SlackerInc
Sunday, November 9th, 2008, 2:04 AM
QUOTE (outsider13 @ Saturday, November 8th, 2008, 3:25 PM)

What are you trying to defend here? That shoving is right or my math is wrong. Your comments make no sense. Sorry that I have learned most recently from people who talk in BB rather than M. Nonetheless, it accomplishes the same thing.
Look, I may be wrong about shoving A8s here, though I find it hard to believe (and I'm not sure either side can prove their case, because it partly depends on speculation as to what range our opponents will call with, and that can never be set in stone). I don't know if Dan H. would shove this hand in this situation, though I'd note that he shoved J9s with a deeper stack (and after having been raised) on PAD (admittedly, he busted out first as a result).
I also question whether analyzing this in terms of average number of chips won is really the right metric. After all, you never win (or lose) that number of chips, and this is not a cash game where you can just deal with averages. You usually take down the blinds and antes, while sometimes you double up (or knock someone out), other times you bust out (or double someone up). As I think Highway Star was saying, there is a debate in the poker world as to whether (as he correctly surmises) Snyder is right that you have to just try to get a bigger stack and go deep, regardless of statistical odds, because it is only with a big stack that you can go deep in a tourney. So if you double up a little less often than you bust out, but when you do double up you can use those extra chips as a club to get more and more chips, a play could be profitable even if the simulations say otherwise.
Anyway, all those points of theory are highly arguable. But what I don't see as arguable is the idea that using BBs "accomplishes the same thing" as using M. That's like saying your GPS unit "accomplishes the same thing" if it stops being able to tell your longitude and can only report latitude!
I will grant, btw, that M is not perfect. An M of 10 because you have t1500 and the blinds are t50/100 with no antes is not exactly the same as an M of 10 because you have t3750 and the blinds are t50/100 with an ante of t25. In the latter case, the starting pot comes more from a diffuse contribution from antes and less from the blinds, so there is more incentive to play (for one thing, with the antes you can limp and perhaps see a flop for about 1/4 of an M, while by definition without antes limping costs 2/3 of an M) and yet less incentive for the BB in particular to defend. But it surely makes no sense to just say you have 15bbs in the first hand and 37bbs in the second, as if you are really over twice as deep!
Harrington also adjusts "effective M" depending on how many are at the table; this has its own problems as while it works in terms of the "how many hands you have until you're blinded out" measure, it doesn't address the ratio of risk/reward in terms of starting pot vs. your stack. But it should still be taken into account in some way.
So maybe we need some complex formula (using logarithms perhaps?), or maybe it just isn't possible to have a single numerical referent in a game that can or cannot have antes and where issues of structure (and Q, how well you are doing compared to others) have to be taken into account. But whatever M's deficiencies, I thnk it's clear that simply speaking in terms of BBs in general terms (for instance, "I wouldn't shove A8s until I was down to X number of BBs") is far worse.
BTW, am I really the only one who uses Tournament Indicator? Because it keeps track of everyone in terms of M.
P.S. This is an interesting discussion, thanks to all for the thought-provoking posts.