Graphic: A game of two halves, but which is better?

We’re nearly there. There are just enough Saturdays left to hand out some Fink Tank prizes for achievement and to pat ourselves on the back for beating the bookies again. But I also have time to squeeze in one more thing: the examination of another piece of prized football lore.

Is it really true that the title goes to the team best able to close out the season? Are champions made by a second-half surge, or is that just another football myth?

If it is, then it’s easy to see how the myth might be sustained. To start with, the title will be won by teams who close out the season in, er, first place by, well, winning a lot. This might look like efficient closing without that really being a good description.

And then there are the hard stats. It does seem as if the gap between the first and second place rises in the second half of the campaign. Again, that is what we would expect to see: the champions will extend their lead as more games are played, if they are the better team.

So, if it is a myth, it is easy to see how it got about. But Dr Henry Stott, Dr Ian Graham and Dr Mark Latham have been seeing what insights might arise from the appliance of science.

The first, and obvious, step was just to see what happens to the points of the champions over the season. Do they generally win more points in the first 19 games, or the second? Looking back at the past eight seasons, you can see from the graphic below that the pattern is that there is no pattern. In some seasons the champions achieved fewer points in the first half of the season than the second, in others more points.

End of story? Not quite. It is possible, after all, that the fixture programme distorted the outcome. It may be that in seasons where the eventual champions did better in the first half of the campaign — thus going against football lore — what really happened was that they had a particularly easy fixture programme. In other words, the teams they faced may have been playing poorly.

So what we need to examine is not only what did happen over a handful of seasons — which might be subject to distortion — but what might happen over 10,000 seasons, when such luck might be ironed out. And it is for such purposes that we have our computer model.

Using present team strength ratings and simulating 10,000 league seasons, you find that on only 44 per cent of occasions are the eventual champions ahead after 19 games. In other words, when a season is played out with teams of a given class, we should not expect the shape of the race to be sorted out immediately. It is more likely than not that the champions will not be ahead at the halfway point, even if the strength of every team is unchanging throughout the campaign.

The final piece of work was a further attempt to adjust for variations in the strength of the opposition. Each team’s actual points at every stage in the season were compared with the number of points we would have expected them to get if they were a side of average ability. Using this method, we were able to see that the relationship between reality and expectation did not change much as the season went on. In other words, season after season, the champions did not improve significantly in strength.

The idea that champions are good closers is a myth. Case closed.