Anytime Goalscorers

[Wiltshire Shot]

I've been working on a system that looks at making a profit from the Anytime Goalscorer market. It assumes that stats can be used to predict how regularly someone will hit the net, then compared with bookmaker prices to see where the bookies have got it wrong. For example, if I reckon a player has a 60% chance of scoring but the bookies are offering evens, then I should have a 10% chance of making a profit. I trialled an early version for the last few months of last sesson and came away with a very, very small profit (I'd have been better off keeping my money in a low interest bank account!) but I reckon that I've improved the staking systen now, so it's a 'real life' system, but with very low stakes at this stage. I'm looking at data for the top five dive divisions in England, plus the top division in each of Germany, Holland, Spain, Italy and France, starting with ten units in each - 100 in total. I don't have time to analyse every fixture and every player, so I'm only looking at teams that are more likely to score - top half sides at home to bottom half sides. Time reasons might also mean I have to drop some of the more profitable leagues. I assume concentrating on the lower leagues will be more profitable as the bookies don't know them as well - but in last season's test for some reason it was Holland that came out top! In those teams, I look for players who score 40% or more. I award one goal for each home game they score in, plus half a goal for a second or sebsequent goal in each game (so two goals in a game would be 1.5, a hat trick 2, etc). This is so that stats aren't skewed by a player hitting a hatful against the bottom club, for example. So, a player with six goals in ten games (home only) scores 60%, so the stats make me assume he has a 60% chance of scoring in his next game. As I'm only looking at games against clubs in the bottom half, the actual chance of scoring is probably slightly greater. The stats obviousy become more reliable after more games, so I don't look at the system until a club has played five home games. So, if our player with a 60% chance of scoring is offered at evens (50%) by bookies, I have a 10% advantage, and I'll put 10x0.1 units on. I only put money on when I'm up by 10% or more, as I want to stay away from borderline value. At least, that's the theory... Last season's testing showed the main problem to be a player starting on the bench and then getting a couple of minutes at the end (can be overcome by waiting for team news - but not always practical as I want to get to matches!). I also need to come up with some sort of weighting for the number of likely goals in a game (if the game is likely to be under 2.5 goals, then I'd be better off betting on 1st goalscorer rather than anytime scorer.

[Wiltshire Shot]

Re: Anytime Goalscorers Most Blue Square Premier teams have now played five home games. The system turns up two players who fit all criteria for Saturday's matches: * Playing at home * Top half v bottom half * (Statistically) undervalued by bookies * Scoring rate of 40% or more Holroyd (Cambridge, v Wrexham) - 3 goals in 5 games = 60% Odds 11/10 (Bet365) = 48% So stake is 1.2 units. Barnes-Homer (Kidderminster, v Gateshead) - 3 goals in 5 games =60% Odds 5/4 (Bet365) = 44% So stake is 1.6 units I particularly like Barnes-Homer, but I'm not so sure about Holroyd as Wrexham don't concede many, but I'll go with it. Further into the season with more data I'd consider 1st scorer, but I think it's too early. Constable (for Oxford v Eastbourne) scored 4%; not enough to meet my 10% criteria, while Green (also Oxford) just missed out on sonsideration.