When would you fold AA's

[GeeBeeQED]

Rek, are we not talking about only a $1000 outlay? If a player is making 50k and can't risk $1000 on such a nice shot even one time, it says something about him that he does not have a surplus of $1000 after an income of 50k or possibly he's just too fearful to play poker. Why would such a person be reading this forum? A $1000 dent in the bank roll would hurt (I've felt it before a few times) but if I never took a $1000 risk, I'd sure be turning down the best of it on occasion. You give me the best of it, I'm taking it, at lest some number of trials.

Really the question and your remarks brings up another thought I've had that relates to the gamblers falicy to some degree. We each know that each roll of the die is independant of the last roll and etc. However the original question really isn't about 1 roll at a time. It's about event grouping. If I choose to put the AA against the KK a fixed number of times then the events are interelated as the results are now based on X# of trials. So now the multiple events become one. In that if I've choosen to take the risk 4 times consecutively then the result is per 4, not per 1. The odds are less that the group of 4 will end happy than making the bet 1 single time. The amount won has a greater spread from the possible loss in the 4 trial example than the single trial.

[alavet]

i would stop after 20k. at least, i try to stop
lol

[Rek]

Hi GeeBee

Maybe I have mis-understood. What I think the OP meant was you have £1K and you get the option AA v KK. Yes I am taking that. You now have $2K and again I would accept. So you keep moving up and putting your whole winnings on the line each time.

Now if I am lucky enough to keep doubling I think, me personally, once I have $64K I am starting to think I will now take the money. I now have to put $64K on the turn of a card (at good odds I know) and this is reasonable money to me now. Where do I stop? I know that one beat and I have lost the lot. Keep going and that one beat will come and if it hasn’t hit you already it soon will. You lose the whole lot in one hit.

[PantsOnFire]

[ QUOTE ]
Really the question and your remarks brings up another thought I've had that relates to the gamblers falicy to some degree. We each know that each roll of the die is independant of the last roll and etc. However the original question really isn't about 1 roll at a time. It's about event grouping. If I choose to put the AA against the KK a fixed number of times then the events are interelated as the results are now based on X# of trials. So now the multiple events become one. In that if I've choosen to take the risk 4 times consecutively then the result is per 4, not per 1. The odds are less that the group of 4 will end happy than making the bet 1 single time. The amount won has a greater spread from the possible loss in the 4 trial example than the single trial.

[/ QUOTE ]
You're right of course in that OP is different because we are reinvesting all of our winnings each time. So theoretically, if you keep doing that you will eventually go broke regardless that you have an 80% edge each time.

This sort of game reminds of me of two games shows, Deal or No Deal and Millionaire. I cringe when a person says "Well, I came here with nothing so let's go for it". In my original post on the million dollar tourney question, I brought up an interesting situation which occurs often in Millionaire.

The player has won $64K and is now at the $125K question. He has no idea of the answer but uses his 50/50. Now his odds are 50/50 where he is putting $32K on the line to win %$61K. Obviously, he is getting the best of it. And lot's of people on the show will take this guess using the "I'm guaranteed $32K and I came with nothing".

However, if you go to an average Joe on the street and ask him to put $32K of his own money on this bet, most would find it risking too much of their life savings or whatever. People on the show forget that they aren't freerolling they are risking $32K of their own money.

I feel this is similar to the OP. Sure you started with $1K of your own money but whatever you have built that to it is still all your own money not a freeroll. When you bet your entire bankroll, you can't handle the variance.

So the anwser to OP question is there is a certain amount of money that people will simply not risk even when getting the best of it.

[Woolygimp]

0$.

I'd lose the first hand.

[JustCuz]

[ QUOTE ]
0$.

I'd lose the first hand.

[/ QUOTE ]

LOL!! That should pretty much end this thread.

Well done!

[jay_shark]

[ QUOTE ]
This is a bankroll management question. It is hard to give a consistent answer to the collection of questions of this type unless you use mathematics. An example of a mathematical approach is to use the Kelly criterion, which says to maximize the expected logarithm of your bankroll. That says when you win 82% of decisive hands, you are better off not betting if you would have to bet over 95.2% ~ 20/21 of your bankroll.

Someone following the Kelly criterion for bankroll management, who has $10k at home, should be willing to risk up to about $200,800, so should try to double up 8 times to $256,000 and then quit. This wins $255,000 about 0.82^8 ~ 20% of the time, and loses $1,000 about 80% of the time.

Having a larger bankroll would mean the player would risk more. Following a more conservative rule than the Kelly criterion means the player should tend to risk less.

[/ QUOTE ]

Pzhon , correct me if I'm wrong here but if you're getting 1:1 on your wager and that p=0.82 , then you should be willing to bet up to 64% of your bankroll .

f*=(0.82*2-1)/1 =0.64

So if you have 10k , aside from the 1k , then you should be willing to double up to 32000 and then stop .

This makes sense , since 16000/(16000+10000)~61.5% , Which means you should give it one last try and try to double up to 32000 and then stop .

[DarkMagus]

From a theoretical standpoint you can't really answer this without more information. It's completely dependent on your hypothetical financial situation which you haven't specified in the OP, so posters are using their own finances as a benchmark. If this $2000 is the only thing you have to your name, and you can't easily get back on your feet if you go broke, you'd be stupid to gamble with it. If you've got a lot of minimum buy-ins behind you (e.g. $10k or more), then keep going until the amount of money you're risking is larger than whatever the Kelly criterion says (which I think someone above put at 95% of your total bankroll). Once you hit that, cash out and start over again at one of the lower tables.

[jay_shark]

Dark , you shouldn't have to invest 95% of your bankroll to maximize your long term growth rate .

If you're an 82% favorite , then you should be willing to risk up to 64% of your bankroll .

f*=(0.82*2-1)/1 =0.64

[pzhon]

[ QUOTE ]
[ QUOTE ]
An example of a mathematical approach is to use the Kelly criterion, which says to maximize the expected logarithm of your bankroll. That says when you win 82% of decisive hands, you are better off not betting if you would have to bet over 95.2% ~ 20/21 of your bankroll.

Having a larger bankroll would mean the player would risk more. Following a more conservative rule than the Kelly criterion means the player should tend to risk less.

[/ QUOTE ]

Pzhon , correct me if I'm wrong here but if you're getting 1:1 on your wager and that p=0.82 , then you should be willing to bet up to 64% of your bankroll .


[/ QUOTE ]
No, that is incorrect. The Kelly criterion says you should be happiest if you can rescale the wager to about 64% of your bankroll, but you should prefer much higher wagers over walking away.

Maybe you should give me a little bit more credit, given our track records, or check what I said with a calculation of the expected logarithm before trying to correct yet another correct statement of mine.