still be clueless as to what the size of my raise really means. Sometimes with a weak tight player, they will fold away much of their stack, especially near the bubble. If they play back, you'll have to slow down. Once they are below that level where you cant fold to a reraise, I just raise enough to put them all in or close to it with anything decent that they may call with a desperation hand. Big stacks in the blinds call for bigger raises which usually means simply better hands. I still may try a steal with a less than average hand once in a while so that if it gets called or reraised and you fold, the next time you raise with a big hand, he will be more likely to play back and give you some action.
Disclaimer: Far from a pro here, may be bad advice.