@*t0m* I know nowt about football betting but I have an 'infallible' system for it

.

Back to the plot, for anyone still with me here is an explanation of the Kelly formula, where you can try to work out your 'advantage' from past bets:

**KELLY CRITERION STAKING STRATEGY**
I believe this is a staking plan used at for casino games of chance in America but it is also advocated for horseracing.

' A computer simulation covering 1000 simulations of a 700 race sample given a 30% win ratio at average odds of 2.75 to 1, after tax, gave the following results. Betting 4% of a 1000 point bank averaged 5054 and never at any time went broke.'

'Invest a percentage of your bank equal to your advantage over the bookmaker calculated as: -

Percentage Advantage = Win % -

__(Loss%) __N.B. The result gives the % for staking at Evens (Odds)

which then has to be divided by the odds of your selection to find the actual percentage to be staked.

To bet more than your advantage guarantees losses in the long run, but betting less than your advantage will win you less than you are capable of.

1000 repetitions of 700 race sample performance. Assumption 30% strike rate at odds of 11/4

Percentage……….Average Bank……….Largest Bank……..Bust

3……………………. 4991…………………. 4991……………..0

5……………………….407………………… 4075……………. 0

10……………………….7…………………….. 70……………. 0

Actual example over 50 repetitions

Strike rate 34% odds 5-2 bets 250

Starting bank 500 starting wager 3% average return over 250 bets = 2107 longest losing run 7

First bet was 3% of 500 = 15

After first 10 bets recalculate the advantage as a percentage and stake that amount. In future calculate the advantage after every runner. If the strike rate drops to 30% the average return drops to 852. If the strike rate rises to 40% the average return rises to 5326. Thus the staking system is sensitive to strike rate which rests with the backer'.

'The fact that the original percentage applies when dealing with Even money chances seems to be understood as can be seen from the examples. With a win percentage of 34 and SP5-2 the stake is 3% of the bank. Th formula would give: -

34% - 66% divided by 5-2 = 34% - 26.4% = 7.6%. The 3% is correct being the 7.6% divided by SP.

However, this is not the end of the story. The formula is correct if we know accurately the percentage advantage over the bookmaker and this is impossible to obtain from samples. Further research indicates that the correct percentage to stake is lower than the formula indicates. Between half or three quarters of that amount. Three quarters of the amount indicated is a reasonable compromise'.

Comments

Although the underlying principle seems sound it is very tedious doing a full recalculation by odds after each bet and even worse deflects concentration away from the selection process. This is a suggested short cut to employ.

After a minimum of 10 results note the strike-rate percentage and stake according to it in the following manner. Thereafter recalculate your wins to runs percentage after each bet. It is assumed average odds bet of about 5/2.

Strike-Rate

20 – 25% bet 2% of the available bank

26 – 30% bet 3% of the bank

31 – 35% bet 4% of the bank

36 – 45% bet 5% of the bank

46 – 55% bet 6% of the bank

56 – 65% bet 7% of the bank

Above 75% bet no more than 8% of the available bank

This is similar to staking operations in the past but where a betting return as a percentage of the bank was required as opposed to betting a percentage of the bank.