|
Models Need A Face Lift
A few large professional soccer punters use computer models as the starting point for their asian handicap bets and, to a lesser extent due to lack of market liquidity, fixed odds match bets.
They are aided and abetted by over-paid and over-valued stats guys who talk a different language to the rest of us. For example factors which I know through experience and long-term profit influence the result of a match they dismiss out of hand as "noise".
These models produce home/away/draw percentages which are converted into asian handicap prices then adjusted for factors which I do agree with such as team news and team spirit. The adjusted prices are compared with the asian handicaps on offer with far east bookmakers (mainland China, Hong Kong and other exotic locations) and where the price available exceeds the pro punters price by the margin he requires (normally 5% ROI) he bets.
I can normally tell exactly which teams these punters will be backing by the time I have Friday night team news in the few leagues in which I operate. Where I know they will be backing a price I like I will make sure that I take it for myself as they are only able to bet in the size that they require on the day of the match and they do not want to give clues before that about where their money is destined to go.
Sometimes however knowing their selection process I realise that their model will strongly oppose the team I am waiting to back and I can sit back and wait until late on Saturday morning when the weight of their money will give me an exceptional price about my team which would have drifted 30 to 40 basis points. I shall only give one example here but last week an hour before match time I was able to obtain 2.419 Partick draw no bet to beat St Johnstone when I had them as 1.875 draw no bet favourites.
These models are far too long term in their assessment of team form in the leagues where I bet. To me the last weeks match statistics and reports are the most relevant information allied to team news. The previous two or three matches are the next most important. The previous season has little or no relevance. Models have no intelligence, emotion or human feelings and therefore, for example, disregard the importance of bogey teams. I can assure you having spoken to many players that they do not and teams they never seem to get a result against are a constant topic of discussion, affect their thinking and, too an extent, adversely affect them next time they play such a team.
Shrewd movements in asian handicap prices on any English or Scottish league can however throw up some excellent value fixed odds prices where these bookmakers have either failed to adjust their prices quickly to follow the asian handicap market or have been left alone by the pro punters due to the maximum size of the bet available being far too small to interest them. This is normally where updated team news or other factors not in the public domain are known only to the pro punter and his associates.
Using models to me is a far too rigid framework in which to operate to maximise profits. Successful betting is both an art and a science and whilst they have the latter in spades they are not even GCSE standard in the former. Your mind needs to be free, able to think laterally and walk the random walk.
When my eyes alight on the first list of prices for the weekend in the leagues in which I bet a I experience a frisson of excitement when I quickly spot over-priced teams. Jim Slater used to say in the late 1960's in his pomp that he just had to look at the share price pages of the FT to pick stock "winners" to invest in. It is a similar skill to one the ultimate pro gambler will strive for. |