I’ve been tinkering with the algorithm model over the past week or so.  The result is lower risk and higher yield - the perfect combination.  This is of course based on the results output from the new algorithm.  In some way this goes without saying but I think it’s always good to remember this point.  If the model’s wrong then performance will be worse than predicted.  How much worse?  The trouble is you cannot measure what you don’t know!

All betting tips from here on will use this new algorithm.  The main addition has been to the stake calculation.  The stakes are now adjusted based the amount of value on offer whereas previously all stakes were calculated using the same percentage of the fund.  This is something which I have always strived for but the results were too variable up to now.  It’s really a minimum requirement for a value model to bet according to the amount of value.  Put simply, if Mars bars were a range of prices you would start by buying the cheapest on offer, not be indifferent between them - even when you would still buy the most expensive one if it was the only one on offer.

The amount of the bet we buy (the stake) varies according to the value on offer (a function of the price).  For example, if the stake % is set at maximum of 5% of the Fund and Chelsea’s market odds on offer are showing 50% value versus the algorithm odds; previously the stake would have been: Fund value x 5%; whereas now the stake size is:Fund value x 5% x 50%.

Moving on…

The following are today’s Premier league betting tips using odds from WBX and Betfair.  Remember WBX has a lower commission of just 3%.  Teams are only layed in this new model due to the model’s ability to better forecast teams not winning rather than winning giving us a better yield.  There is quite a wide variance in market odds versus algorithm odds in today’s tips but that is no more than coincidence; previous games have not shown the same variance.  Nor is the algorithm only laying away teams.

Arsenal v Newcastle. With Newcastle at market odds of 15 this is the sort of game which can hit the fund hard.  A loss here would be roughly 40% of the fund.  There is no understating the significance of this.  However, the algorithm odds of a Newcastle win are 50 which therefore measures the risk as low, around 2%.  The theory is that the Fund earns sufficient income during the period between these large losses (when they do occur) to outweigh the loss when it comes.  As long as the algorithm odds are more accurate than the market we win.  If not, then we lose, simple as that.  Interestingly the old algorithm stake was lower at around £7.

Arsenal beat Newcastle just this weekend 3-0 in the F.A. Cup. Whilst Newcastle may have learned something they are unlikely to produce a good enough performance to win which is now the issue for the Fund.

Bolton v Fulham. Neither team has performed well in recent matches.  Bolton have lost Anelka to Chelsea which is of concern but again we want to avoid the loss not necessarily win.

Middlesbrough v Wigan. There’s not a wide gap between all of these teams and the algorithm odds look too heavily weighted against Wigan.  As it true for Fulham.

Sunderland v Birmingham. Same as above but it’s all coincidence I assure you!

HOME AB Odds Ex Odds Value Stake AWAY AB Odds Ex Odds Value Stake
Arsenal 1.12 1.32 18%  £-    Newcastle 50.35 15.00 -70%  £9.79
Bolton 1.16 1.95 68%  £-    Fulham 29.41 4.80 -84%  £12.41
Middlesbrough 1.20 2.10 74%  £-    Wigan 20.52 4.30 -79%  £12.20
Sunderland 1.18 2.24 89%  £-    Birmingham 23.72 3.70 -84%  £13.52