Who pays the vig !

[AussieVamp2]

ok, another question wm,

when playing with your simulations and considering the plateau type thing, any comments on when is a good time to step it up?

[AussieVamp2]

or down, of course......

the Miller deal was to have a threshold, and go up once you got there, but not go down until you got a certain way back again from memory, like the midpoint of the last level, to prevent what he called the 'whiplash' effect, so you dont bet 2, 1, 2, 2, 1, 1, 1, 2 sort of thing presumably

[buckeye]

I side with Miller, and irish, on this question. The best way to illustrate it is to compare 11/10 to no vig! If there is no vig, like acesgold's special, then the loser still loses $110 betting $110. But the winner gets $110 betting $110, with the vig he gets only $100. So the WINNER is the one that paid the vig. The loser always loses 100% of his wager, but the winner only gets 90.9% profit out of what they would have if there were no vig. Don't see how you can see it any other way! The loser always loses 100% of their wager while the winner always collects LESS than true odds, whether its spread or $ line betting.

Miller is also one of the few "pros" who accurately knows the advantages and disadvantages of parlays, and their true nature/value.

[zippy]

Buckeye, you say "Don't see how you can see it any other way!" OK, let's compare 11/10 to no vig, like in your example. Instead of risking $110 to win $100, I'll be risking $100 to win $100. If I win, I get $100, which is the same as if there were vig. If I lose, I lose only $100, instead of that plus the extra $10 vig I'd normally have to pay as a loser. So the loser pays the vig.

Again, both ways of describing the situation--this way and your way--are exactly equally accurate, as long as you are consistent. You can say either that you get shorted on your wins or that you pay extra for your losses, just not both.

[greg]

zippy makes sense. If a book was offering a zero vig special the risk to win 100 would only be 100 therefore the losing wager would only lose 100 when trying to win 100 as opposed to losing 110 at normal vig wagering. Lets say i have 110 dollars to my name i wager to win 100 at zero vig i win my cash worth is now 210, now lets say I make the same wager laying 110 and win my cash worth is now also 210, therefore It can be argued that the winner pays zero vig as evidenced by this scenerio where successful wager has the same result on bankroll independent of presence of vig on wager.

[wintermute]

aussie

I generally set levels based on the increments at which books handle wagers. Say I have 20 dimes capital and I'm betting 1 percent of capital per game. So I'm wagering $200 a contest. I'll up my wagers to $250 when my capital reaches 25 dimes. I would do the same thing on the downside. Bet $150 after my capital drops to 15 dimes.

I kind of pull a JR when I start betting at a certain plateau. I'm reluctant to backtrack to the previous level too quickly, not because of any whiplash but because I believe in myself - if I wager less I win less. You either have faith in what you can do or you don't and you either believe in Kelly or you don't.

I had one truly terrible day this baseball season, losing 15 out of 16 wagers but I'm still ahead by a decent margin ( some wagers were parlays so it's not quite as horrific as it looks ). It all averages out in the end.

'mute

[AussieVamp2]

yeah, makes sense to me

but at the levels at which they handle wagers?

you can bet 247 if you want to at at a time at most places can't you?

It is human nature to do the divisible by ten thing though

[irish]

I just shake my head, Sometimes there is no point in argueing,illustrating or even infact proving,,,,,,,,,,whatever,,,,,Sincerely Irish.......

[wintermute]

Multiples of 10 make the math easier to check. Also some of my favourite books like WSE won't let you bet except in multiples. It's too bad - otherwise they'd be perfect.

'mute

[AussieVamp2]

yeah, canbet and DAS do that too - although DAS has a bunch of them so you can construct any 5 dollar multiple

but it is annoying, for instance in the case where I might want to construct my own half goal handicap, to get it right you do want decimals etc., then

text boxes are easier to code than list boxes too, so weird

[reno]

I'd like to see what you guys who have studied and practiced various bankroll-management systems think of the program that my ex-scalping partner is using for his baseball plays.

He ran his computer-program baseball system through--on a cumulative basis, so there was no backfitting--for the last 4 seasons. An all-underdog program, it averaged 52.6% winners at a +117 take over the four years. The program won every year, and results varied from good to great. The program generates, on average, three to four plays per day. He continually re-ran the program through the 4-year data base, trying different money-management systems to see which offered the highest average return over the four years. And it turned out to be an almost pure Kelly that worked best. Consequently, even though he expects wild fluctuations in his bankroll, he has decided to bet 16% of the bankroll on each play because he believes that he will make the most money by doing that.

Based on the information that I've provided, do you agree or disagree with his decision?

[sportshobby]

I wanted to first say that it was halarious, then I realized that you were serious.

Eventually he'll have an 0-4 day. I know I don't want to wake up one morning and have 64% of my bankroll in the toilet.

[Fedya Fussball]

Sportshobby:
A 0-4 record would drop your bankroll by 50.3% if you bet at a 16% clip...

A 0-6 record would drop your bankroll by 64.9%...

You are look in a bad streaks, but if reno is winning at 52.6%, then he is losing at 47.4% and to lose 6 in a row he has a p(6 loses) is 1.13% or pretty close to 60/1.

FF

[sportshobby]

FF, here's one things people often overlook when testing out the application of the Kelly criteria. Assuming you have four plays in one day, and they are all night games, they have to be put in all at once.

The Kelly criteria assumes one wager at a time is placed, graded, and the outcome used to determine the amount of the next play.

I assumed that he would put in four plays at 16% of bankroll each.

[reno]

Sportshobby, you are right. He puts in all of his plays in the morning. So an 0-4 record would drop his BR by 64%. Given the possibility of that happening, which he is well aware of, he still chooses to bet 16% of his bankroll on each play because, according to his computerized replays of the previous 4 seasons, that percentage generated the highest return.