Proving Computer Football Predictions Are Good!

We get many comments from people who have quite obviously pre-judged our website and written-off the value of our computerised predictions, although the comments are usually disguised as rhetorical questions. Typical of the comments we get are such things as "How can a computer possibly predict the outcome of a football match?" and "How come your computer program said last week that such-and-such team would win 3-0 when it lost 2-3?" And perhaps you think exactly the same way.

First, though, ask yourself whether or not you think that it should be possible to get reasonably close to deciding if a team ought to win a match (or otherwise) if you did enough background research. If you don’t think it is possible, then let me turn the tables and ask you a rhetorical question: How do you think the Bookies can set the Odds in order to make themselves rich if they aren't able to predict the probable match outcomes correctly, often enough? Yes, once you have been forced to think about it properly, it is perfectly obvious that the Bookies must employ a worthwhile system of some sort, otherwise they wouldn't be able to stay in business for long. So what we did was spend a considerable amount of time, money and effort, over the course of very many years, to try to emulate what the Bookies do – by devising a valid basis for predicting what the most likely outcome would be of any League-based football match.

We could bury you in prediction theory here, and bore you to tears about how our computer program works. But we won't do that. Instead, we'll just jump to the facts. Firstly, overall (and which is very easy for you to check via the interactive tools you can find on our website), the Bookies expectations and our posted predictions have about the same chance of success (although the Bookies have the advantage of the Over-Round (mark-up) to cushion them when things don't go as expected, and which happens more times than you might imagine, although usually not through any fault of theirs). Secondly, we have gone the extra mile to invent a tool (SoccerPAT) that allows you (the visitors to our website - all of you) to analyse what happened to our predictions, right down to the individual team level if you so want. You would not be able to do that without a computer program to compile the vast array of databases that are necessary.

Thirdly, before you select a team to bet on, you can get an initial clue as to the team's likely performance just by analysing the Odds being offered by the Bookies, plus you have your own brain to rely on (but only if you can be bothered to do the necessary research work, of course). So, that means that nobody has to rely blindly on what a computer prediction says about the outcome of any particular match. But where a website such as ours supplies you with all the checkable background data showing precisely what lies behind its computer's predictions, then using your own brain in conjunction with the computer's output will surely make it much easier for you to make more valid betting selections than if you were relying on just your own intuition alone, even if backed up by the Bookies' expectations.

But then we come to the fourth and final point: the proof of the pudding is in the eating. So go take a look at the SoccerPAT output for Week 3 of the European soccer season. You can do that by clicking on the big red SoccerPAT icon on the Home screen or by clicking this link to see the screen shot. You will see 6 matches listed where our computer program predicted a Home Team win with a score-line of 2-0, and the Bookies too thought 4 of them would be Home Wins. You will also see that 3 of them did actually finish with 2-0 as their score-line. And, overall, the return was 49% on the singles "win" stake. But now you ask, so what about the overall situation if we had bet on the exact same parameters for Weeks 1, 2 and 3? Well, all you have to do is click on the "Summary for Season" link on the SoccerPAT Week 3 screen to see the answer: 3 straight weeks of wins, where betting on the Doubles would have given you a return of 65.35% profit (again, you can also see the screen shot here).

So, what was your question again? How can a computer possibly predict the outcome of a football match? I think we have answered that, and also proved that the pudding is very tasty! The SoccerPAT facility could help you find lots of excellent betting opportunities if you are prepared to open your mind to the possibility that computerised soccer predictions might be a good thing!