A team of academics in Salford Business School has developed a machine which can, among other things, predict sports results. And they are currently testing their creation against one of the top names in UK football.
SAM, the Sport Analytics Machine, has been developed by statisticians using a series of algorithms, and is able to forecast the results of a variety of sporting fixtures.
Developed by Professor Ian McHale and Tarak Kharrat, the machine takes into account multiple factors to come up with its predictions, such as recent results, which team is at home, the strength of teams that both sides have played as well as the quality and the form of the players on the pitch.
And now the machine is pitting its predictive powers against ex-Liverpool and Republic of Ireland player Mark Lawrenson, who every week on the BBC website forecasts the results of the weekend fixtures, usually against a celebrity, to see who comes out on top. SAM has joined the fray and is adding its skills to the mix every week. It is currently the most successful predictor in the league with an average of 110 points so far.
Currently the machine is being used to predict top flight football results in England, but it has scores from all of Europe’s top leagues from the past seven years and could also be used to predict results from other leagues. In addition, SAM can lend its hand to predictions in other sports such as cricket, golf, rugby or tennis.
Ian McHale, Professor of Sports Business, said: “Data is the fuel of our system. It is entirely done using numbers, there is nothing subjective about the results that SAM comes up with.
“We have developed this over the last three years and it’s looking good at the moment, our tests have shown that we will probably outperform the expert in the long run and we’re currently top of all the people who have predicted results.”
The statistical model developed estimates probabilities of each score line occurring in any match between two given teams at a given location. It gives probabilities of each result, score line and number of goals in each game. It can even be used to predict the times of goals.
Professor McHale explained: “There are several ways to go about making predictions for the results of a football match. In fact, lots of academics and people in the gambling industry have tried. The thing all of the methods have in common is that they look to the past to get an idea of what can be expected in the future.
“For example, if a team is in the habit of scoring lots of goals in recent matches, it is reasonable to think that that team might score more than the average number of goals in the next match. Conversely, if a team is having a run of conceding lots of goals, it might be expected to concede more than the average in the upcoming game. Of course, more recent matches reveal more relevant information than matches a long time ago, and so that is taken into account.
Tarak Kharrat added: “There are two additional factors from past results that can be used to signal what the future holds.
“First, it matters where the team has been playing: at home, or away. This effect is known as home advantage, and it is pretty easy to detect and measure. In fact, if you do this over several years, it looks like home advantage is such that the home team scores about 1.2 times the number of goals the away teams score.
“Second, it matters which teams the team in question has been playing against. We assign each team an attack strength and a defence strength. The attack strength quantifies how good the team is at scoring goals, whilst the defence strength quantifies how good the team is at stopping opposing teams from scoring. These strengths are estimated from the data. It is this data that is fed into SAM in order to estimate the team strengths.”
http://www.bbc.co.uk/sport/0/football/35200491