THE first time I saw Robert Huth play, he scored. And what a goal it was. From miles outside the box the ball cannoned into the net. From that day, every time Huth got the ball in an attacking position, the crowd shouted “shoot”. He frequently followed their advice. I never saw him score again.

The advice Huth was being given was rarely sensible. If you shoot from miles away you are unlikely to score. Psychological tests conducted in the 1960s, and recorded by Ken Bray in How To Score, show that professionals overestimate the chances of scoring from long distances. Huth needed no encouragement.

But you may have noticed a weasel word. I did not say that that the advice was not sensible. I said that it was “rarely” sensible. The reason? Every decision carries a cost. The best place to shoot, obviously, is right on the goalline.

Unfortunately you have to get the ball there first. And every time you pass or for every second you keep the ball, there is a chance of losing it.

So the decision about shooting from outside the area is a trade-off.

I have been thinking about this question because we have been working on improvements to the Fink Tank model. Not possible, you cry! Well, Dr Ian Graham and Dr Henry Stott have a hunch that it is and the hunch is connected with shooting.

At the moment, match probabilities are based on a weighted calculation involving not only goals but also shots on target. But the shots on target made by teams and conceded by them differ not just in number but also in type.

Some clubs may shoot all the time from miles out, other teams may do so rarely. Some clubs may let sides into their box pretty readily, but are very good at keeping those close shots out of the net.

The profiles of sides may have an impact on the outcomes of games. Thus a side great at scoring from far out may do especially well against a side particularly bad at keeping those sorts of goals out. Our present model does not account for such aspects. We intend a new one to do so.

Along the way, we are finding useful data on clubs. For instance, adjusting for the quality of opposition, Manchester United shoot at goal more often than any side in the league. They are, however, spectacularly bad at getting those shots on target. In a ranking of conversion rates (the proportion of their shots that are on target) they come well down in 39th, just behind Bury.

Chelsea are fairly poor at generating shots (10th-ranked), poor at getting them on target (12th) but good at getting those that are on target into the net (3rd). Arsenal are excellent at all three, which is why Fink Tank rates them more highly than most pundits.

Liverpool? Great at shooting, terrible at getting them on target, not great at converting those into goals. All that shooting by Steven Gerrard is good when it works, but it may not be the most sensible strategy. Moving to the next level down, the data on shooting from inside the six-yard box is also fascinating. Arsenal turn out to be the best at getting these, while Chelsea are the best at preventing them. If you were trying to improve Fulham, you might start with keeping the ball outside the six-yard box more often. And Martin O’Neill, of Aston Villa, has some work to do here as well.

But these are just the first fruits of our new shots work. There’s plenty more to come.

finktank@thetimes.co.uk