|
|
#1
06-13-2005, 03:14 AM
|
|||
|
|||
Spark Notes on Gigabet\'s Post
I sometimes make -EV plays if winning would upgrade my stack category from medium/surviving to large/dominating.
I sometimes pass up +EV opportunities if losing would downgrade my stack category, but winning would not upgrade it. 
|
|
#6
06-13-2005, 12:35 PM
|
|||
|
|||
|
Re: Spark Notes on Gigabet\'s Post
I'm guessing he probably means $ EV.
I'm pretty sure I know what you're getting at, and it's something that I've noticed before. We really need different language to clarify whether we are talking about the EV of a given move or a given situation. The EV of a situation is going to be the EV of the optimal move in that situation. Passing on one +EV move in a given situation does not mean that we are making a play with lower +EV, necessarily. Thus, there may be circumstances where every move that can be made is + EV, but one is the most + EV. In particular, to paraphrase ilya, there may be situations where one passes on a +EV opportunity to put your chips in the middle because the $ EV gained by folding and allowing the situation to play out in a different manner might be higher.
|
|
#7
06-13-2005, 12:58 PM
|
|||
|
|||
|
Re: Spark Notes on Gigabet\'s Post
[ QUOTE ]
I'm guessing he probably means $ EV. [/ QUOTE ] What do you mean by $EV? There's actual $EV and there's $EV as estimated by some model. If he means actual $EV then it's nonsense by definition, and I think a disturbing number of people seem confused about this. If some chip position gives you some strategic advantages, that's already accounted for in your actual $EV. It's complete nonsense to say you'd pass on some $EV now for more $EV later. You'd never pass on $EV, ever, because it would be throwing money away, by definition. Then there's $EV as estimated by some model, and the usual fave is ICM. I think talking about "passing on $EV to gain $EV later" in this sense isn't necessarily nonsense, but it's kind of a backwards approach to how you should think about $EV. You should be trying to figure out what factors give an actual $EV that differ from $EV_icm, and talk about what your actual $EV is, rather than saying "you should give up $EV_icm here." [ QUOTE ] I'm pretty sure I know what you're getting at, and it's something that I've noticed before. We really need different language to clarify whether we are talking about the EV of a given move or a given situation. The EV of a situation is going to be the EV of the optimal move in that situation. Passing on one +EV move in a given situation does not mean that we are making a play with lower +EV, necessarily. Thus, there may be circumstances where every move that can be made is + EV, but one is the most + EV. In particular, to paraphrase ilya, there may be situations where one passes on a +EV opportunity to put your chips in the middle because the $ EV gained by folding and allowing the situation to play out in a different manner might be higher. [/ QUOTE ] That's a good point as well and another thing that people often talk about in a confused way. eastbay
|
|
#9
06-13-2005, 01:46 PM
|
|||
|
|||
|
Re: Spark Notes on Gigabet\'s Post
[ QUOTE ]
What do you mean by $EV? There's actual $EV and there's $EV as estimated by some model. If he means actual $EV then it's nonsense by definition, and I think a disturbing number of people seem confused about this. ... Then there's $EV as estimated by some model, and the usual fave is ICM. [/ QUOTE ] I think giga's post would've been better understood if the Q3 hand never existed. Everyone thinks he's talking about when to make -EV plays, and comparing that hand to curtains' J4s hand and other hands that have nothing to do what he's trying to say. eastbay is right. giga is trying to describe his model. I've been thinking of it as the Gigabet Block Model, or GBM. I'll never completely understand his model, but I'm pretty sure this was the point of his post. He was trying to describe how to calculate EV using the GBM. Much like ICM, the value of a chip is not constant in the GBM. In ICM, each chip is worth more to a shortstack than to a big stack. Chips lose value as your stack increases. Therefore, 50/50 coin flips are always -ICM. In GBM, chips also change value, but in a less linear and more complex way. A coin flip could easily be +GBM if winning gets you another block, and losing doesn't cost you a block. I would imagine that EV results for ICM and GBM are pretty close on most hands. However, giga is using the Q3 example as a hand where they are not. It doesn't matter why, just accept that this hand is worth more according to GBM. I'm guessing that the Q3 hand is still -GBM, but the play is not nearly as negative as it appears to ICM and giga's opponents. He is advertising with a play that appears insane, but is really only marginally -GBM. He is getting a discount on this advertising. He is giving up EV, but nearly as much as his opponents think. This is just 1 example of exploiting the spread between GBM and ICM. Most of us will never play a $1000 SnG, so the advertising aspect of the Q3 is not important. What is important is that chips change value in different ways than the ICM uses. This is valuable to understand at any limit. ilya's summary might cover some of what giga was trying to say, but I think its dangerous to try to simplify his ideas. Just look what happened with the "PVS" despite being a much simpler and less important concept. If you think giga's point was: "you should take risks to get a big stack", you are really missing his point and are better off forgetting than he ever posted this. Hopefully, giga will come back and answer the questions that eastbay posed in his response in the original thread. Until then, I will try not to guess what he means by a "block" or "line" and add to the confusion.
|
|
#10
06-13-2005, 01:47 PM
|
|||
|
|||
|
Re: Spark Notes on Gigabet\'s Post
Ilya should answer, but I took it to mean -cEV can be +$EV, not that -$EV can be good.
Seems like you could put some sort of extra value in having greater than some multiple of the blinds, and on having [your stack - some other stack] be greater than some multiple of the blinds. This could create local spots where additional chips are worth more instead of less than previous chips.
|
 |
|
|
Linear Mode
