Quote:
Originally Posted by madhusudana
Here is what I'm talking about. As you can see, BB barrels off with 2pair on a 4 flush board. Also, the small blind calls the turn, and would've called the river if the condition was >=tp. In fact, it happens for any action. How do I prevent this? I cannot use the layer system because I want to run the sim not knowing the flop. I cooked the flop just to get the right sample tun quicker to show you what I'm talking about.
How can I use the board filters for that?
I just did a quick calc and there's roughly 22k different flops out there.
With an additional ~2.5k for the turn and river to come off after that, that means there's roughly 50 million different ways for the board to turn out.
I think that it would be a bit too ambitious to try to define play on an unknown flop+turn+river.
Personally, I'd just stick to defining the flop play as best as I could and then just check down the turn.
The errors that would be involved in going any further than that would strongly outweigh the benefit of specifying the actions in more detail.
That's not to say that it's not possible to define all sorts of board structures with the "board texture" tab of the postflop conditions. But the tree that would have to be built here would have to be amazingly detailed, given all the different boards out there.
In fact, even if the flop were known here at Qh9h4h I would tend to be really careful in defining play any further than the turn. There's roughly >2k ways for the turn+river to come off and the board is already very very coordinated.
Quote:
Originally Posted by madhusudana
Just to clarify a few other things..
Lets say the I'm looking at BTN vs BB scenario, 1-2 NL. I happen to find out that a certain hand for the BB is +1$ EV (looking at the preflop all hands popup after the computations). Does that mean that my "strategy" ev in the BB is actually -1$ ? i.e 2$ blind minus 1$ profit? But it is still more profitable than folding and loosing the big blind, right?
If BB's EV is $1 and he's posted a $2 blind then he would indeed lose $1 in this hand, due to him not regaining the blind he posted. However, this is not to say that he shouldn't play the hand. He's got an EV of $1 after all. But
at the time of his decision the blind is already posted. Thus, the blind no longer belongs to BB, but to the pot. He's faced with a choice of folding, which will mean just giving up and "winning" $0, or he can "fight back" and win $1. If given the choice between giving up for $0 or playing back and winning $1, he should fight back.
Fwiw, losing money in the blinds happens to virtually every player. Playing to win back $1 after you have posted $2 sounds like a smart plan. Two hands later you'll be in a winning spot on the button ... which is why you post those blinds in the first place. Winning back $1 in the blinds will reduce your losses, allowing you to get to that winning spot more cheaply.
Quote:
Originally Posted by madhusudana
One more clarification question:
As in the image above, lets saw I'm looking at the very first EV stat of the SB when he limps in preflop. From what I understand that is the EV of his limping action weighted by all the EVs to the right of tree, correct?
Yes, all future actions are taken into account in the calculations.
For example, please see the pic below.
Button raises, bb calls, bb bets flop and Button folds.
Given that Button invests $14 preflop, only to fold the flop, his EV is -$14.
After all, every single time he makes that raise to $14, he ends up losing it.
Quote:
Originally Posted by madhusudana
So if this number is positive, it means it is profitable to limp in given all of the conditions I've made up pre and post flop?
Yes, correct.
Last edited by scylla; 01-15-2013 at 06:37 PM.