Sportsbetting on the edge

[bblight]

I was in Vegas a few weeks ago and dropped by the Gamblers Bookstore. I've read most of the football books they have so I started browsing and found a notebook authored by John C Tarbet titled "Sportsbetting on the edge - a gamblers handbook for success"
In general this is a very detailed, 200 page statistical wagering guide whereby you determine the size of your bet based on the size of your bankroll and your proven winning percentage, using a statstical formula laid out by Mr Tarbet.
In effect, you increase your bet when you win and decrease it when you lose, but based on a statistical % rate.
My instincts tell me to stay with the tried and true, but you can't grow in a static environment, so I'm considering giving this a small try with a seperate bankroll.
Has anyone ever read the book and tried this formula? What do you think? Is it worth a try?

Any feedback would be appreciated.

Good luck!

 

I haven't read the book you mentioned, but I'm almost sure formula, Mr Tarbet is talking about, is called the Kelly Criterion. At least formula sounds like that.
In US people seems to use KC as you mentioned, based on your previous success. I have never understood what betting history has to do with a size of a bet. Maybe if your betting history includes 1 000 bets on football and you have gone 600-400, it's OK to use that kind of formula. But I doubt many of the bettors doesn't have that kind betting history. At least betting only NFL it's impossible. Most of them doesn't even have a history from the season 01-02.

In Europe, or at least in Finland, people uses the KC based on the probability of the match. Own estimation vs bookmakers odd. Here is the formula:

B=(ES*ODD-1)/(ODD-1)

ES=own estimation
ODD=bookie's odd

One problem for US bettor is that you can't use -110,-150,+160...type of odds. Or at least you have to create a formula to convert those odds to decimal numbers. Easiest way is to use decimal odds.

Lets say you have a bankroll worth of 2 000. You estime that Anaheim has a 55% chance to win. Odd on Anaheim is 2.20 (+120).

B=(0.55*2.2-1)/(2.2-1)= 0.083

So you should invest 8.3% of your current bankroll, which in this case is 2000*0.083=166 units. When using the KC, ALWAYS USE DIVIDER. Bigger the better. If you a damn good capper, formula/4 might be good, but I recommend formula/10 for most of the bettors.

 

[SixthSense]

I have the book and I understand the theory behind it but I disagree with it. I should also say the book is very, very good with numbers but I disagree with the KC theory.

I don't have the time to go into great detail right now but, in general, the problem with it is you better win at the percentage that you picked or you stand a chance of losing a lot more money than you would with flat betting.

There are better ways, IMO, to determine your betting percentage and flat bet and make money. Obviously, maximizing your winning potential and minimizing your losing potential is the goal.

I'll try and give more of my reasons for this later, when I get more time.

 

[couch_potatoes]

I use a formula derived from the Kelley criterion. But Sixth Sense makes an excellent point, namely your assumption about the overall win rate. It is a crucial factor.

If you can't win 55 - 60% of bets that pay near even money, you're going to lose your bankroll if your wagers are too large.

The trouble with gambling is the losing streak. Even if you make bets that win 6 out of 10 times, you will frequently encounter losing streaks of 5 and 6 in a row. You must have enough capital to make it through the streak. If you are betting even a quarter of your bankroll, you will wash-out during that losing streak. The question is, what fraction is optimal?

My solution is to bet a smaller fraction. Whether it's a formula, or a rule-of-thumb such as 5%, you will benefit greatly by keeping emotion out of the equation, and use logic/discipline instead.

============
Here's an example of the formula I am using:

A= % of wins (expressed as a decimal) --> This is your EXPECTED overall win rate.

B = Average profit per game won
C = Average loss per game lost
D = Size of your bankroll today
E = Percentage of bankroll to risk on 1 game

For example, estimating that you can win 55% of the time, A = 0.55

Assuming you make $10 profit for every $11 wagered when you win, B = $10

Again, assuming you lose the amount wagered ($11) whenever your team loses, C = $11

Say your bankroll is $500. D = $500

Here is the math:

E = A - [(1-A)/(B/C)] <----- Formula

1-A = (1 - 0.55) = 0.45
B/C = ($10/$11) = 0.91
E = 0.55 - (0.45 / 0.91), or
E= 0.55 - 0.4945, or 0.05, or 5%

Since D = $500, the amount to risk = 5% x $500 = $25

 

[winbet]

Sorry to put a dampner on this lads, but over the years(many)I have seen plenty of staking plans. The problem with all of them, as couch potatoes pointed out(I think) is you do not know when you are going to win and avoiding the losing run. The best pros I have read about always bet level stakes, but they seemed to have a bottomless pit when it came to their bank. What I do nowadays is start the season (whatever sport) betting as careful as possible (considering the down side totally) using 20% of my bank. If I lose 5 straight I call it a day, I can afford to take the loss and you have to question your judgement. If I happen to get off to a flyer I use the bookies money to really have a go, pressing up all the time Im winning. Normally what happens is somewhere in between, but the main thing is be very careful with your money and go for the jugular with theirs.

 

[nepenthe]

I used to follow the Kelly Criterion and I am sure it IS the optimal betting method provided you can win at least 55% of your near-even-money bets over the long haul. In an ideal world, I would structure each of my bets to conform to 1% of my current bankroll, but this is impossible when I have to bet a bunch of games at a single time. The variance is also greater using this method. Furthermore, the emotional aspect of it is what really got me here. Nothing is more frustrating than having to decrease wagers while on a losing streak, get on a winning streak and still end up at a big net loss despite going, say, 6-5.

What I do nowadays is flatbet 1.5% of my bankroll on every bet, regardless of perceived strength. When my bankroll increases by 50% or so, then and only then will I increase my bet size. My bet size stays the same even through my losing streaks, because I am confident I will eventually make it back. This method may not be optimal, but it gives me a little additional peace of mind I didn't have with Kelly.

 

[~Falco~]

Just some common sense writing. Check it out.

http://professionalgambler.com/debunking.html

Not schilling just sharing

 

[winbet]

I cannot reiterate enough that using staking plans will preoccupy your thinking. Betting level stakes will focus your attention to the important part, selections. The bottom line is always, can I win, if you cannot win at level stakes then you are deceiving yourself and getting into maths which prove that you could have won.

 

[crismasty]

it doesnt matter if your lose 5 or 6 in a row because you have decreased you bet upon losing the first game to your smallest unit ..
wehn you win and increase your bet you are allowing to catch a streak and bet more when you are winning ie as in most progressive betting systems in blackjack or sports..
if you do no bet more when you are winning you will be punished by vigorish or the fact that you lose the same amut as when you win ,
the percentage of bankroll system is what most professional bettors use excepting the fact that stronger plays obviously will have larger bets made on them but never more than a certain % of br unless your talking about the highest level of professional that will fire it all when he thinks he has the nuts..

 

[SixthSense]

crismasty -

I'm not sure what most "professional" bettors do but the percentage of bankroll method is the worse option (between that and flat betting). If you do the math, you will see the numbers and will just about always end up on the short end of it. If you hit a percentage of 59% or higher, you have a chance to come out ahead, with the percentage of bankroll. Otherwise, you are simply throwing money away.