SNG endgame -- how to play 44

[M.B.E.]

[ QUOTE ]
IMO, the deciding factor in your decision is how often your opponent will fold to an all-in. If you think that he will do so at least 1/4 of the times, then go ahead and raise.

[/ QUOTE ]
Where did you get the 1/4 figure?

[ QUOTE ]
Given that he has you covered and is getting 2:1 on the call, this seems very unlikely however, so without any other information, I'd fold in this spot.

[/ QUOTE ]
Yes, he has me covered, but barely. If he calls and loses his stack will be 491, with the blinds at 300/600. Whereas if he folds he will have 3145, still the short stack but a reasonable shot at getting back into things. (With the 491, he'd need a miracle.)

I think the chance of his folding here is certainly more than 1/4, maybe as high as 1/2.

[M.B.E.]

In the actual hand, I reraised all in. My thinking was that the button could have made this raise with a variety of hands, including any ace and KT/KJ/KQ, maybe others. My stack now was big enough that if I raised him all in, folding would be a reasonable choice. Yes he had me covered, but not by much; if he called my allin and lost, his stack would be less than the big blind on the next hand.

Anyway, he thought a while, let his clock count down a bit, and finally called. He showed AKo. The board didn't help him so I won the hand, which made me the chip leader. I went on to finish second in the tournament. (I got a bad beat on the very last hand, otherwise I would probably have won, but that's another story. Anyway second place paid $1,080 for an $865 profit, nice.)

In retrospect, the fact that his hand was as good as AK made me think that maybe the potential distribution of his hands at the time was narrower than I thought. If that's right, then folding would have been correct because there'd be a greater chance of his having a pocket pair.

Also, the fact that he had to think about calling with AK makes me think I was right that there was a good chance he'd fold a lesser hand to my allin.

As I mentioned, in retrospect I'm happy with the result but I think folding would have been better, since I had enough chips to wait a little while, hoping someone else busts so I get the $360 in fourth-place money. On the other hand, if the prize structure had been 50-30-20 instead of 40-30-20-10, then raising all-in would clearly have been the correct play. But that's just intuitive, I'd love to see a rigorous calculation.

[Ignatius]

> Where did you get the 1/4 figure?
.
Feel free to pick your own number here. Each times he folds two overcards, you earn T1000. Each time he folds an overpair to your 44 (unlikely), you gain T4000. If you assume that he is 4 times more likely to fold overcards than and stick to the original assumtion that your all-EV is zero, then you end up with an EV of 1000*0.2+4000*.05 = 400 which is slightly more than 10% of your stack. For me, this is about the absolute minimum that I would risk my stack for in this situation.
.
> I think the chance of his folding here is certainly more than 1/4, maybe as high as 1/2.
.
In that case, your EV should be somewhere between T500 and T1000. Even with the flat payout, moving in should be clearly profitable then. (If the button would have been me, then your chance would have been 0 however, as the only hands I might open-raise less than all-in here are exactly AA and KK.)

[curmudgeon]

Your software is in dire need of reprogramming, and there is a distinct possibility of a bug in your hardware. [img]/images/graemlins/grin.gif[/img]

PDOH+PDIH*EOS=IT!
(the probability of me dancing on my opponents head +
the probability of something happening to me IF he does something)
times various estimates of my opponents stupidity = the answer I probably want to hear, but only if I did whatever I did do... on one million almost exactly equal situations. [img]/images/graemlins/confused.gif[/img] [img]/images/graemlins/crazy.gif[/img] [img]/images/graemlins/wink.gif[/img]



"Pay no attention to that man behind the curtain, I am the Great and Powerful..."