Why do football pools systems work (in general)? Simply, because the performance of football teams is not in the long run random – teams do in fact play broadly to form, but they only work if they are based on sound statistical analysis.

There are two parts to solving the problem of how to win the football pools. The first is the performance analysis – the designer of the system should have looked at past results of teams and matches, and then produced a mathematical model which explains the results with a reasonable degree of accuracy. It should enable predictions to be made on the basis of the mathematical model.

Actually accuracy only needs to be slightly better than random. An edge of a few percent above random is adequate in the long run, but we would expect to do much better occasionally, and win. Of course, we want to win enough to cover the stakes on those weeks we did not win, so that it becomes a profitable exercise. But, there is another factor – a long run small edge is not enough, you have to have enother element in your plan if you are to reduce the odds to realistic levels some of the time (and by that we mean odds of about 3/1).

So, the second part in the football pools problem is what we call coverage. It is not economic to cover every possible combination on a coupon of 49 matches – there are 450 million ways of lining up any eight score draw results, if there are eight – there could be as few as three some weeks (thankfully rarely). That is where ‘plans’ and ‘perms’ come in.  These are ways of covering say 20 to 24 selections on a coupon in such a way that when there are maybe 12-15 score draw results on a given coupon, then there is a reasonable expectation of winnings if we have forecast say six or eight of those draws. Put simply, the plan may cover the 20 forecast selections in such a way that 25,000 possible 8 match combinations are staked. This reduces the odds significantly. Some plans are illustrated at football pools plans.

There are a number of systems available. Some are described in books available, others are downloadable online; there are spreadsheets available; there is software. There are books which specialise in ‘perms’ and software which also does so. So, there is quite a range of choice for those who are prepared to dig around. But, before you commit, check that published results for the system are available. One or two football pools system providers do publish results, and some have been know to offer free forecast trials. Predictions should be succeeding above 50% regularly, and as well as 80% on occasion.

However, the whole process is based on sound prediction of match results, and to do that (assuming that you cannot afford to regularly stake thousands of pounds or dollars), then, having bought a system, the user has to put in work to maintain the basic data on which the analysis is based. There are organisations which offer tips and there are comprehensive sets of match data and predictions available online (these are usually subscription only). If the football pools fan does not want to subscribe, then several hours work will be required to prepare each entry.

The final key aspect is that to be successful in this arena, then persistence is important – that one week that an entry is skipped may be the week that would have been the winner. We are, after all, working on a statistical basis.

In conclusion, football pools systems can work if they are well designed and properly used (and that means not allowing one’s views on a likely match outcome to affect the selections).

And, do not forget that the football pools season is year-round, as Australian football pools can be done during the northern summer! 

(c) 2010 Phil Marks