I don't know if this was addressed in the gigabet stack size thread...

[mythrilfox]

And I'm sure there have been hundreds of threads on it since, so forgive me if this was already addressed, but it's hard to search for things like this...

A number of different stack situations were outlined in the thread and it was speculated where the line would lie, but suppose the stacks are like so:

4x20,000
4x10,000
blinds 250/500

Where is the line now? There is no clear median. What would your stack have to be to consider taking a -cEV gamble?

Also, I assume that there is a second line above the first line, and it is above this line that you are able to make those +EV plays that you would have missed out on otherwise. So if your stack is too big and is already above the second line you should no longer take the -cEV plays? Likewise, if a play will get you closer to the line but not there, should you still take a -cEV gamble? i.e. suppose the line were 20k and you had 14k and could put a 3k stack allin but knew you were getting the worst of it. Losing 3k isn't that big of an issue if the extra 3k will compensate when you win, but if you are 6k below the line it will not, immediately at least. You'd have to be sure you could get the other 3k soon, before the line shifted again. (Of course, it could shift down as well, but in general it seems like it shifts upward)

[LearnedfromTV]

I *think* the idea with those stack sizes would be that taking the 49 end of a 49/51 v. another big stack if you are a big stack would be a good idea, because it would create a new plateau with you at 40K at at least double everyione else's stack. But that taking the same coin flip with a 10K stack is not a good idea because winning doesn't distinguish your stack. An extension of this is that the 10K stacks should be slightly more afraid of playing for their stack vs. the 20K's, because doubling up only puts them on par with several others (as opposed to a 20K/10K/10K/10K/10K/etc setup, where becoming the only 20K has some additional value).

Edit: there's a catch-22 here though in that if there's is 1 20K and 7 10K's, the advantage the 20K supposedly has disappears if all of the 10K's play as though getting to be the only 20K has an additional advantage (that is, if they are willing to gamble to double up and become him, and push around all the 10K stacks who are afraid to gamble...). I think Gigabet-theory may be eating its tail some of the time, but I don't claim to understand it well enough to be sure.

[mythrilfox]

I don't think a 49/51 is what's being referred to. If you're talking about things that close then switching a single suit of the cards will shift you from being a dog to a favorite, it's realistically impossible to know if you're the 49 or the 51. This is just a neutral cEV play for all intents and purposes.

Plus, why wouldn't you take it against the 10? You'd have 30k and you can threaten elimination to anyone at the table while still having 10k chips left over if you make a wrong move, and since 4 others also have 10k you're not that bad off.

And I thought you were supposed to avoid close gambles with other bigstacks later in the tournaments anyway. Most of the examples of the play deal with you having more chips than the guy you're going up against but having enough left over to still be a reasonable force if you lose.

[LearnedfromTV]

Ok, then the shy end of a 45/55...

I was saying the 10K stack shouldn't try to double up just to become a 20K because he doesn't gain a fundamental relative advantage and risks elimination. A 20K does gain a relative advantage by becoming 40K (while risking elimination). Maybe the theory would argue that a 20K taking slight -EV to become 30K would be the correct, since becoming 10K isn't a fundamental relative disadvantage (there are other 10K's), but becoming 30K is. I 'm just speculating.