Analyzing the Gigabet Dilemma

FakeKramer got flamed alot for this post, but I thought it prompted some really interesting issues. That thread had already been demolished by flames, so I started this new one.

In the hope of salvaging a learning experience from all the flaming going on here, I'm going to try to explain why I don't think this is a good example of "The Gigabet Dilemma". So a while ago Giga made a post about, using a pretty innovative theory of tournament equity, one could justify some pretty -cEV plays. According to this theory, a stack's size relative to others at the table and in the tournament may have more to do with its value than the percentage of the overall chips that it represents. For example, imagine the stacks at OP's table were as follows:

Button (t5066)
SB (t6100)
Hero (t22492)
UTG (t27652) (I subtracted 10K from this guy's stack- that's important)
UTG+1 (t24190)
MP1 (t6638)
MP2 (t24930) (Subtracted 15K here- again, important)
MP3 (t6105)
CO (t26779) (Subtracted 5K)

In this situation, there are two "clusters" of stacks. The shorties are floating around 5-6K, with 4-5 BB's. Hero is on the low end of a cluster around 22-27K, with roughly 20 BB's. Couple of important things to note about Hero's current stack in my hypothetical:

1. He can force any of the shorties to make a decision for all their chips while jeopardizing about 10% of his own stack.

2. He can lose to a shorty and STILL threaten the other shorties at the table without much risk to himself.

3. Although he can make a respectable dent, he cannot actually bust any of the larger stacks (this distinction is important late in the tourney, as outlasting a single player can translate into some extra $).

4. He cannot even threaten any of the larger stacks without putting his entire stack in jeopardy.

Now, let's look at the same call again. It costs Hero 5363 to call into a pot of 12,588, so he's getting better than 2:1. If he wins, his stack goes up to 35K. Now he can bust anyone at the table, force them to make a tough decision without risking too much of his own stack, and even survive a loss against one of the other large stacks (at which point he'd be around 6-8K, still competitive with the shorties).

If he loses, he's down to about 19K: still a comfortable lead over the shorties (though one of them will be about tied with him after tripling up here), and still a non-trivial threat to the larger stacks but without the ability to push them around.

The most important thing to recognize here is that those 5363 chips in my example ARE NOT DOING HERO VERY MUCH GOOD. Chips are tools, and in this instance there is not a lot he can do with this set of tools. The Gigabet Dilemma is a way to make these tools useful. If they aren't necessary to threaten anyone with Gambler's Ruin, what good are they doing? Well, they are essentially an expendable resource that can be used to take a (potentially -cEV) shot at acquiring some useful tools (ie a more threatening stack).

This may have been a bad hand to take a shot with anyway, but the actual stacks at Hero's table make it very wrong. Even if he gets lucky and picks up this pot, he is still not in a commanding position.

I'm not interested in arguing over whether my adjustments turn this particular situation into a good play, because 43o is indeed a very poor hand. But I am interested in exploring the thinking behind this kind of play and when it could be useful. Comments on that are warmly encouraged!

[FakeKramer]

I'm having a hard time finding the original "Gigabet Dilemma" post. If someone could provide a link to it, that would be great. I would love to read it.

Alright... don't let this post get in the way of Foucault's request for conversation. I'm eager to read the replies and learn a little more.

This is the Gigabet Dilemma post

I found some of the responses much more elucidating than Giga's initial, very convoluted explanation. It might make you feel better to see how much he got flamed for what I think is the most mind-opening thing I have ever read about poker. That's why I hate flamers so much.

[DonkeyChip]

Foucault beat me to it, it's from the STT forum. I agree that there are some great posts in that thread. But I often wonder if the fact that he won the hand makes the play look better. Or on the other hand, if the fact that he was so severely dominated pre-flop makes the play look worse.

People love to make 'fancy plays' and they also like to win with crappy cards. They also like to think that they have a deep understanding of poker that's over most other players heads (and in gigabet's case, that may be so). What better illustration of someone's superior poker skills is there than beating good cards with bad cards and 'advanced' plays? Don't get me wrong, I'm not saying gigabet isn't an excellent player or that the GD-hand was wrong. I'm just saying that it appears there exists incentive for some to believe the play was correct.

I agree with much of the logic used to make the play (and it was all 'new' to me), but I still feel that in this particular hand (i.e. the GD hand) that it was incorrect. But then again, I have trouble winning last-longer bets with friends much less TLB-challenges to the world. [img]/images/graemlins/wink.gif[/img]

[pooh74]

[ QUOTE ]
FakeKramer got flamed alot for this post, but I thought it prompted some really interesting issues. That thread had already been demolished by flames, so I started this new one.

In the hope of salvaging a learning experience from all the flaming going on here, I'm going to try to explain why I don't think this is a good example of "The Gigabet Dilemma". So a while ago Giga made a post about, using a pretty innovative theory of tournament equity, one could justify some pretty -cEV plays. According to this theory, a stack's size relative to others at the table and in the tournament may have more to do with its value than the percentage of the overall chips that it represents. For example, imagine the stacks at OP's table were as follows:

Button (t5066)
SB (t6100)
Hero (t22492)
UTG (t27652) (I subtracted 10K from this guy's stack- that's important)
UTG+1 (t24190)
MP1 (t6638)
MP2 (t24930) (Subtracted 15K here- again, important)
MP3 (t6105)
CO (t26779) (Subtracted 5K)

In this situation, there are two "clusters" of stacks. The shorties are floating around 5-6K, with 4-5 BB's. Hero is on the low end of a cluster around 22-27K, with roughly 20 BB's. Couple of important things to note about Hero's current stack in my hypothetical:

1. He can force any of the shorties to make a decision for all their chips while jeopardizing about 10% of his own stack.

2. He can lose to a shorty and STILL threaten the other shorties at the table without much risk to himself.

3. Although he can make a respectable dent, he cannot actually bust any of the larger stacks (this distinction is important late in the tourney, as outlasting a single player can translate into some extra $).

4. He cannot even threaten any of the larger stacks without putting his entire stack in jeopardy.

Now, let's look at the same call again. It costs Hero 5363 to call into a pot of 12,588, so he's getting better than 2:1. If he wins, his stack goes up to 35K. Now he can bust anyone at the table, force them to make a tough decision without risking too much of his own stack, and even survive a loss against one of the other large stacks (at which point he'd be around 6-8K, still competitive with the shorties).

If he loses, he's down to about 19K: still a comfortable lead over the shorties (though one of them will be about tied with him after tripling up here), and still a non-trivial threat to the larger stacks but without the ability to push them around.

The most important thing to recognize here is that those 5363 chips in my example ARE NOT DOING HERO VERY MUCH GOOD. Chips are tools, and in this instance there is not a lot he can do with this set of tools. The Gigabet Dilemma is a way to make these tools useful. If they aren't necessary to threaten anyone with Gambler's Ruin, what good are they doing? Well, they are essentially an expendable resource that can be used to take a (potentially -cEV) shot at acquiring some useful tools (ie a more threatening stack).

This may have been a bad hand to take a shot with anyway, but the actual stacks at Hero's table make it very wrong. Even if he gets lucky and picks up this pot, he is still not in a commanding position.

I'm not interested in arguing over whether my adjustments turn this particular situation into a good play, because 43o is indeed a very poor hand. But I am interested in exploring the thinking behind this kind of play and when it could be useful. Comments on that are warmly encouraged!

[/ QUOTE ]

Great post. But I think you overemphasize too much that just because Hero does not cover the other big stacks, that he is somehow much less threatening to them than when he does cover them.

If Villain has 27K and hero has 23K or 31K, either scenario will be devestating to villain and ALMOST equally intimidating.

[prana]

[ QUOTE ]
For example, imagine the stacks at OP's table were as follows:

Button (t5066)
SB (t6100)
Hero (t22492)
UTG (t27652) (I subtracted 10K from this guy's stack- that's important)
UTG+1 (t24190)
MP1 (t6638)
MP2 (t24930) (Subtracted 15K here- again, important)
MP3 (t6105)
CO (t26779) (Subtracted 5K)

In this situation, there are two "clusters" of stacks. The shorties are floating around 5-6K, with 4-5 BB's. Hero is on the low end of a cluster around 22-27K, with roughly 20 BB's. Couple of important things to note about Hero's current stack in my hypothetical:

1. He can force any of the shorties to make a decision for all their chips while jeopardizing about 10% of his own stack.

[/ QUOTE ]

10% of 22,492 is 2,249, in this example to bust the shorties he is commiting almost 22.5% from the shortest stack and over 27% from the other stacks in the 6,000+ range.

[DonT77]

I agree that taking a SLIGHTLY -cEV situation might be a +$EV play, but 43o doesn't figure to be SLIGHTLY -cEV against the range of hands that 2 all-ins figure to have.

Below I've assigned the 2 all-ins 2-unsuited high cards (AKo) and a middle pair (88). It looks like I might call with T9s and take slightly the worst of it, but with 43o and 67s - you figure to not be SLIGHTLY -cEV.

Note that a dominated hand as big as AQo is a 4:1 dog, so getting ~2:1 is still not in the category of SLIGHTLY -cEV.

from twodimes.net:

cards win %win lose %lose tie %tie EV
Ks Ac 533278 38.90 833466 60.80 4010 0.29 0.390
8d 8h 635958 46.39 730786 53.31 4010 0.29 0.465
4c 3d 197508 14.41 1169236 85.30 4010 0.29 0.145

cards win %win lose %lose tie %tie EV
Ks Ac 532802 38.87 835602 60.96 2350 0.17 0.389
8d 8h 593351 43.29 775053 56.54 2350 0.17 0.433
7c 6c 242251 17.67 1126153 82.16 2350 0.17 0.177

cards win %win lose %lose tie %tie EV
Ks Ac 515185 37.58 852664 62.20 2905 0.21 0.377
8d 8h 436439 31.84 931410 67.95 2905 0.21 0.319
Tc 9c 416225 30.36 951624 69.42 2905 0.21 0.304

cards win %win lose %lose tie %tie EV
Ks Ac 439420 32.06 907277 66.19 24057 1.76 0.329
8d 8h 638113 46.55 728874 53.17 3767 0.27 0.466
As Qd 269164 19.64 1077533 78.61 24057 1.76 0.205

[ QUOTE ]


10% of 22,492 is 2,249, in this example to bust the shorties he is commiting almost 22.5% from the shortest stack and over 27% from the other stacks in the 6,000+ range.

[/ QUOTE ]

I didn't say to bust the shorties, I said to threaten them. For 10% of his chips, he can force them either to push or fold. In other words, he commits 2K, they commit 6-7K, THEN he gets to make decision about whether to commit another 4-5K.

[prana]

[ QUOTE ]
[ QUOTE ]


10% of 22,492 is 2,249, in this example to bust the shorties he is commiting almost 22.5% from the shortest stack and over 27% from the other stacks in the 6,000+ range.

[/ QUOTE ]

I didn't say to bust the shorties, I said to threaten them. For 10% of his chips, he can force them either to push or fold. In other words, he commits 2K, they commit 6-7K, THEN he gets to make decision about whether to commit another 4-5K.

[/ QUOTE ]

Your argument might make sense if he wasn't CALLING 2 all ins. I've RAISED with hands just as bad before but this isn't that at all it's a call of 2 allins so he IS CALLING TO BUST THE SHORTIES and it is at or above 25% of his stack committed to do this.

[prana]

[ QUOTE ]
Now, let's look at the same call again. It costs Hero 5363 to call into a pot of 12,588, so he's getting better than 2:1. If he wins, his stack goes up to 35K. Now he can bust anyone at the table, force them to make a tough decision without risking too much of his own stack, and even survive a loss against one of the other large stacks (at which point he'd be around 6-8K, still competitive with the shorties).

[/ QUOTE ]

Seems to me like you were talking about the same thing as me after all. This call is almost right at 25% of his stack.