The Secret to Sports Betting – Part 4

The Secret to Sports Betting – Part 4

Martingale and the Kelly Criterion.
It is time for the last part of the secret to sports betting. In the secret to sports betting part one, we talked about the importance of having a separate gambling budget so that the betting does not go beyond the household money. In addition, we began to accept that all matches can end with home win, draw or away win and that you must never bet with your heart. Finally, I warned against combining many objects on one coupon and made a strong recommendation to always bet where you get the best odds.
In the secret to sports betting part two, I discussed having knowledge of what you are betting on as well as specializing in something, not listening to “experts” who cannot show results, taking breaks in betting when necessary and the importance of keeping accounts.
In the secret to sports betting part three we looked at how to find the matches you want to bet on before this week we put 
these games into a system.

The 10 basic rules

1) Only bet for money you can afford to lose 
2) Realize that all matches can end with home, draw or away win 
3) Place your bets where you get the best offers 
4) Don’t make judgement based on your team preferences 
5) The fewer games on each coupon, the better 
6) Have knowledge about what you are betting on 
7) Only listen to experts who can prove they are worth listening to 
8) It is allowed to take a break 
9) Specialize in something not everyone knows everything about 
10) Do the accounting on all bets placed 


It is easiest to start with the system that no one ever should use! Every now and then a “new 
system” comes up that someone has found. Mostly, it always involves a variation of the 
Martingale that has been known for over 100 years. The strategy is simple: You find 
something that gives 2.00 in odds and place your bet. If you lose, you double your bet until 
you win. Then you go back to the starting bet. Thus, you will win back everything you have 
lost + the starting profit, when you finally hit. The problem naturally comes when the losing 
streak becomes too long. In the book “Fixed Odds Sports Betting: Statistical Forecasting and 
Risk Management” by Joseph Buchdahl from 2003, advanced Martingale calculations are 
made which show that during 1000 bets at odds of 2.00 you will have a losing streak of ten atleast once . And with a value of at least five percent on all games, the risk of bankruptcy is 38 percent! With an initial investment of 500 euro, the account looks like this:

Game 1: 500 at 2.00 in odds = loss 
Game 2: 1,000 at 2.00 in odds = loss 
Game 3: 2,000 at 2.00 in odds = loss 
Game 4: 4,000 at 2.00 in odds = loss 
Game 5: 8,000 at 2.00 in odds = loss 
Game 6: 16,000 at 2.00 in odds = loss 
Game 7: 32,000 at 2.00 in odds = loss 
Game 8: 64,000 at 2.00 in odds = loss 
Game 9: 128,000 at 2.00 in odds = loss 
Game 10: 256,000 at 2.00 in odds = loss 

You have now bet a total of 511,500 and have received zero. The system says that on game 11 
you should bet 512,000 to get back all the losses plus the five hundred you wanted to win in 
the first place. Tempting?

Martingale is also called the “Monte Carlo system”. It comes from the fact that at a Casino in 
1913, 26 times in a row the roulette showed black. The players started betting massively on 
red along the way because “it must be red soon”. Of course, it didn’t have to. The chance that 
there will be a head on the next coin toss is 50 percent, regardless of how many times there 
have been tail in a row.

Another example of this flawed psychology of many players is the “Monty Hall problem”. 
There are three doors. One has a new car behind him, the other two have an orange. You 
choose door A. The host opens door C and shows an orange. You will then be asked if you 
want to switch to door B or stand on door A. Then you must always switch to door B! The 
math is simple:

Before starting:

Door A: 33.33 percent chance of car 
Door B: 33.33 percent chance of a car 
Door C: 33.33 percent chance of car 

You choose door A. There is then a 33.33 percent chance of a car and a 66.66 percent chance 
of an orange.

The hosts, knowing where the car is, open a door with an orange. It doesn’t change your 
chance that you hit from the start. There is still a 66.66 percent chance that you missed the 
initial selection! 

The Kelly Criterion

Now that the Martingale is hopefully buried once and for all, we can look at the Kelly 
Criterion. Here, you are expected to play with value in your bets, as we talked about in Part 3. 
The Kelly Criterion uses the expected value to tell you how much to bet.

The formula for finding the stake is:

(BP – Q) / B 
B = the odds minus 1.00 
P = chance of hit 
Q = chance of miss

Say we give Tottenham a 55 percent chance of winning at home against Arsenal and the odds 
are 2.00. Then the formula becomes:

B = 1.00 
P = 0.55 
Q = 0.45 
(1.00*0.55-0.45)/1.00 = 0.10 

You must therefore bet ten percent of your fund value on this game. In our eyes, this is far too 
aggressive. There are therefore some calculators that recommend setting 50 percent or 25 
percent of what the Kelly criterion recommends. Then you are on a somewhat healthier 
management of your gambling money, but using the Kelly criterion requires that your 
percentage setting matches 100 percent. We think it is demanding enough to find value in 
odds as told in part three. Using it also for determining how big the bet should be is maybe a 
bit too much. 

Our model

Our model is thus the one we recommend. Here we settle the accounts once a week and 
always put two percent of the fund value on our games. Of course, we could have the 
potential to earn a lot more with more aggressive percentage bets, but our whole goal is to 
have a healthy relationship with betting. We set aside the maximum amount we want to lose 
during an entire calendar year and manage our stakes accordingly. If we made a profit last 
week, the bet goes up, if we made a loss, the bet goes down. We don’t seek to win back losses 
quickly – we have an overall strategy where we know that as long as we percentage matches 
well, we will always end up in profit!