I have always found Benn’s comment richly comic — an hilarious attempt to suggest that defeat was really victory, if you look carefully enough. But politicians are mere amateurs at this sort of stuff. In football, it has been raised to an art form.

Take Sir Alex Ferguson’s comment last week that, against Exeter, “Ronaldo could have scored five and Rooney could have scored a hat-trick”. Well, er, yes. Perhaps I could have scored, too, if I had not been a few hundred miles away writing an article on welfare reform. In fact, dammit, I could have bagged four (less speed than Ronaldo, better temperament than Rooney).

The point is, I did not. And neither did Rooney nor Ronaldo.

A whole series of what I call Compensation Statistics has been developed. You know the ones I mean, the stuff that runs across the screen on Match of the Day while a manager laments the way Keano missed a sitter and expresses disappointment with the referee.

The job of the Fink Tank, of course, is to assess whether these Compensation Statistics — shots on target, shots off target, corners, possession, yellow cards — should give any real comfort, or whether they are merely something to cling to. “We may have lost, but at least we had a lot of possession.”

Goals are scarce, rather random, events. It is easily possible for one team to do more of the sort of things that often result in goals, yet still lose. So it is entirely conceivable that statistics other than goal numbers will tell you that your team did well and provide hope that, although, unluckily, all that good play did not pay off this week, next week it will.

The question is, which stats to look at and which to ignore. Dr Henry Stott and Dr Alex Morton have been doing some calculations to find out.

Their favoured method has been to look at what, using a favourite statistician’s phrase, “explains the variation” in goals. It will not surprise followers of the Predictor that the stat that best indicates what the score will be is “number shots on target”. Indeed, this is so important that it has been incorporated into the model. Simply put, if two previously equal teams on a neutral ground have a goalless draw, in an immediate rematch the team who shot more on target in the first game would be the favourite.

Note the phrase “shoot on target” rather than “shoot at the target”. The number of shots off target is frequently cited, but our calculations show that it has no influence on the outcome of a match.

The favourite Match of the Day stat, the one that comes up first, is possession. And, indeed, the proportion of possession that your team have is positively and statistically significantly related to the goals you score. This is not particularly surprising. More eyebrow raising is that the role of possession, while positive, is actually very small. It explains only 1.1 per cent of the variation in goals. By contrast, shots on target explains 18.8 per cent.

Furthermore, possession is comfortably the most important of the remaining stats. Yellow cards for your opponents explains only 0.6 per cent of the variation. And, fascinatingly, the number of corners you take is almost irrelevant. The number of free kicks you take or give away hardly matters either.

In case this all seems a bit negative, we did come up with some more helpful findings. More attention should be given to passing. The pass rate — in other words, the proportion of passes that are successful — has almost twice the explanatory power of possession.

Narrowing this down even farther, the pass rate while playing in your opponents’ half explains 2.3 per cent of the variation in the number of goals you score. In other words, it is the accuracy of your play going forward that really tells you how the team is doing.

What was it they said at school? “Guys won’t ignore passes if they take statistics classes.” Or something.

finktank@thetimes.co.uk