Sunday 5 October 2008

Sunday Value

Am I missing something, or are West Ham terrific value at 1.91 on BETDAQ to beat Bolton Wanderers later today?

I have been maintaining a rating system from the start of the season based on Elo principles, and now that October is here, and it seems to have settled down somewhat, I'm going to start using it to find value bets.

I currently have Bolton rated as the worst team in the Premier League with West Ham rated 8th, but some 58.3% higher than Bolton. By comparison Chelsea are 'only' ranked 21.87% higher than Aston Villa and are priced at 1.6.

Liverpool also look to be good value to win at Manchester City, (2.28 on BETDAQ, 2.24 on Betfair right now).

Small stakes on these until I gain some confidence that the ratings are mature enough to start using live, and all bets will be placed on BETDAQ since I will usually let these run as pure punts and 2% commission versus 20% is really no contest.

Update: I guess I was missing something in regard to the West Ham price, but overall in profit with Liverpool's come from behind win.


Anonymous said...

Interesting. I also had a decent bet on West Ham although I took 11-10. I think it was a good bet irresp3ective. Not so sure on the Liverpool bet though. I think they may be overrated and possibly got lucky today.

Cassini said...

11/10 was certainly excellent value. I was pleased to get my 1.91. Not all 'value' bets win though. Liverpool did get a little lucky with the sending-off when 1-2 down. I took that bet because they looked value when I compared their rating with Man City's - also probably overrated. Thanks for the comment, and good luck.

Anonymous said...

Cassini, you mentioned that your current system on football is based on ELO prinicples. I hope you're aware that the accuracy only comes into place after around 30 matches (for each team) have been rated. Apologies if you are already aware of this, plus you may be doing an adapted/hybrid version which has different characteristics. Just thought I better mention it just in case. (BF forum name: JPG)

Cassini said...

Thanks for the comment JPG - much appreciated. I wasn't aware that it was as many as 30 matches, but I was aware that it would probably be the new year before it would really start to (hopefully) pay its way. Last week was the first bet using it and I was just 'dabbling' when I saw a big discrepancy, but as shown by the West Ham result, and indeed the St Mirren v Rangers result, there's nothing certain in football.

Anonymous said...

No worries mate. Believe it or not, after only 9 or 10 matches, ELO is apparently around 75% "accurate" already, the remaining 20 0dd matches make it about 98-99% accurate. If you think its worthwhile though, you can start one or more seasons back, update from that point and get a lot more data together. A lot of people think that results 3 years ago dont matter. They do - but only to a lot lesser extent. This is something ELO incorporates automatically though.