What we should learn from the FA Cup third-round tie between Arsenal and Leeds

A game of two halves? They were Jekyll and Hyde performances. Did you watch the FA Cup third-round tie between Arsenal and Leeds? It was one game typical of many. Leeds were so much better than Arsenal in the first half it was astonishing. As astonishing as how much better in the second half Arsenal were than Leeds.

In the first half Leeds had 15 shots and Arsenal three. In the second half Leeds had three shots and Arsenal 13. No replay this week was required, though it might have been. One of Arsenal’s second-half shots went in.

Nothing in the first half gave any hint of what was to come in the second half. Arsenal players said that in the dressing room manager Mikel Arteta was “angry” and “shouted a lot”. For whatever reason, the play in the second 45 minutes was unrecognisable from the play in the first 45 minutes.

Sometimes I write that the result of a match could have been different on another day. I feel it is a hard point to get across, and I wonder how many of you who read it will accept it.

There is a sense, though, in which teams play for two short periods one after the other on the same day – the first half of a match then the second. This is as close as we can get to reproducing circumstances exactly. Pretty much the same players with the same referee on the same pitch in front of the same spectators in the same weather. And more often than not, what happens in the second half is different from what happened in the first half.

As you will see if you look at the startling graph. It was compiled from the first and second half scores of Premier League and EFL games in the 20 seasons 1999-2000 to 2018-2019. It shows that the better a team did in the first half the more likely they were to do worse in the second half, and the worse a team did in the first half the more likely they were to do better in the second half.

Traders tell me that when a team are leading most punters back them in various ways to do at least as well in the rest of the match. This is the opposite of what usually happens. More often than not, such teams do worse in the rest of the match.

As your eyes move across the graph from left to right they start on games in which the home team did really badly in the first half and end on games in which the home team did really well in the first half. The blue line shows the proportion of games in which the home team did better in the second half. The pink line shows the proportion of games in which the home team did worse in the second half.

As you can see, the blue line swoops down while the pink line swoops up. This means home teams who did badly in the first half tended to do better in the second half and home teams who did well in the first half tended to do worse in the second half. What was good for home teams was of course bad for away teams, and vice versa.

More often than not, a team leading at half-time achieved a poorer goal difference in the second half. And the higher their goal difference was in the first half the more likely it was to be lower in the second half.

Of course a team leading at half-time could record a worse score in the second half and still win, though they did not always. My point is about the rest of the match.

We never seem to learn from experience what we should learn from experience – which is how often what has just happened turns out to be a poor guide to what will happen next. Pierre Bezukhov is a character in War and Peace who said: “All we can know is that we know nothing. And that’s the height of human wisdom.”

We can speculate on the influences at play in football matches. The better a team do in the first half the more likely they are to have benefited from good fortune, which might not be repeated, and the better the position they are in now the less well they need to do in the second half. But we do not know the relative strength of those influences, or of any others there might be.

Dr Jekyll did not know what turned him into Mr Hyde. He thought he did then found out he did not. Jekyll mixed a potion, drank it and turned into Hyde. Every time he drank the potion he was transformed. Until he ran out of one of the ingredients and bought some more. Afterwards the potion did not work. Perhaps, he speculated, the ingredient had changed – perhaps there was an impurity in the old supply that was not in the new supply. He could not tell.

We do not need to know why football teams give Jekyll and Hyde performances. Only that they do.

Goal feasts and famines

For neutrals the difference between first and second halves is the same as for partisans. If you are interested in how many goals are scored rather than who scores them, you will witness a similar phenomenon. The more goals there are in the first half the more likely it becomes there will be fewer in the second half.

Again these figures come from Premier League and EFL games in the last 20 seasons. The proportion of games in which fewer goals were scored in the second half than in the first half was as follows. After one first half goal it was 22 per cent, after two it was 57 per cent, after three it was 81 per cent, after four it was 93 per cent, after five it was 98 per cent and after six or more it was 100 per cent.

Today's top sports betting stories

Follow us on Twitter @racingpostsport

Like us on Facebook RacingPostSport