A Spanish Tragedy

[Matthew]

All, In the same vein as my predict the premiership thread: http://www.punterslounge.com/forum/showthread.php?t=37165 I've done the same for the Spanish League (top flight only). Read my other thread for specific details of the approach and what goes into it. Simple recap: Loads of variables spanning 6 seasons are put into a statistical package which generates a probability for a HW, D or AW. Simply taking the most likely outcome is not enough to be profitable - I need to be selective - to do this I set hurdles that the probabilities have to get over. These differ for each outcome. Onto the interesting bit.............the cutoffs as they stand have only picked AW to date. The spanish top flite is a bit different to the english top flight in that more games end in a draw or aw. If you had blindly backed the AW at level stakes this year then you would be 8.45pts up based on a 1pt per game level staking strategy. A yield of 5.3% and a strike rate of 31.4%. Average AW odds in the spanish top flight season to date are 4.33 (min. 1.45 & max. 15). So I think there is a bit of value in the old AW in Spain. Season to date using the cutoffs described above and level staking 1pt per game things would look like this.

Stake	£75.00
Returm	£98.87
Profit	£23.87
SR	36.0%
Yield (%)	31.8%
HW Odds:	#DIV/0!
D Odds:	#DIV/0!
AW Odds:	4.54

That to me looks pretty healthy - although I owe a lot to Recreativo beating Madrid in their own back yard. Simple thing is - Madrid do this - yet the bookies refuse to adjust the odds. I saw Mallorca thrash the 3-0 at the bernabeu a few years back - although a little known player called Samuel E'to did all the damage that game. Wondered off a bit there. Lets see how I get on........ Next fixtures:

06/01/2007		Ath Madrid	V	Gimnastic
06/01/2007		Sociedad	V	Osasuna
06/01/2007		Zaragoza	V	Sevilla

[davidc]

Re: A Spanish Tradegy Hey plopplop - good luck with this. One possible problem that you may already have identified is the long odds winner. Have you back tested to see if the stats hold up by removing (e.g) the top 10% highest odds winners in order to try to even out any statistical anomaly?