Max % of Bankroll per Day

[Tarvold]

Hi Guys,

I'm just wondering what the maximum % of your bankroll do you would bet in a single day. Any info or pointing me in the direction of a few articles would be most appreciated.

Tarvold

[RickySteve]

∞

[SlappyWhite]

I do not have a max per day, max for a game for me would be 10%, but that is pretty rare maybe 50 times a year tops, most the time I play between 1 and 2 % on a single game.

[gordon gekko]

2% should be your max.

[LT Profits]

No such thing as a daily max. If a play has an edge, you bet it, be it two plays or 18 plays.

[gm2022]

Everyone has their own thing. Let me put it this way, if you have a roll and your willing to lose xx and will be ok if that happens than that should be your max. Some its 2% hell some its 30%, just do what makes you feel ok inside win or lose

[donjuan]

Kelly.

[JRS21386]

It varies for each bettor

[accuscoresucks]

3.5%
any given day

that would be alot of picks and stars

[ico2525]

I have read LT saying 1%. That doesn't mean you can't bet 10 games totaling 10%.

[JRS21386]

Betting 10 games a day will get you nowhere...

[durito]

Betting 10 games a day will get you nowhere...

i prefer 50

[JRS21386]

And your successful at this?

[element1286]

Betting 10 games a day will get you nowhere...

I usually have 10 or so baseball plays per day.

[LT Profits]

ico,

When did I ever say 1%? 99% of my plays are for 2.5% of BR, and the rare best bets range from 3.5% to 4.5%.

[reno cool]

Everyone has their own thing. Let me put it this way, if you have a roll and your willing to lose xx and will be ok if that happens than that should be your max. Some its 2% hell some its 30%, just do what makes you feel ok inside win or lose

this is my view. need to consider empirical factors.

[Brady2Moss]

I bet 1-2% of my bankroll and I usually have 10-15 plays a day.

[I.R.B]

I bet profit... Somedays i will bet 50% of my profit some days i will bet 10%... The thing about betting profit is you have to make profits to bet profits...

[rory borealis]

Simultanaeous Kelly Bets

2% should be your max.

2% would correspond to a Kelly win % of 53.31% @ -110

As I've stated elsewhere....If you take a look at the 4 or 5 year records of the pay-for-picks handicappers at the monitoring forums, you don't see many hitting over 55%.

The late Huey Mahl gave a formula for betting management which was authored by an electrical engineer from Ma Bell named John Kelly showed that at 54.5% proficiency one would calculate .545-(.455/.91)= 4.5% of one's bankroll.


At -105(5Dimes...GuardianGuaranty...BetJam on Fridays)the proficiency level of 54.5% would dictate a Kelly Bet of .545-(.445/.952)=7.7%


at -107(The Pig owned by LEGENDZ)...=6.8%



Kelly Bets get pretty hard on the GI tract, if one is even fortunate enough to have funds grow ("Honey! I shrunk the bankroll" was what I heard yelled out at Reno's Cal-Neva Sportsbook), and dictate utilization of anything which consumes 47 times its wieght in excess stomach acid(Rolaids sportsbook)

The loss of 1/2 point is devastating to one's optimal bet...the gain of 1/2 point adds an incredible safety buffer.


SIMULTANEOUS KELLY BETS

SBR Forum Mathematical Maestro, Ganchrow as well as Statistician, Stuart Holland of The United Kingdom & The MODERATOR at The Math Help Forum all have pointed out that MULTIPLE GAMES going off at the same time are independent events
They then provided a formula to calculate each bet





Richard Epstein, Peter Griffin, William Ziemba, Huey Mahl, Jim Barnes, Bill Whatmough and Mike Lee have been the authors who have discussed Kelly betting the most if I recollect correctly.





Again, if my recollection is good....many of the above have worked off of Griffin’s calculations….





If one looks at Wagerline’s site one will see that over the long run of more than 3 seasons one finds very few of the leading handicappers picking much better than 53%.





Gordon Gecko spoke earlier of dividing one’s stake into perhaps 50 betting units (2% Kelly).....


54.5% winning proficiency is tough to achieve and....t one had better be better over the long run than the best of those Wagerline cappers….



FAILING TO ACCOMPLISH SUCH AN HERCULEAN FEAT OF PROGNOSTICATION, we are pretty much left looking at 53% proficiency as an approachable, long-run benchmark to attain…

That then would correspond to a 1.35% Kelly bet when the book’s vig is 11/10.….

In this case…you may as well take up ditch-digging if you can’t beat 53%… and I would then recommend that the divisor be “74”(1.35%) instead of the previously-mentioned “50”(2.00%).





Turning to the question of simultaneous Kelly bets(sporting events occurring at the same time during the afternoon), we are fortunate to have the figures for 53.1% proficiency for DEPENDENT EVENTS


We'll use this DEPENDENT EVENTS MATH since at these low percentages... I don't think the curve between INDEPENDENT & DEPENDENT would deviate that much... if it does, I'd be grateful for one of the SBR Math Whizzes to step in & provide the correct figures

…so 53.1% proficiency over the long run of many seasons would mean dividing each stake into “67”(1.50% )betting units.





IN THIS CASE THEN of the simultaneous DEPENDENT Kelly bets…





1 BET…………..divide entire bankroll by 67 to determine optimal bet size





2 BETS………….divide entire bankroll by 58 to determine optimal afternoon’s betting pot and divide the result into 2 equal bets for the afternoon’s games





3 BETS…………divide entire bankroll by 55 to determine optimal afternoon’s betting pot and divide the result into 3 equal bets for the afternoon’s games


You won't get into too much trouble following that math since 55 is still less than 2% of bankroll bet

[tomcowley]

I might be able to understand your posts if they didn't remind me of ransom notes.

[rory borealis]

I might be

able to understand your posts if they didn't remind me of ransom notes.

[HeeeHAWWWW]

I might be able to understand your posts if they didn't remind me of ransom notes.

Very good, I award you a



To the OP's question: theory might say no max, but there are some days your judgement is just off, and/or weird results clump together. Setting maybe 10% max/day limits your exposure to this sort of thing.

There's also your betting psychology. Can you handle losing 25% of your roll in one day without going on tilt? I've had days where i won 9/10, and days where I've lost 9/10 - just out of random chance, both will happen eventually. If you're maxxed at 10%, it's never going to hurt too much.

[VegasVic]

I bet profit... Somedays i will bet 50% of my profit some days i will bet 10%... The thing about betting profit is you have to make profits to bet profits...

...and they give you cash, which is just as good as money.

[VegasVic]

My money management system is very simple and probably very boring to most, but I’ve never depleted a bankroll and in fact usually turn a profit. I keep sports betting in proper perspective. I'm not trying turn a toothpick into a lumber yard or even make a living doing it. I already have a job. The last thing I want to do is make the thing I love to do for recreation become work.

I wager on hockey alone and devote what limited time I can afford to that one sport. I invest in all aspects of the sport whether it be futures, exotics, or daily plays. They all contribute to the bottom line at the end of the year but I do keep a separate bankroll just for my daily plays.

I bet 1%, (and occasionally 1.5%), of my starting bankroll per bet, whether it be on a side or total. I start with a $10,000 bankroll.

However like most others, I will not adjust my unit size down if my bankroll balance falls below $10,000. I have faith in my ability that I can withstand the occasional losing streaks that are inevitable. If I did begin to lower my unit size down it will take that much more of a winning percentage, when things do turn around, to get back to even. For example…If I open the season by losing 10 straight bets at $100 each my bankroll falls to $9,000. If I adjusted my unit size to $90, (or 1% of bankroll), I’d now be forced to win 11.1 bets to get back to even, (providing I’m betting all even money wagers).

On plays that I feel are particularly strong due to a soft line and deserve more financial attention I will press it up to a unit and a half. Possibly up to 10% of my wagers will be of this variety. But I never risk more than 1.5 units on any wager, (unless there’s a scalp involved and that‘s a whole different animal and separate cash reserve).

[Arilou]

I find that on an NFL Sunday or an NCAAF Saturday the percentage of bankroll you have in play should approach 100%, and counting money rebet as it comes back to you it should sometimes top 100%. It can't get to 100% straight up all the time because you'll often have some money at the wrong sportsbook where you don't have a play, but where there are enough reasons that you need to keep it funded, but the key is that not all plays are wagers. If you have a bunch of accounts and can't find arbitrage or market-based plays for your money on a football day you are not trying all that hard or need better math. Even a baseball day presents a ton of these opportunities.

However, if you are making plays rather than playing markets or using mathematical models then that is different. Using only close to even-money wagers at market prices makes the above approach seem beyond reckless, but there is no max. A solid capper could easily have 20 plays on an NCAAF card between spreads and totals, which is far tighter than 10 MLB plays in a day, and risking 2% per play at -105 for 40% of your bankroll would be perfectly reasonable there. VegasVic, however, makes an excellent point that you need to be able to take the loss. If you can't handle losing what you're putting in play, you'll have to tone it down.

But overall my approach is that between sides, moneylines, totals, teasers, covariant parlays, pickoffs, arbitrage, halftimes, team totals and props if you can't find a good use for the bulk of your money on a football weekend then you're not trying all that hard. Someone like VegasVic who bets Hockey obviously has a different situation.

[Mr Winters]

I had a significant %age of my (small) BR in play last Sunday in NFL games (about 80%) and had a bad day. I learned a very important lesson (for probably the 5th time in as many years) to watch big lines. Giving 13 to the Giants in an away game, or 9.5 to a brady-less pats is just too much and shouldn't be touched. Actually, always take the home-dog, not the favorite. You make money like that. Not like what i did, which was go 10-16 or something like that (spreads and o/u's).

Now of course, a field goal in each of two games and i would have had a great day. But since those FGs weren't kicked, and when you account for the couple of parlays i also lost, i had a very bad day and set myself back seriously on the month.

If i put myself out like that again, i might have a great day and win it back, but i also might stink it up again and double my losses, even at 5%/wager.

Bottom line is that my recent experience would recommend that something like 10%/day be on the table, that way a 1-9 day doesn't f*** up your week.


As an aside i would say that i figured more "good" bets would have a better chance of "evening" out and giving me a hallowed 55% kind-of-day. But if it doesn't work out, that thinking goes into the toilet along with the supposed "experience" and "common-sense" that made me think of it that way.


I am trying to figure out the best way to ride out winning and losing streaks. Betting 1% per with such a small BR is so boring as to be almost futile. However, the life-lessons i am learning about how this stuff applies to our (my) lives is probably very valuable indeed.

You've got to play to win - and you can't play when you are broke. DON'T chase your losses (did that one again last night and lost a couple units on the second half of the Tulane game), and DON'T put more on the table than you can afford to lose (who can afford to lose more than 10% of their BR in one day (besides me who is working with very little to begin with and shot myself in the foot anyway)?!?

Handicapping is an INCREDIBLE way in which to multiply and "invest" your money - but you must understand probability MORE than you understand sports. The way of the SQUARE is to think that your knowledge of sports outweighs the law of averages. From time immemorial those guys are the smartest losers in the world.

Don't be "that guy," eh?

GL everyone! May you hit 56% this year and every year!

[Peep]

I have never used a Kelly, I don't feel I have ever had a proper feel for my edge on any one event.

But is this true with a Kelly?

showed that at 54.5% proficiency one would calculate .545-(.455/.91)= 4.5% of one's bankroll.

If so, I don't think overbetting would be a real problem.

"It is not what we don't know that kills us. It is what we KNOW that is NOT TRUE that gets us everytime".

Lots of people thinking they have a 55% edge on a game, not having any me thinks that causes bankroll shrinkage (a can be fatal betting disease).

[roasthawg]

I use Kelly and normally end up with most of my bankroll in play on the weekends. With 55% confidence it takes 18 games to get the whole thing in play. I normally don't rebet during the weekend...I let all of the bets play out before I recalculate my bankroll. I'm probably losing money this way but it's mostly superstition...if I'm up early in the day I don't want to put more on the later games because I feel like I'm bound to lose. I was 7-0 against the NFL last week...bad strategy.

[rory borealis]

@ -110 ----54.5% winning proficiency then 4.5% is the correct bet size.

I do not think 54.5% is achievable over the long run

I have never used a Kelly, I don't feel I have ever had a proper feel for my edge on any one event.

But is this true with a Kelly?

showed that at 54.5% proficiency one would calculate .545-(.455/.91)= 4.5% of one's bankroll.

If so, I don't think overbetting would be a real problem.

"It is not what we don't know that kills us. It is what we KNOW that is NOT TRUE that gets us everytime".

Lots of people thinking they have a 55% edge on a game, not having any me thinks that causes bankroll shrinkage (a can be fatal betting disease).

[roasthawg]

Ok, instead of starting a new thread I'll post this here first...

I'm a full Kelly bettor and I feel that I have a 55% edge on two games and bet accordingly. I also feel that I have a 55% edge on combining those two teams in a 6 point teaser...is the proper play to treat these as 3 individual bets each at a 55% edge?

[Thremp]

A day is an entirely arbitrary time period.

[ico2525]

If you think you have a 55% advantage individually on multiple games, then you can assume that you have a .55^3 (55% cubed) advantage on all three games. That comes out to 16.6%. The juice on a 3 team parlay is much worse relatively than the -110 straight up though. The same thing applies to teasers.

If you bet a teaser, the added points should theoretically increase your estimated edge. If you feel you're 55% on a game at +2, you should be much more confident when you're laying +8.

[roasthawg]

If you bet a teaser, the added points should theoretically increase your estimated edge. If you feel you're 55% on a game at +2, you should be much more confident when you're laying +8.

Yeah, exactly. So my question is that if I feel I have a 55% edge on two games individually and also have a 55% edge when I combine them in a -110 6 pt teaser then do I treat them all as seperate bets? Or is this over betting since I already bet about 1/18 of my bankroll apiece on these two games before the teaser?

Basically what it is is that I'm 55% confident in each game individually ats and 75% confident with the added 6 points. .75 squared is 56% so is this a seperate bet?

[pats3peat]

I bet profit... Somedays i will bet 50% of my profit some days i will bet 10%... The thing about betting profit is you have to make profits to bet profits...

what about when you lose the profit

[Ganchrow]

Yeah, exactly. So my question is that if I feel I have a 55% edge on two games individually and also have a 55% edge when I combine them in a -110 6 pt teaser then do I treat them all as seperate bets? Or is this over betting since I already bet about 1/18 of my bankroll apiece on these two games before the teaser?

Basically what it is is that I'm 55% confident in each game individually ats and 75% confident with the added 6 points. .75 squared is 56% so is this a seperate bet?

You've got 4 separate bets (7 if you include the 3 two-team teasers), 3 of which are all mutually independent (the 3 singles) and 1 (or 4 with the t2o-team teasers) which is positively correlated with the other 3.

As such, it would not be correct to treat all the bets as independent. What you'd need to do would be build the complete outcome matrix, determine the outcome probabilities, specify the utility for each outcome, and then maximize the expected utility.

So for example, if the three single win outcomes are denoted by A, B, and C, and the three component teaser outcomes are denoted by A_T, B_T, and C_T, (where A, B, and C, are proper subsets of A_T, B_T, and C_T, respectively -- meaning that a straight bet win implies a teaser leg win) then the 27 discrete outcomes would be:

1. A_T B_T C_T
2. A_TB_TC
3. A_TB_T
4. A_TBC_T
5. A_TBC
6. A_TB
7. A_TC_T
8. A_TC
9. A_T
10. AB_TC_T
11. AB_TC
12. AB_T
13. ABC_T
14. ABC
15. AB
16. AC_T
17. AC
18. A
19. B_TC_T
20. B_TC
21. B_T
22. BC_T
23. BC
24. B
25. C_T
26. C
27. all lose

So this means that for each of the three singles you'd need to specify two probabilities: 1) the probability of winning the straight bet, and 2) the probability of winning the teased spread but NOT the straight bet. Subtracting the sum of these two probabilities from 1 would then give the probability of losing the both the straight bet and the teased spread.

The attached spreadsheet details the problem and solution. Probabilities and odds are the indepedant variables, bets sizes represent the 4 solutions. (Note that the problem is specified is technically incomplete as the 4 constituent parlays should also be a part of the complete solution.)

Go to Tools=>Solver... to bring up the solver menu. Click the "Solve" button to solve for bet sizes.