EV of blind steal vs play for different hand ranks?

[gasoltub]

At what rank does your hole cards in Holdem have less than 0.75BB EV?
In other words, at what rank is a successful blind steal more profitable than calling/raising and playing the hand?

I know it's impossible to say exactly but there is perhaps a reasonable estimation.

Obviously having everyone fold after you raise AA UTG is depressing, but how about JTs from LP? ATo? Other hole cards?

I posted this in Poker Theory yesterday, but haven't got any replies so far so I'm giving it a shot here.

[Nick Royale]

[ QUOTE ]
At what rank does your hole cards in Holdem have less than 0.75BB EV?
In other words, at what rank is a successful blind steal more profitable than calling/raising and playing the hand?


[/ QUOTE ]
These question is not equivalent. Do you see why?

[gasoltub]

[ QUOTE ]
[ QUOTE ]
At what rank does your hole cards in Holdem have less than 0.75BB EV?
In other words, at what rank is a successful blind steal more profitable than calling/raising and playing the hand?


[/ QUOTE ]
These question is not equivalent. Do you see why?

[/ QUOTE ]

No.

But to define my question a bit more:
Everyone folds to you in CO or Button. There is a good chance that a raise will steal the blinds.
At what rank is a successful blind steal more profitable than calling and playing the hand?


Everything else equal, in a set position, with a set pair of hole cards, against the same opponents, with the same action before you, the expected value of your hand will approach a specific number of BB after an infinite amount of hands.
This could be more or less than 0.75BB.
AA expects to win more than 0.75 I assume, and 72o doesn't expect to win at all. In my PT database with ~70K hands, AA has been making 2.84BB/hand and T4o -0.13BB/hand.
So a successful blind steal with T4o would be great (although it probably shouldn't be advised) but a raise with AA where everyone folds is bad.

Is my question clearer?

[W. Deranged]

I think your question is a very interesting one but not all that strategically important. Here's why:

Given that you are choosing between EXACTLY a successful blind steal and calling and playing out the hand, then your question is relevant. BUT, you have to consider that raising will almost always have extra value (in addition to blind-steal-value) from those hands that get called and then you play out and win. Tactically speaking, whenever you have a good hand in late position, a raise is excellent because your opponents, particularly in the blinds, will not give you credit for the hand. Hence, when you raise with goods hands and do get played with, you will tend to make consistently more than you would by calling. Add this to the blind steal value and I can't think that calling will ever show better EV than raising.

[CallMeIshmael]

Most hands want the blinds. There arent a whole lot that dont.


FWIW... you cant really just use PT numbers here...


For example, your EV with AJo might not be 0.75BB/hand, but, it probably is when it is folded to you on the button.

(or... if that hand doesnt fit that criteria, there are certainly many that do)

[Nick Royale]

First of all, as you realise, when folded to you on the button the expected value for AA goes down by alot. This is because less players than usual will get into the pot and by doing so make a mistake (they would have folded if they knew what you had) If your average is 2.84 BB/100 your average from button after all have folded to you is probably less than less than 1.5 BB/100 (I'm guessing). But really, the problem is you'll get called more than 50% of the time when you raise because people will suspect a steal or just unwilling to let go of their blind. Because of that you need to raise any hand that have +EV on average, and even more hands than that. That's the way to get the most value out of your good hands.

[ QUOTE ]
At what rank does your hole cards in Holdem have less than 0.75BB EV?
In other words, at what rank is a successful blind steal more profitable than calling/raising and playing the hand?

[/ QUOTE ]
As you now see these questions aren't equivalent. You're assuming the blinds will fold 100% of the time but that's very far from the truth.

[gasoltub]

[ QUOTE ]
Most hands want the blinds. There arent a whole lot that dont.


FWIW... you cant really just use PT numbers here...


For example, your EV with AJo might not be 0.75BB/hand, but, it probably is when it is folded to you on the button.

(or... if that hand doesnt fit that criteria, there are certainly many that do)

[/ QUOTE ]

Even if I can't rely entirely on PT numbers they should at least give a hint, right?
When I check my 50K hands of 1/2 and 2/4 I see that the only hands that have won more than 0.75BB/hand are:
AA, KK, KQs QQ, AKs and JJ

So it seems right that a blind steal is worth more for most hands.

[gasoltub]

[ QUOTE ]
You're assuming the blinds will fold 100% of the time but that's very far from the truth.

[/ QUOTE ]

Ah, that's what I thought you where picking on [img]/images/graemlins/smile.gif[/img]
But my question and the discussion I was after really was about a 100% fold.

[HolyBejeesus]

You could set up a filter for when you are first to enter the pot from CO or button, and went to showdown. Unfortunately, this leaves out all games in which the blinds played but the hand ended before SD. Also, I tried doing this with my 20k hands and I didn't have close to enough data to get any worthwhile conclusion. If you have 200k+ hands it might be worth looking at.

[CallMeIshmael]

[ QUOTE ]
You could set up a filter for when you are first to enter the pot from CO or button, and went to showdown. Unfortunately, this leaves out all games in which the blinds played but the hand ended before SD.

[/ QUOTE ]

You dont need to set it to only include hands that were shown down.

Just filter to: chance to steal blinds and raise, and set the flop to 2+. So, this only includes when you have been called.

[ QUOTE ]
Also, I tried doing this with my 20k hands and I didn't have close to enough data to get any worthwhile conclusion. If you have 200k+ hands it might be worth looking at.

[/ QUOTE ]

I have that many, and, without even looking, Im quite confident it is nowhere near enough.

I have a pretty healthy attempted to steal blinds %, and nowhere near enough hands for any individual hand to be able to get an idea of its winrate.