Exploiting being CL: Short-handed high blinds Reshoves w/o Fold Equity

[Indiana]

actually this is a pretty good post, but the topic is difficult which is why everybody shys away from posts about overpushing. The math alone for that one problem took us a few hours to get right.

[Gramps]

Colin, the wheel called, it's been invented already. And although you articulate it's square existence very well, it's actually round. [img]/images/graemlins/smirk.gif[/img]

Don't neglect the future affect on calling ranges. If CO pushes and folds out the blinds, you have a bit of evening out of the short stacks, with close to a 3:1 chip edge on the nearest stack. That's a hugely pw3nable situation and you should be up against pretty tight calling ranges, giving lots of highly +$EV puhsing spots. People are human, people take notes, people don't like to be [censored] with (like the big stack coldcalling/overpushing with Q6o or J8o), if you play with any regularity, the "spite equity" will come back to majorly bite you in the ass. P3wning big blind poker is about building chip edges and controlling/manipulating calling ranges. Don't [censored] that up with a marginal (er...-$EV) play that loses you even more $$ in future hands.

I mean, if a player is constantly pushing into/through you on the bubble when you're the big stack and he figures you don't want to [censored] up your chip edge by calling loose (thus allowing him to profitably push a much wider range), then it's fine to make a loose -$EV call every now and then (even a -cev call that's quite -$EV) that is a bit of a $$ "write-off" on that hand, because it lets him know you're a looser caller than he's giving you credit for, and that he has to tighten up in the future (meaning more blinds get folded to you, and you get to take more +$EV open-pusing opportunities) - but you are losing $$ every time you do that, and there's an optimal % to do that for each player (and some that it may not make much of a difference). I think if you asked the biggest 50 winners in online SNGs over the past few years about what their overpushing range would be in the hand described, I'd be surprised if 5 were close to what you propose - most would be much, much tighter than the ranges suggested, for the reasons stated above - that and the fact that it's very arguable that a +cev play here is necessarily +$EV in a 50/30/20 payout structure (even assuming no affect on future variables such as calling ranges, image), but I'll leave that issue to the more hardcore math/stat guys.

Also, I disagree with your overcalling assumption - a lot of people are smart/experienced enough to understand the dynamics of a 3-way all-in on the bubble when they're 2nd/3 in chips - that in order for one of the blinds to be knocked out in 4th place, it has to go CO>CL>them which should happen probably < 15% of the time (1 of 6 possibilities, that will happen < 1/6 of the time since they will have a stronger hand on average than CO (and CL given the proposed range)). If you do this loose overpush play with any regularity, if the blinds have observed this in previous SNGs (people do take notes) and are intelligent players they will overcall with at least top 10% of their hands here, if not more. If CO>them>CL, well - they pretty much have the same stack they started with, no big deal. If CL>CO&them, they get 3rd place, they got the 20% equity jump from 4th to 3rd, which is a biggie in the 50/30/20 structure.

[bones]

Gramps,

Please write a book.

[Velocity]

[ QUOTE ]
Gramps,

Please write a book.

[/ QUOTE ]

I don't like people who quote for truth so I figure I better add something.
Whould it help if we got down on our knees and begged? How about if everyone got strippers to beg for them?

[Sparta45]

[ QUOTE ]
[ QUOTE ]
Gramps,

Please write a book.

[/ QUOTE ]

I don't like people who quote for truth so I figure I better add something.
Whould it help if we got down on our knees and begged? How about if everyone got strippers to beg for them?

[/ QUOTE ]

I have no shame....QFT

by the way Gramps, how did you come up with that 15% calculation?? I'm not doubting it's accuracy, I just don't see how you calculated it. thanks

[Slim Pickens]

OK, so I did a little math on your first scenario and got an unsurprising result. It's not terribly important exactly what ranges and such I used because that's small compared to the point I'm going to make.

blinds 200/400
CO (t1400)
Hero (t7000)
SB (t2550)
BB (t2550)

Preflop: <font color="red">CO raises all-in t1400</font>, Hero ???

Folding leaves Hero at 37.67% if it folds around, and I'll estimate that happens 80% of the time. The blinds should call more than that, but they typically don't. In any case, that can only raise your equity for folding. UTG getting called by the BB the other 20% of the time and having BB be a 60/40 favorite leaves your overall prize pool equity at 37.82%.

Assuming the blinds never call if you push, and you're a 55/45 underdog to CO's range, the push breaks about even. It hurts you when the blinds start calling, but that's not the point. We can argue that this is a breakeven spot at best, so there has to be some future considerations.

Tell me how much better it is to have the stacks this way if you push and lose:

Hero (t5600)
Button (t2350)
SB (t2150)
BB (t3400)

...than the way they are if you just fold.
blinds 200/400
Hero (t7000)
Button (t2350)
SB (t2150)
BB (t2000)

I prefer three short stacks, rather than two short stacks and one medium stack on my right. Again we can argue, but I'm coming to the same conclusion I came to in the Indiana thread. Under ideal circumstances, this will break even. If you're the least bit off, things get way worse very quickly.

[Gramps]

[ QUOTE ]
by the way Gramps, how did you come up with that 15% calculation?? I'm not doubting it's accuracy, I just don't see how you calculated it.

[/ QUOTE ]

You mean 1/6 doesn't equal 15%? Kind of cut through showing any work, but if the 3 hands are roughly on par, the 2nd stack should be eliminated 16.7% of the time. When he overcalls with say top 10% of hands, since CO is pushing very wide, and CL is now supposedly overpushing quite wide, his hand should be better than CL's on average, and quite a bit better than CO's on average. So...that 16.7% drops a bit, without going through calculations, it's probably safte to say it drops below 15%. Pokerstove is helpful in showing how often each player will win (and from that you can roughly figure out how often each player will get 2nd and 3rd), but I don't know of any software that will do a full 3-way analysis (showing order of finish and what % each of the 6 possiblities occurs given the hand ranges plugged in).