Football - Risk adjusted return method

[dalkent]

I’ve been running a football thread for a while now and found that I am struggling to get bets on some of the leagues using the criteria i chose. I have decided to try a risk adjusted return approach and incorporate Asian Handicap ratings also. To that effect I have upgraded the models to produce handicap probabilities and therefore calculated value. Unfortunately I cannot back score the model as I don’t have history of Asian handicap odds. Therefore I am going to run a test on here independent of my normal thread and approach. Once I have done 100 bets (or the bank has depleted too much) I will decide whether to pursue it and upload it to the website. Here are the my first weekends predictions. I will have a starting bank of £100 and bet 3% per bet with bank updates every £20 (i.e. will stick with £3 bets unless above or below 120 and 80) :hope League 1 Carlisle to win £3 at 3.12 Colchester +0 (DNB) £3 2.86 Leyton Orient +0.5 £3 at 1.775 Tranmere +0.5 £3 at 1.685 League 2 Dag+Red +1 £3 at 1.24 Torquay +0.5 £3 at 1.847 York +0 (DNB) £3 at 1.794

[dalkent]

Re: Football - Risk adjusted return method Some more for tonight. Last week was quite poor Bradford -1 £3 at 2.64 Bristol Rvs. +.5 £3 at 1.925 Oxford -0.5 £3 at 2.19 Carlisle -1 £3 at 3.2 Colchester+1.5 £3 at 1.2 Gillingham+1.5 £3 at 1.14 Port Vale +.25 £3 at 1.862 Shrewsbury w 3 at 2.8 Torquay w £3 at 2.88 Accrington H+.25 £3 at 2.19

[dalkent]

Re: Football - Risk adjusted return method I was a bit late posting tonight because of delayed trains all over London. Anyway spent a lot of time trying to derive the standard deviation of a split asian handicap bet. Couldn't find it anywhere on the web. Anybody else done this before or seen it anywhere?

[dalkent]

Re: Football - Risk adjusted return method A few for the weekend. Still struggling for positive results. [TABLE=width: 311]

[TR] [TD]Newport +0.25[/TD] [TD=align: right]£3.00[/TD] [TD]at[/TD] [TD=align: right]1.83[/TD] [/TR] [TR] [TD]Crewe +0.25[/TD] [TD=align: right]£3.00[/TD] [TD]at[/TD] [TD=align: right]1.99[/TD] [/TR] [TR] [TD]Leyton Orient -0.5[/TD] [TD=align: right]£3.00[/TD] [TD]at[/TD] [TD=align: right]1.98[/TD] [/TR] [TR] [TD]Gillingham +1.5[/TD] [TD=align: right]£3.00[/TD] [TD]at[/TD] [TD=align: right]1.475[/TD] [/TR] [TR] [TD]Shrewsbury +0[/TD] [TD=align: right]£3.00[/TD] [TD]at[/TD] [TD=align: right]2.64[/TD] [/TR] [TR] [TD]Notts County +1[/TD] [TD=align: right]£3.00[/TD] [TD]at[/TD] [TD=align: right]1.523[/TD] [/TR] [TR] [TD]Port Vale +1.5[/TD] [TD=align: right]£3.00[/TD] [TD]at[/TD] [TD=align: right]1.375[/TD] [/TR] [TR] [TD]Reading +0.5[/TD] [TD=align: right]£3.00[/TD] [TD]at[/TD] [TD=align: right]1.862[/TD] [/TR] [TR] [TD]Leeds +0.5[/TD] [TD=align: right]£3.00[/TD] [TD]at[/TD] [TD=align: right]1.752[/TD] [/TR] [TR] [TD]Blackpool Win[/TD] [TD=align: right]£3.00[/TD] [TD]at [/TD] [TD=align: right]2.15[/TD] [/TR] [/TABLE]

[Areme]

Re: Football - Risk adjusted return method Dalkent, could you expand a bit on your risk adjusted return method?

[dalkent]

Re: Football - Risk adjusted return method It's all about risk adjusting the expected returns by dividing expected return by the standard deviation of returns. I will use an example to explain where i compare two bets. Assume two matches one with odds of 1.5 and the other 10. I have calcualted the corresponding probabilities to be 0.73 and 0.11. Expected values of each bet are Match 1- 0.73*(1.5-1)-(1-0.73)=10% Match 2- 0.11*(10-1)-(1-0.11) =10% So they both have exactly the same expected returns Now i risk adjust by dividing each by the standard deviation of returns. Match 1 -10%/63% = 14.3% Match 2 - 10%/313% = 3.2% So now clearly Match 1 is much better since the chances of winning each time you bet are higher. Sorry about the briefness of explanation. I can provide more later if you wish.