This book is widely considered to be the definitive publication on football betting systems. I've tracked down a copy of it, as I believe it is worth seeing if the systems and analysis therein have stood the test of time, and whether they can bring something to the party so far as our own investments are concerned.
Paul Steele sets out fifteen systems in the book. He rates the complexity of these on a scale of one to three, with three being "computer necessary". From having studied his systems, the necessity of the computer is for historical number crunching. You could use a spreadsheet, but keeping all the data in an enterprise class database makes things so much simpler.
Another advantage we have is that all the historical odds data is available. I have read criticisms of Steele's selections as the available odds were not known at the time. For the analysis that will be undertaken, I will be using my "best available" odds - I have data from at least six bookmakers for every fixture, but which actual firm of bookies it is varies. This being the case I take the best price across the range that's available. After all, that is what we would do in practice.
Early in the book, Steele discusses the concept of "True Odds" - in which he sets out a formula for calculating the likelihood of Home, Away or Draw.
He gives this as:
He provides an example thus - if team "A" has won 4, drawn 1 and lost 3 at home and the opposing team "B" has won 1, drawn 4 and lost 3 away, then the formula looks like:
This equates to 43.75% and this, according to Steele, is the "True Odds". Bookmakers then add the over-round of 10-12%, so to our figure we should add 5.5% which is half the over-round range.
Let's take a look at the 2014-2015 season, and see whether the "True" odds are better than those on offer from the bookmakers, and, if so, what the outcome was. If the best available odds are more favourable than the Paul Steele "true odds" then in theory it would be advantageous to back these selections.
Steele recommends a betting season of October to March to allow for form to settle down at the start, and to minimise the effects of impending promotion or relegation at the other end. This being the case, we will limit our investigation to fixtures within this period.
I have performed this calculation for all fixtures within the date range above, below are listed the top ten or so biggest differences between the Paul Steele true odds and the best odds readily available in the marketplace:
I have added 1.055 to the values calculated, the 0.55 is the 5.5%, and the 1 is because we are comparing against decimal odds.
What jumps out at you is that whilst the bookies were offering seemingly mega generous odds on the home team, it was because they were utterly outclassed by the opposition. It bears further analysis but my initial take is that this means of assessing the true odds may be too blunt an instrument.
Jamie
Paul Steele sets out fifteen systems in the book. He rates the complexity of these on a scale of one to three, with three being "computer necessary". From having studied his systems, the necessity of the computer is for historical number crunching. You could use a spreadsheet, but keeping all the data in an enterprise class database makes things so much simpler.
Another advantage we have is that all the historical odds data is available. I have read criticisms of Steele's selections as the available odds were not known at the time. For the analysis that will be undertaken, I will be using my "best available" odds - I have data from at least six bookmakers for every fixture, but which actual firm of bookies it is varies. This being the case I take the best price across the range that's available. After all, that is what we would do in practice.
Early in the book, Steele discusses the concept of "True Odds" - in which he sets out a formula for calculating the likelihood of Home, Away or Draw.
He gives this as:
((Home Wins + Away Losses) x 100) / (Total Games)
He provides an example thus - if team "A" has won 4, drawn 1 and lost 3 at home and the opposing team "B" has won 1, drawn 4 and lost 3 away, then the formula looks like:
((Home Wins (4) + Away Losses (3) = 7) x 100 = 700) / Total Games (16)
This equates to 43.75% and this, according to Steele, is the "True Odds". Bookmakers then add the over-round of 10-12%, so to our figure we should add 5.5% which is half the over-round range.
Let's take a look at the 2014-2015 season, and see whether the "True" odds are better than those on offer from the bookmakers, and, if so, what the outcome was. If the best available odds are more favourable than the Paul Steele "true odds" then in theory it would be advantageous to back these selections.
Steele recommends a betting season of October to March to allow for form to settle down at the start, and to minimise the effects of impending promotion or relegation at the other end. This being the case, we will limit our investigation to fixtures within this period.
I have performed this calculation for all fixtures within the date range above, below are listed the top ten or so biggest differences between the Paul Steele true odds and the best odds readily available in the marketplace:
I have added 1.055 to the values calculated, the 0.55 is the 5.5%, and the 1 is because we are comparing against decimal odds.
What jumps out at you is that whilst the bookies were offering seemingly mega generous odds on the home team, it was because they were utterly outclassed by the opposition. It bears further analysis but my initial take is that this means of assessing the true odds may be too blunt an instrument.
Jamie
