So the producers of Soccer AM, the cult television programme, have no reason to celebrate acquiring my support. Their campaign is to establish the word “bouncebackability” firmly in the English language. The word was first used by Iain Dowie, the Crystal Palace manager, after his team had equalised against Arsenal.

I think it is a fine addition to the English language, albeit that Dowie embedded it in a sentence that bore only a passing resemblance to English — “Crystal Palace have shown great bouncebackability against their opponents to really be back in this game,” he said.

The moment I heard of the campaign, I knew what the Fink Tank’s contribution could be. I recruited Dr Henry Stott and Dr Alex Morton to the cause and we agreed that we would set about examining whether there really was such a thing as bouncebackability.

I have to admit that at the beginning I was worried. Arsenal scored against Palace in the 63rd minute and Aki Riihilahti replied only two minutes later. If what Dowie meant by bouncebackability is a version of the myth that teams are most vulnerable just after they’ve scored, then I had bad news for the Soccer AM team. There is no truth in this idea, nor its counterpart that some teams are at their most dangerous immediately after they concede.

Fink Tank research demonstrates that equalising goals are evenly spread across the time remaining in a game after one side takes the lead. Yet this finding does not blow bouncebackability out of the water. Dowie would still be right to identify it as a phenomenon if it is true that some teams show a greater propensity to score once they have conceded than they do in normal circumstances.

It is important to understand that the chances of a goal being scored rise as the match progresses. In each five-minute period, a goal is more likely than in the preceding five minutes. This is one of the reasons that I get irritated with people who leave with five minutes left. They are not only obstructing my view of the game’s denouement, they are also departing just when a goal is most probable.

One cannot, then, demonstrate that bouncebackability exists simply by showing that teams are more likely to score at some point after they have conceded. This may just reflect that the game was progressing and goals were becoming more likely anyway. Stott and Morton had to adjust for the timing of goals in order to establish whether some teams show the ability to bounce back.

My gift to the Soccer AM campaign is this — I can confirm that science supports Dowie’s intuition. Bouncebackability exists.

It is, of course, possible that it wasn’t intuition at all and that Dowie had read A Birth Process Model For Association Football Matches, a fine paper written by Mark Dixon and Michael Robinson. This was published in 1998 in The Statistician, volume 47, and basically reached the same conclusion.

Teams have, on average and adjusting for time, a 21 per cent greater scoring rate (number of goals scored per unit time) after they have conceded than when they are on level terms.

There is also a big variation in bouncebackability between sides, allowing the Fink Tank, using data for the past two years, to produce the world’s first bouncebackability league table. Arsenal come top, and by some distance. Their scoring rate goes up by 118 per cent when they fall behind. Portsmouth and Manchester United are among those with negative bouncebackability as their scoring rate actually drops when they concede.

And Palace? It’s a little early to say because they are newly promoted, but thus far they do no better than average. So Dowie was right that there was bouncebackability on display in the game against Arsenal. He just had the wrong team.

The Retaliators