Huge Value On THE TRIBE

[Thremp]

[ QUOTE ]
[ QUOTE ]
CLE +151 is a no-brainer. Send it in.

[/ QUOTE ]

27% of your bankroll?

[/ QUOTE ]

If you have terrible analysis like yourself and he wants to get rich the fastest, then yes. Instead he probably 1) doesn't trust math 2) is going to insult my wager size and call me a recreational player 3) doesn't want to make money faster than he can otherwise.


I'm really open for a solid argument about why you should not bet Kelly other than "Losing monies makes me more unhappy than winning money makes me happy, ergo my utility curve is not log like Kelly assumes." That is a legit argument that I can live with.

[TomCowley]

Agree, the biggest concern is margin of error. I prefer to think of that as betting full kelly on a "safe" estimate of edge, but you can call it betting 1/3 kelly on your original estimate of edge, and you're going to get ballpark similar bets. I was responding to a post where the estimate of edge was taken for granted to be correct, in which case you really have to stretch for a reason to voluntarily bet less.

[sublime]

[ QUOTE ]
[ QUOTE ]
[ QUOTE ]
CLE +151 is a no-brainer. Send it in.

[/ QUOTE ]

27% of your bankroll?

[/ QUOTE ]

If you have terrible analysis like yourself and he wants to get rich the fastest, then yes. instead he probably 1) doesn't trust math 2) is going to insult my wager size and call me a recreational player 3) doesn't want to make money faster than he can otherwise.


I'm really open for a solid argument about why you should not bet Kelly other than "Losing monies makes me more unhappy than winning money makes me happy, ergo my utility curve is not log like Kelly assumes." That is a legit argument that I can live with.

[/ QUOTE ]

#1 you are a thick headed tool.

#2 kelly cant be disproven by math. the problem is the real life issues are far too important to just glaze over. its not about being a 'pussie' its about having a mid six figure bankroll that would be very hard to replenish outside of gambling.

here is what i see. somebody relatively new to sports betting (like a year or so) who ran into somebody who both has a clue and ran hot (crockpot) and decided wow this is easy, have a big edge bet big....thing is, guys who have bet year round for multiple years have run into bad stretches that consume 50-60% of their bankroll, betting only 1-2% per bet. now obviously this wont happen to you because you're a genius. however, it can and will happen to most. using kelly over the worst of that period would have put the PRO (we are not counting people who bet for relatively small amounts and can replenish their rolls via a real job w/o much stress) pretty much out of commission.

the other issue with kelly is that its really hard to define your edge (again, you are so sharp this is not an issue for you). its hard enough to be on the 'correct' side, never mind estimating your actual edge.

[sublime]

kelly is a great tool btw, i dont disagree with that. i just think for somebody who bets for a living (or most of it) anything > than half kelly is just asking to go close enough to broke that they will need a real job to continue. even half kelly is v v aggressive.

[TomCowley]

Hell, having 2 dimes on the ALCS prop doesn't make me *happy*, but that's a situation where I'm really sure that I can assign a lower bound on edge with >95% confidence, and I believe in math. The only justification for betting less, given my roll and that it won't lock me out of bets on 5D while it's pending, is "I'm a pussy". As much as I didn't like it, I knew it was the right play.

[NajdorfDefense]

[ QUOTE ]
The only danger in full kelly is that you're continually overestimating your edges. If you make a conservative estimate of your edge, and you have a realistic definition of your bankroll (rent due in 3 days is not part of your bankroll), failing to bet full kelly when possible is like hating money but worse, like raping it and putting it in porno films in return for jujubees.

[/ QUOTE ]

[NajdorfDefense]

sublime, I disagree that figuring out the correct side is that incredibly hard [if you have a clue] - the BSP pick threads here for 3+ years are evidence of that, there are books written that predate Wong and Yao showing where the value is in certain sports year after year after year.

I do agree that most bettors overestimate their edge. This is why I don't bet full kelly except for extremely rare occasions where you find a line that is both 'solid value' and 'off-market' or I get injury/news info before the books. I think this is why most of us bet < full Kelly, but does not mean that is correct for people with linear utility functions for their net worth.

Of course, Kelly betting doesn't work for sizing my craps bankroll for AC next month. [img]/images/graemlins/mad.gif[/img]

[TomCowley]

[ QUOTE ]
kelly is a great tool btw, i dont disagree with that. i just think for somebody who bets for a living (or most of it) anything > than half kelly is just asking to go close enough to broke that they will need a real job to continue. even half kelly is v v aggressive.

[/ QUOTE ]

This is so backwards. Your bankroll is the maximum amount of money you can AFFORD to lose. If you have future costs that come out of your "bankroll", then your real bankroll is the amount you can lose and still cover those costs. Betting Kelly (adjusted for margin of error concerns, of course) on that bankroll is correct. Betting some consistent fraction of Kelly on an overestimated bankroll has no mathematical basis. At best, if you're lucky and your fraction of Kelly approximates the fraction of your "bankroll" that really is your bankroll, your numbers will be close. At worst, you're losing a ton of value.

[MyTurn2Raise]

[ QUOTE ]
[ QUOTE ]
[ QUOTE ]
[ QUOTE ]
CLE +151 is a no-brainer. Send it in.

[/ QUOTE ]

27% of your bankroll?

[/ QUOTE ]

If you have terrible analysis like yourself and he wants to get rich the fastest, then yes. instead he probably 1) doesn't trust math 2) is going to insult my wager size and call me a recreational player 3) doesn't want to make money faster than he can otherwise.


I'm really open for a solid argument about why you should not bet Kelly other than "Losing monies makes me more unhappy than winning money makes me happy, ergo my utility curve is not log like Kelly assumes." That is a legit argument that I can live with.

[/ QUOTE ]

#1 you are a thick headed tool.

#2 kelly cant be disproven by math. the problem is the real life issues are far too important to just glaze over. its not about being a 'pussie' its about having a mid six figure bankroll that would be very hard to replenish outside of gambling.

here is what i see. somebody relatively new to sports betting (like a year or so) who ran into somebody who both has a clue and ran hot (crockpot) and decided wow this is easy, have a big edge bet big....thing is, guys who have bet year round for multiple years have run into bad stretches that consume 50-60% of their bankroll, betting only 1-2% per bet. now obviously this wont happen to you because you're a genius. however, it can and will happen to most. using kelly over the worst of that period would have put the PRO (we are not counting people who bet for relatively small amounts and can replenish their rolls via a real job w/o much stress) pretty much out of commission.

the other issue with kelly is that its really hard to define your edge (again, you are so sharp this is not an issue for you). its hard enough to be on the 'correct' side, never mind estimating your actual edge.

[/ QUOTE ]


sublime, Thremp is really good at arbing and is pretty good at picking off the low-hanging fruit of bad lines and line movements as well as knowing who to follow

[sublime]

[ QUOTE ]
Your bankroll is the maximum amount of money you can AFFORD to lose.

[/ QUOTE ]

ok

[ QUOTE ]
If you have future costs that come out of your "bankroll", then your real bankroll is the amount you can lose and still cover those costs.

[/ QUOTE ]

no. the future costs come out of expected winnings earned via the bankroll. if you lose 80% of it betting like a mad man, then you cant bet enough to cover those expenses and you need to either dip into savings (bad) or get a job (bad) until you regrow your roll to the point where you CAN pay those expenses.

or you can avoid that (it will happen) and bet via a fractional kelly method.

anybody who is a proponent of full kelly just hasnt seen what variance can do.

i am really done w this now.