Ganchrow, how about using Kelly to wager on Phoenix Suns Over Totals?

[imgv94]

What would be the best system to utilize this strategy?

O.K we all know the Suns are notorious for scoring an obscene amount of points and it's uncommon for a long string of their games to be under the total..

What would be the best way to start this? Honestly I don't see Suns totals losing more than 4 times in a row all season..

Is there a way we could make profit on this with minimal risk?

[imgv94]

A better team might be the Washington Wizards, they have never had more than 3 consecutive unders the entire season..

Imagine the money that could of been made?

[Ganchrow]

I'm afraid I don't really understand your question as it relates to Kelly.

[VegasDave]

I'm not sure how to solve Kelly, but I found this amusing...

Phoenix once went for 6 unders in a row this season, which would devestate the system of doubling every game. Washington, however, has at worst gone 3 games in a row with the under.

Assuming you had $2000 to gamble with, and were aiming to make $100 per over...

at Cleveland U (194) Bet 110 (-110, 1890)
vs Boston O (200) Bet 231 (+210, 2100)
at Orlando O (204) Bet 110 (+100, 2200)
vs Indiana O (202) Bet 110 (+100, 2300)
vs Milwaukee O (208½)Bet 110 (+100, 2400)
vs New Jersey U (201½)Bet 110 (-110, 2290)
at New York U (211½)Bet 231 (-231, 2059)
at Detroit U (200½)Bet 485 (-485, 1574)
vs Cleveland O (196½)Bet 1019 (+926, 2500)
at Dallas U (204½)Bet 110 (-110, 2390)
at Houston U (193) Bet 231 (-231, 2159)
at Memphis U (190½)Bet 485 (-485, 1674)
vs Detroit O (197½)Bet 1019 (+926, 2600)
vs Atlanta U (199) Bet 110 (-110, 2490)
vs Charlotte O (200½)Bet 231 (+210, 2700)
at Chicago O (202) Bet 110 (+100, 2800)
vs Dallas U (203½)Bet 110 (-110, 2690)
at New York O (206½)Bet 231 (+210, 2900)
at Philadelphia O (202) Bet 110 (+100, 3000)
vs Houston O (189) Bet 110 (+100, 3100)
vs Denver U (221½)Bet 110 (-110, 2990)
vs Miami O (199) Bet 231 (+210, 3200)
at LA Lakers O (213½)Bet 110 (+100, 3300)
at Denver O (213½)Bet 110 (+100, 3400)
at Sacramento O (211½)Bet 110 (+100, 3500)
at Phoenix O (229) Bet 110 (+100, 3600)
vs Memphis O (207½)Bet 110 (+100, 3700)
at Charlotte O (208½)Bet 110 (+100, 3800)
vs Orlando O (211) Bet 110 (+100, 3900)
at Milwaukee U (229) Bet 110 (-110, 3790)
vs Milwaukee U (228) Bet 231 (-231, 3559)
vs LA Clippers O (213½)Bet 485 (+441, 4000)
at Toronto O (220½)Bet 110 (+100, 4100)
vs Chicago U (216½)Bet 110 (-110, 3990)
at New Orleans U (203) Bet 231 (-231, 3759)
at San Antonio U (208½)Bet 485 (-485, 3274)
vs Utah O (215½)Bet 1019 (+926, 4200)
vs New York U (219½)Bet 110 (-110, 4090)
at Orlando O (205) Bet 231 (+210, 4300)
vs Boston O (209) Bet 110 (+100, 4400)
vs Phoenix O (231½)Bet 110 (+100, 4500)
at Detroit U (205) Bet 110 (-110, 4390)
at Boston U (205½)Bet 231 (-231, 4159)

Needless to say, that 4th straight under would be crippling. But anyone who did this would have doubled their money this season

[Ganchrow]

I'm not sure how to solve Kelly, but I found this amusing...

Phoenix once went for 6 unders in a row this season, which would devestate the system of doubling every game. Washington, however, has at worst gone 3 games in a row with the under.

This is of course the infamous Martingale strategy.

And it's been the ruin of many a poor boy.

[VegasDave]

This is of course the infamous Martingale strategy.

And it's been the ruin of many a poor boy.

I can imagine.

No matter how long it works out for you, all it takes is the one time it doesnt.

[imgv94]

Sorry for the confusion..

Let's say you used a smaller base wager?

[Ganchrow]

Sorry for the confusion..

Let's say you used a smaller base wager?

You mean smaller than that recommended by full Kelly? Sure. You could do that. But you could do that with any bet. I just don't see why this bet is special.

[imgv94]

Well the Wizards were a better example.. Does seem risky though..

Thanks usckingsfan for you're help..

[Ganchrow]

Well the Wizards were a better example.. Does seem risky though..

Thanks usckingsfan for you're help..

But what does this have to do with Kelly?

[imgv94]

But what does this have to do with Kelly?

Martingale actually was what I was referring too.

[tacomax]

If you are seriously considering going down the Martingale route then you should seriously consider stopping gambling.

[Ganchrow]

Martingale actually was what I was referring too.

Ahh. Ok. Now I get it.

Yeah, you should probably stay away.

[imgv94]

If you are seriously considering going down the Martingale route then you should seriously consider stopping gambling.

You are probably right..

[VegasDave]

Well, if you have enough money, the Martingale route is fairly safe...

Granted, the super high risk to low reward looks terrible, but that is because the actual likelihood of losing that much is so slim.

Take your same 2000 bankroll and aim to make 10 per over...

(Just using estimates here)
11 / 23 / 48 / 102 / 210 / 430 / 980 / 2000 / BANKRUPT

It would take you 9 unders to go bankrupt. Assuming you had a 50 % chance per over/under (which, judging from how often they hit it, you've got better than that...), you would only have a 1/512 chance of going broke.

And "10 dollars a game, so what?"... You profited $250 using the Wizards model so far, and came no where close to 9 straight unders.

Sucks Ganchrow that you pointed out to me that it was Martingale... when I was like 11, I came up with this theory based on roulette wheels (playing black or red infinitely doubling up each time)... I thought it was an original idea

[VegasDave]

By my calculations, 2000 to win 250 is a -800. Would you not bet 2000 at -800 with odds of 512 to 1?

[MrX]

Is there a prize for being the first to post the flaw?

[Mudcat]

One major flaw is how it eventually costs you all your money.

Other than that, Martingale is pretty good.

[VegasDave]

The prize, MrX, is saving the person who tries this out.

Is there a fallacy in my math? The only flaw I see is the inevitablity of the 10 straight losses - but it would be pretty unlikely.

[MrX]

By my calculations, 2000 to win 250 is a -800. Would you not bet 2000 at -800 with odds of 512 to 1?

For starters, the odds of losing 9 in a row is 511 to 1 at any given starting point. The odds of losing 9 in a row at any time in the trial is, of course, much higher.

[VegasDave]

For starters, the odds of losing 9 in a row is 512 to 1 at any given starting point. The odds of losing 9 in a row at any time in the trial is, of course, much higher.

True.

Going on this season's numbers, only two teams have put together streaks of 9+... Memphis had a 12 game OVER streak, and Toronto is in the midst of an 11 game UNDER streak.

Assuming you took your favorite team and stuck with the OVER or UNDER game in and game out, you'd make money on 29 of the 30 teams in the league.

[MrX]

The betting scheme you posted has you losing $3804 after 8 losses in a row.

[VegasDave]

Explain? Are you referring to the one in post #15?

[Ganchrow]

Well, if you have enough money, the Martingale route is fairly safe...

Granted, the super high risk to low reward looks terrible, but that is because the actual likelihood of losing that much is so slim.

Take your same 2000 bankroll and aim to make 10 per over...

(Just using estimates here)
11 / 23 / 48 / 102 / 210 / 430 / 980 / 2000 / BANKRUPT

It would take you 9 unders to go bankrupt.

If you were betting to win $10 each round, your progression would look like this: 11, 23.1, 48.51, 101.871, 213.9291, 449.25111, 943.427331, 1981.197395. This means after 8 losses you'd have already wagered $3,772.285936. Your chances of losing all 8 games over an 8-run stretch are 1 /256. This means that your expected loss, after playing this one round, is -$4.77 (you lose $3,772.285936 with about 0.391% probability and win $10 with 99.609% probability. Play for ten rounds and you're probability of going bust is 3.838%. Play for a hundred and your probability of going bust is about 1 out of 3. If you play for 377 times (so your winnings, were you to win, would be the same as what you'd lose after a single bad round) are better than 3 out of 4.

Assuming you had a 50 % chance per over/under (which, judging from how often they hit it, you've got better than that...),

If you really believe that either the over or under hits at 52.38% or more over the long haul, then you should just go out and play the -110 (or better if you can get it) with Kelly or even a flat betting scheme and completely scrap the idea of using the Martingale.

In Joseph Buchdahl’s canonical Fixed Odds Sports Betting: Statistical Forecasting and Risk Management, Buchdahl uses Monte Carlo simulations to determine risk/return characteristics for 6 different staking strategies while varying player edge, unit size (with a “unit” be differently defined under each staking strategy), and average odds level (from large dog to large fave). The staking strategies Buchdahl examined were:

1. Level staking
2. Percentage staking
3. Fixed profits staking
4. Martingale staking
5. Pyramid staking
6. Kelly staking

What he determined was that of all these systems, Martingale and Pyramid staking had the most risk, the highest probability of bankruptcy, the highest probability of failing to make a profit, the lowest expected finishing bankroll size (for bettors who had not secured an edge).

Sucks Ganchrow that you pointed out to me that it was Martingale... when I was like 11, I came up with this theory based on roulette wheels (playing black or red infinitely doubling up each time)... I thought it was an original idea

Personally, I wouldn't be quite so proud of this achievement.

Basically, for any given bankroll target, Martingale maximizes your chances of hitting that target at the expense of the penalty you'd face were you to lose. This would be the only sort of scenario where using Martingale would make sense:

You're in a foreign country and have no way of reaching any friends or family. You're stricken with a deadly disease that will kill you in 20 minutes unless you can afford to pay a doctor the $794,280.0466 he would charge for your treatment. Your current bankroll is $793,280.0466 and you are in a casino that only features roulette. Using the Martingale on red/black (or odd/even, or any other bet that pays 1:1) would represent your best opportunity to make that last life-saving $1,000.

This way you'd die with 0.195% probability and live with 99.805% probability.

[Ganchrow]

Assuming you took your favorite team and stuck with the OVER or UNDER game in and game out, you'd make money on 29 of the 30 teams in the league.

Sure.

But the 29 times out of 30 you won, your wins would be very small, while the one time out of 30 that Martingale didn't come through, your loss would be huge.

[Ganchrow]

For starters, the odds of losing 9 in a row is 511 to 1 at any given starting point. The odds of losing 9 in a row at any time in the trial is, of course, much higher.

You got it.

10 points to the first person who can tell me the approximate probability of losing 9 in a row at some point (assuming a 50/50 proposition) during an 82-game stretch.

[VegasDave]

Good stuff Ganchrow... I was just trying to play devil's advocate... but obviously not doing a very good job mathmatically

I'm still proud! Just because communism only works perfectly in theory doesn' t mean it wasn't a good idea!

[Mudcat]

Sucks Ganchrow that you pointed out to me that it was Martingale... when I was like 11, I came up with this theory based on roulette wheels (playing black or red infinitely doubling up each time)... I thought it was an original idea

Actually there must be a million people that have come up with this theory. It just keeps coming up over and over.

Fortunately there is a guy like ganchrow who has the patience to explain the flaws. Horrible people like me who have been around for awhile might take one brief shot to tell you it doesn't work and if you still argue, I'd just roll my eyes and say, "Sure dude, go for it."

[Arilou]

Assuming no game will ever push, your probability of going broke during 82 games is 6.8402%

[Ganchrow]

6.8402%

How'd you come up with that?

[Korchnoi]

You got it.

10 points to the first person who can tell me the approximate probability of losing 9 in a row at some point (assuming a 50/50 proposition) during an 82-game stretch.

(82-9)x(1/512) = 14.25%?

[Ganchrow]

(82-9)x(1/512) = 14.25%?

Nope, that's not quite right.

[Korchnoi]

Nope, that's not quite right.

i'm assuming he keeps playing even after he goes busto, which I should assume for the sake of your question right? i'll give it some thought tonight.

I'm assuming he's betting a specific way (the under, or "tails") also.

[JDK192]

½^9*(83-9)+1/2^10*(83-10)+…+1/2^82*(83-82) / 4

Answer is approx 7%

[Korchnoi]

½^9*(83-9)+1/2^10*(83-10)+…+1/2^82*(83-82) / 4

Answer is approx 7%

I don't quite get your methodology, could you explain what you're doing? I ran a montecarlo a few hundred times and got around 7%.