The flip side of dominating the bubble

[Lloyd]

[ QUOTE ]
The points are the same.

If a payout is something like 40/24/13/9/6/4/2/2 (ie steep) then you would care less about coming in 8th or 6th.. so you naturally tend to take chipEV decisions as is... with a view to taking close decisions and accumulating.

if the payout is something like 18/16/15/12/11/10/9/9 (ie flatter) then you tend not to want to take a gamble so much because not making the money is more of a major concern, and the ultimate prize for winning is not as significant.

I agree with Durron 100%

[/ QUOTE ]
Okay, I re-read your post for the 4th time and see that your on the same page. I don't know why but I've had a hard time comprehending your posts. And the few I've seen (and understood) it seems like you've got a good poker mind so I'm hoping to work through it.

[ekky]

I am from England, so I guess there could be a culture/language barrier there somewhere?

I always used to think I was quite literate with my expressives [img]/images/graemlins/mad.gif[/img]

[adanthar]

[ QUOTE ]
If you double up+blinds here you should be able to dominate this table for the rest of the bubble. Plus this tournament's payout is SO steep that there really is no real difference between cEV and $EV, even on the bubble.

[/ QUOTE ]

Well, first things first: on these bubbles, you can dominate the table whenever you cover everyone behind you. The number of your chips is kinda irrelevant as long as they can't call, so you don't *need* a double up to do this.

Second, the math is fairly simple to figure out: what you want, I believe, is to take your *current* $EV and examine what happens to it after you purposefully put yourself all in.

Looking back at today's tourney, I find that there are around 1,200,000 chips in play, so the average at the time was around 22,000. The prize pool was $33,000, which, divided by 55, is $600. Therefore, since a 22K stack is worth $600, a 14K stack now has an $EV of ~$380. (It's more or less linear up to this point, but if you go any lower the EV will precipitously drop unless your goal is to fold in.)

*However*, the minute the bubble bursts, all 50 remaining players will be guaranteed at least $180. Thus, a few hands from now, with $9,000 removed from the pool, an average stack of a little under 25,000 will be worth $480 and a 14K stack will be worth ~$270.

This is where my math skills give out and I can't figure out how to proceed. But what I am intuitively guessing is that you need $110 worth of cEV overlay - the gap between the two numbers - in order to push with no folding equity, not including the times when somebody else pushes behind you (so the actual number is higher.) Also note that when the lowest rung of the ladder is very low in relation to the overall payouts - such as in the Super - far less money is removed from the pool and the gap is much lower, meaning you should push more hands.

Somebody else can finish it from here, but I believe, *given the original read*, my gut instinct is correct and you should be giving serious thought to folding jacks in that spot. I can't do the altered read example on the fly, but I believe Axs/QJ are both much too loose and calling with 22 is really *really* terrible.

[ekky]

I dont like this, simply because it doesn't take into account skill considerations of good players (like yourself) with a larger stack.

If your chances of winning are X with a small stack.. I dont think doubling up will increase your chances of winning to 2X. I think it will increase it more then that.. so That has to come into consideration when laying out ChipEV scenarios...

so if you double up.. your chances will be 2X+Z.. where Z is the increased liklihood of you winning with a doubled stack.. and I should imagine on first glance that this Z figure in $ terms is greater then the cEV differential.

If its 4 handed and you all have 25% of the chips.. you might assume that you all have a 25% chance of winning.. but if you double up.. do you really think that having 50% of the chips only means you are winning the tourny 50% of the time? In the hands of a good player I contest this hotly.

[adanthar]

Well, my skill exactly on the bubble basically consists of pushing every hand into people that fold a lot, and the fun part about this particular spot is I can still do it without doubling (in fact I pushed 65o two hands earlier and somebody said they folded AJ. Cool!)

More importantly, with this particular payout structure, finishing 50'th and finishing 10'th are about the same, while finishing 55'th is very bad. If you take this to an extreme, like a sat structure payout where everyone wins the same amount, not folding aces is terrible. It follows that the closer you get away from a ladder and towards a flat payout, the more you should fold when someone that has you covered raises.

I can't prove it but I am reasonably sure this is the case unless your Z is just enormous.

[Lloyd]

[ QUOTE ]
Well, my skill exactly on the bubble basically consists of pushing every hand into people that fold a lot, and the fun part about this particular spot is I can still do it without doubling (in fact I pushed 65o two hands earlier and somebody said they folded AJ. Cool!)

More importantly, with this particular payout structure, finishing 50'th and finishing 10'th are about the same, while finishing 55'th is very bad. If you take this to an extreme, like a sat structure payout where everyone wins the same amount, not folding aces is terrible. It follows that the closer you get away from a ladder and towards a flat payout, the more you should fold when someone that has you covered raises.

I can't prove it but I am reasonably sure this is the case unless your Z is just enormous.

[/ QUOTE ]
I'm looking forward to delving into this later. It's so counter to some of the MTT wisdom that you could be on to something.

[ekky]

Right.. but if paying 50th and 10th are the same.. that must mean the leaps at the final table (especially in the final 3) are very steep.. so you should have no real care about going out on the bubble.

[ QUOTE ]
If you take this to an extreme, like a sat structure payout where everyone wins the same amount, not folding aces is terrible.

[/ QUOTE ]

Right.. but if you invent an uber-sat.. where 50th to 10th get a place in the stars rebuy... 9th gets a $215 entry to the $500k on stars, all the way up to first that gets an entry into the WPT main event.. you can see that you really dont give a monkeys about coming 50th--10th.. but you really want to get to the FT.. this is where the Chip value discrepancies are less distinct.

If the payout is uniformly flat from 50th to 1st.. you are correct that survival is more important.. but if they are flat from 50th to 10th..and then steep from thereon in.. it stands to reason that you should be taking any Chip edge however small, to get the chips... bubble or no bubble.

[adanthar]

Yeah, this is true. What happens in those cases is that each extra chip is worth a lot more $EV. You can tell that from looking at that Paradise million dollar freeroll.

However, with 55 left, there's still enough of a diffusion effect so that a 2x average stack isn't worth 2x$EV. It's worse on the bubble (when you don't have any FE and are up against a bigger stack - if this guy had 3 BB, pushing tens would not be a question because your $180 is not at nearly as much risk) because you're giving up that extra $. Conversely, at the final table, that same stack is worth much more than 2x average ("Gigablocks"), because the ladder is now super steep and taking flips for smaller amounts of it to eliminate an opponent is perfect play.

It's just something that seems instinctively correct.

[ekky]

[ QUOTE ]
However, with 55 left, there's still enough of a diffusion effect so that a 2x average stack isn't worth 2x$EV.

[/ QUOTE ]

probably not in terms of immediate EV.. but EV isn't linear in any case. Some poeple play worse with a bigger stack, so they are better off grinding up the places, and vice versa.

However, I think that in most pay off cases, the difference is going to be so small that its hardly worth thinking about.

Also something that came to mind... how much *EV* are you blinding off, folding your JJ, in order to make the money? it seems like if you take it to the absolute, you are going to be securing a LOT of medium place finishes, just by trying not to lose a tiny amount of immediate EV.. which really seems counter-intuitive.

It would be correct if the payout was so flat that accumulating chips would not increase your potential payout significantly.. but in most tournaments.. this is not the case.

[adanthar]

You can't take it too much past this, for sure. You have to open push tons of stuff and you also have to push tons of stuff when you have the other guy covered/close enough. Also, once the bubble bursts, the difference in payout levels (in this case) is like $16, 1/10 of a unit, so all of this goes out the window.

This basically only applies when a guy that has you covered is pushing/raising enough to commit you with an actual range of hands you can put him on *and* there's a comparatively huge pay jump coming up. If you get anything else out of that post you're probably taking it too far.

But it's funny, because it turns out that the short stacks are very often absolutely correct to fold a better hand. I set out thinking about this to get a sense of how incorrectly they are playing when I am bullying on the bubble, and it's not actually that bad (well, it is sometimes, but not when I 'only' push like top 50%).