Expected goals are instructive but fail to account for luck
The Soccer Boffin with his weekly dose of betting wisdom
The ball is round, Sepp Herberger said, so it can roll in any direction. Herberger managed West Germany when they won the World Cup in 1954. When the ball rolls your way you win, when it rolls another way you lose. Usually anyway.
Expected goals stats show that winners are lucky and losers are unlucky. They may be other things, but they are also these things. Usually anyway.
Expected goals tell us how many goals would be scored on average with that number and calibre of chances. I have kept a note of Opta’s expected goals figures for Premier League matches this season.
As you would have anticipated, about half of teams score more than their expected goals and about half score fewer. What is interesting is what happens in matches that somebody wins and somebody loses – matches that are not drawn.
Over 70 per cent of winners score more than their expected goals – they get an above-average number of goals for the quantity and quality of their chances. Over 70 per cent of losers score fewer than their expected goals – they get a below-average number of goals for the quantity and quality of their chances.
In just over 50 per cent of all games that are not drawn the winners score more than their expected goals and the losers score fewer than their expected goals.
Sometimes a team will exceed their expected goals because their strikers were brilliant. Other times a team will not reach their expected goals because the opposing goalkeeper was brilliant. Even in those matches, however, there may be an element of fortune in the exceptional individual performances. Are they always that good?
Luck plays a part in most successes and failures but rarely is it discussed let alone accepted. “Why did they win?” the presenter asks. How often does a pundit answer: “They got lucky”. He might be right but everyone else in the studio would laugh and next week somebody else would be sitting in that seat.
Burnley have achieved surprisingly good results. They are seventh with 28 points, having scored 15 goals and conceded 12. Opta’s expected goals stats say Burnley should have scored 11 and conceded 22.
Everybody knows that Burnley have achieved good results by winning or drawing low-scoring games. What Opta’s expected goals stats tell us is that with the same number and calibre of chances for and against a team would typically have scored four goals fewer and conceded ten goals more. And instead of being just below the European places they would probably be just above the relegation places.
West Ham and Crystal Palace have posted surprisingly bad results.
West Ham, 18th in the table, have scored 14 goals and conceded 32. Opta say they should have scored 17 and conceded 22. Playing the same way, with the same type and volume of chances for and against, they would typically have scored three goals more and conceded ten goals fewer. And they would probably be somewhere in the third quarter of the table instead of the bottom quarter.
Palace, who are last, have scored ten goals and conceded 27. Opta say they ought to have scored 22 and conceded 21. With the same quantity and quality of chances for and against a team would on average have scored 12 goals more and conceded six fewer. And they would probably be close to where Burnley are now.
The tennis film Match Point starts with a voiceover by the actor who portrays a retired player called Chris Wilton: “The man who said ‘I’d rather be lucky than good’ saw deeply into life. People are afraid to face how great a part of life is dependent on luck.
"It’s scary to think so much is out of one’s control. There are moments in a match when the ball hits the top of the net. And for a split second it can either go forward or fall back.
"With a little luck it goes forward and you win. Or maybe it doesn’t and you lose.”
It goes to show that good and bad luck play a part in any sport. Usually anyway.
Walter makes a splash in 1954 final
Sepp Herberger managed West Germany when they won the World Cup for the first time in 1954.
Their captain and best player was Fritz Walter, who had contracted malaria during the Second World War, marching through southern Europe. Since then he had felt uncomfortable playing in warm weather. He could only perform at his best in rain.
The 1954 World Cup was staged in Switzerland. There is not much rain in Switzerland in the summer. On July 4, 1954, the day of the World Cup final, there was a downpour. And West Germany, with Walter at his best, beat Hungary 3-2.
How general play can reveal a team's secrets
When Ruud Gullit managed Chelsea in the 1990s he once asked for videos of their next European opponents. Chelsea approached the opponents, who sent recordings of some recent matches – with the goals wiped out.
Gullit laughed when he told the story. He learned more about how the team might try to score from watching the rest of their play than he would have done from seeing the goals. As Bobby Robson once said, it is because of all the things that happen in between that goals are scored.
Gullit wanted to know how this team might try to score against Chelsea. He could do that by watching the many times and ways in previous games that they had tried to score and failed. He did not need to see the much smaller and randomly selected set of attacks that had succeeded.
Follow us on Twitter @racingpostsport
Like us on Facebook RacingPostSport