Too aggressive on the bubble?

[dfscott]

The hand before this, I was down to 2 BB after posting the BB and doubled up by going over the top of the 2nd stacks min-raise with KTo. Other than that, I had been very tight, so I felt like this was a decent steal opportunity (which obviously backfired), especially since I thought big stack would like to see it stay 4-handed for a while. Should I be tighter than this here and try and wait out button?

Party Poker No-Limit Hold'em Tourney, Big Blind is t300 (4 handed) converter (http://www.selachian.com/tools/bisonconverter/hhconverter.cgi)

Hero (t2010)
BB (t2735)
UTG (t1965)
Button (t1290)

Preflop: Hero is SB with T/images/graemlins/spade.gif, Q/images/graemlins/heart.gif.
<font color="#666666">2 folds</font>, Hero raises to t2010 (all-in), BB calls t1710.

Edit: you know, I just busted out again playing QTo. I think I've answered my own question...QT is trash.

[lorinda]

I thought big stack would like to see it stay 4-handed for a while.

This is the error in your reasoning.

Most players don't think this way, and even if he did, he's not big enough stack here anyway.

Lori

[lastchance]

SB can raise with any two here. I don't think this is too out of the ordinary.

[lorinda]

I don't think this is too out of the ordinary.

I don't think it's that much out of the ordinary either, (although personally I fold most times here) It's very player dependent, I just think the reasoning was incorrect /images/graemlins/tongue.gif

Lori

[eastbay]

[ QUOTE ]
The hand before this, I was down to 2 BB after posting the BB and doubled up by going over the top of the 2nd stacks min-raise with KTo. Other than that, I had been very tight, so I felt like this was a decent steal opportunity (which obviously backfired), especially since I thought big stack would like to see it stay 4-handed for a while. Should I be tighter than this here and try and wait out button?

Party Poker No-Limit Hold'em Tourney, Big Blind is t300 (4 handed) converter (http://www.selachian.com/tools/bisonconverter/hhconverter.cgi)

Hero (t2010)
BB (t2735)
UTG (t1965)
Button (t1290)

Preflop: Hero is SB with T/images/graemlins/spade.gif, Q/images/graemlins/heart.gif.
<font color="#666666">2 folds</font>, Hero raises to t2010 (all-in), BB calls t1710.

Edit: you know, I just busted out again playing QTo. I think I've answered my own question...QT is trash.

[/ QUOTE ]

Looking at four different calling assumptions, from tight to maniac:

http://rwa.homelinux.net/poker/tight-qto.PNG

A tight player who will call with TT+,AQ+ is a clear push.


http://rwa.homelinux.net/poker/average-qto.PNG

An average player who will call with 77+,AT+ is a closer push.


http://rwa.homelinux.net/poker/loose-qto.PNG

A looser player who will call with any pair, any ace, KQ, or KJs, it becomes very marginal and I fold, even though the analysis shows a tiny gain in equity. I don't think it's worth the risk.


http://rwa.homelinux.net/poker/maniac-qto.PNG

A maniac who adds any K to his arsenal will kill you if you push here.

You can see that this is either a clear push or a clear fold depending on what you expect to get called by. So as Lori says: definitely player dependent.

I fold here unless I have BB pegged as fairly tight.

eastbay

[dfscott]

[ QUOTE ]
I fold here unless I have BB pegged as fairly tight.

[/ QUOTE ]

Thanks.

But now I have to ask the bigger question: where did you get that cool program? Both Google and a forum search came up empty.

[eastbay]

[ QUOTE ]
[ QUOTE ]
I fold here unless I have BB pegged as fairly tight.

[/ QUOTE ]

Thanks.

But now I have to ask the bigger question: where did you get that cool program? Both Google and a forum search came up empty.

[/ QUOTE ]

I'm writing it. I am trying to release an initial version in the next few weeks.

eastbay

[dfscott]

[ QUOTE ]
I'm writing it. I am trying to release an initial version in the next few weeks.

[/ QUOTE ]

Cool -- let me know if you want a beta-tester. /images/graemlins/smile.gif

[rbilabronze]

Eastbay, that program looks amazing!!!

Are you considering adding a "R" (for random hand) button after the T, A, L, M to cover the situation where the BB may call with any two given pot odds? Or, I guess given the situation, M could be given the range of hands of "any-two" to cover that calculation.

[eastbay]

[ QUOTE ]
Eastbay, that program looks amazing!!!

Are you considering adding a "R" (for random hand) button after the T, A, L, M to cover the situation where the BB may call with any two given pot odds? Or, I guess given the situation, M could be given the range of hands of "any-two" to cover that calculation.

[/ QUOTE ]

You can specify a random hand by putting a "*" in the calling range box. I figured this was just as easy as putting a preset button for any two, and in addition, it's easier to understand than any two in the range notation. Note that the T,A,L,M buttons are just convenient presets, you can enter any range you like in the notation shown.

I also plan on making the T,A,L,M presets customizable, because an "average" calling standard in an $11 is probably not the same as "average" for $109, so you can tweak it to your likings.

I will be launching a website and forum in the next couple of days.

eastbay

[rbilabronze]

Thanks for the info. Does it also calculate scenarios where they may be more than 1 caller (such as pushing from the button or cut-off)?

[eastbay]

[ QUOTE ]
Thanks for the info. Does it also calculate scenarios where they may be more than 1 caller (such as pushing from the button or cut-off)?

[/ QUOTE ]

Pushing from any position is possible. It is assumed that action is two-way currently (if someone calls, the rest fold), as this covers the bulk of the cases without adding tons of complexity.

I have a plan for a v2 with a completely different interface that will be fully general as to the possible action.

eastbay

[rbilabronze]

Thank you. I had one more quick question if you don't mind.

When I first started making these calculations manually I was separately breaking outs ties. Doing it that way seemed to add a layer of complexity and the results weren't much different than just doing an ICM for the winning stack size (and using pot equity as the EV as opposed to the actual EV of a win -- and then a second calculation applying the EV of a tie against a second ICM percentage). Slightly different results but not meaningful enough to be worth the actual work involved.

I was trying to understand the underlying methodology of how your new program treats ties? Does the tool run a second ICM for relative stack-sizes if there is a tie (with corresponding EV for a tie)? Or does it use one ICM based on winning stack size (and use the "pot equity" percentage as EV)?

[rbilabronze]

Eastbay, how are you calculating the 0.381 for the $EV win in the program?

When I manually calculate it with the ICM Calculator I am getting .3739 (based on stack sizes if SB wins of 4020SB; 725BB; 1965UTG; 1290B). The only way I can get the 0.381 is if the stack sizes after the win were 4170SB; 575BB; 1965UTG; 1290B). It seems like at 0.381 the SB of 150 is somehow getting added into the SB's stack an extra time?

I am sure I'm missing something but wanted to point this out.

[eastbay]

[ QUOTE ]
Thank you. I had one more quick question if you don't mind.

When I first started making these calculations manually I was separately breaking outs ties. Doing it that way seemed to add a layer of complexity and the results weren't much different than just doing an ICM for the winning stack size (and using pot equity as the EV as opposed to the actual EV of a win -- and then a second calculation applying the EV of a tie against a second ICM percentage). Slightly different results but not meaningful enough to be worth the actual work involved.

I was trying to understand the underlying methodology of how your new program treats ties? Does the tool run a second ICM for relative stack-sizes if there is a tie (with corresponding EV for a tie)? Or does it use one ICM based on winning stack size (and use the "pot equity" percentage as EV)?

[/ QUOTE ]

Option #2. I agree with you that considering ties separately is in some sense more correct, but I also could not find a case where it seemed to make a difference big enough that it might change the conclusion.

If someone can present such an example, I would include the tie case separately.

eastbay

[eastbay]

[ QUOTE ]
Eastbay, how are you calculating the 0.381 for the $EV win in the program?


[/ QUOTE ]

There were actually a couple of small blind or ante-sized bugs in there for who picks up how many chips in the various cases. This is why the program is not released yet. /images/graemlins/wink.gif

Thanks for pointing that out.

eastbay