Why the Premier League games of Leeds and Burnley have become less unusual

Burnley’s first 16 games this season averaged 1.9 goals. Leeds’s first 16 games averaged 3.8 goals. Anyone who follows football would have been able to come up with a plausible explanation.

Sean Dyche is said to be a defensive coach. Marcelo Bielsa is said to be an attacking coach. Burnley established themselves in the Premier League by edging tight games. Leeds won the Championship and became a team everyone loved to watch in the Premier League by playing open, expansive football. They created a lot of chances but also gave away a lot.

Burnley’s last 16 games have averaged 2.8 goals. Leeds’s last 16 games have averaged 2.3 goals. Over the past four months Burnley’s games have yielded more goals than Leeds’s games.

Goals in football matches can be strange things, frustrating things for nearly everyone who bets on them. Everywhere.

Swansea’s first 22 games in the Championship averaged 1.7 goals. Their last 22 averaged 2.4. Rochdale’s first 22 games in League One averaged 3.4 goals. Their last 22 averaged 2.7.

There is something noteworthy about all these examples beside the fact that there were big differences between one part of the season and another. A high number or a low number was followed by one closer to the middle of the range. For most teams in most competitions in most parts of most seasons goals per game are somewhere between two and three. Unusual numbers are not always followed by more normal ones, but more often than not they are.

Let us delve into this more deeply.

Teams near the top or bottom of a table tend to be involved in higher-scoring games than teams near the middle. This is because one thing we can say for sure about goals is that they are a function of competitiveness. The wider the difference in ability between two teams the higher the score is likely to be. Teams near the top or bottom of a table feature in lopsided contests most frequently.

Even so, the impact of them is quite small. And variations other than those in most cases are random and not repeated. Even for teams with a distinctive playing style.

For Burnley and Leeds it was the first part of this season that was out of the ordinary, not the second. Burnley’s Premier League games in previous seasons were not particularly low-scoring for their position in the table. Leeds’s games in the Championship were not high-scoring for their position in the table.

Overall there is more randomness in goals totals than in corners totals, which means the past is a less reliable guide to the future for goals totals than for corners totals.

This might sound odd. Goals are what the players are focused on. They are trying to score them and not concede them. Corners are just things that happen now and again in the process. But there are more of them. About four times as many on average in a game. So goals are rarer than corners. And rare events tend to be more volatile.

Rare events are like small samples, which can be deceiving – even for clever people, such as those at the Gates Foundation of Microsoft pioneer Bill.

They heard of research from Pennsylvania in the United States showing that a disproportionately large number of high-performing schools were small. They donated nearly $2 billion to encourage the creation of small schools, sometimes by splitting up big schools. If they had asked more questions before handing over the money they would have found out that a disproportionately large number of low-performing schools were also small.

Small schools are small samples, and thus more likely to be extreme – extremely good or extremely bad. And to be a lot different in the next survey than in the last.

The graph below shows what happened in the first and second halves of football seasons. It was compiled with data from the last 25 completed Premier League seasons, 1995-96 through to 2019-20. The numbers along the bottom are goals per game in the first half of a season, the numbers up the side are goals per game in the second half of the season.

The data jerks about but I could draw a straight line through it, and if I did the straight line would be close to horizontal – rising slightly from left to right, but only slightly.

Teams whose games in the first half of a season were low-scoring tended to be involved in the second half of the season in games that were higher-scoring – still below average, but not by much. And teams whose games in the first half of a season were high-scoring tended to be involved in the second half of the season in games that were lower-scoring – still above average, but not by much.

This should be encouraging news for anyone who bets on goals. Okay, it does mean we should not expect unusual totals from the past to be reproduced in the future. But now we know we should not.

Not got a bet365 account? Sign up and get up to £100 in bet credits

Up to £100 in Bet Credits for new customers at bet365. Min deposit £5. Bet Credits available for use upon settlement of bets to value of qualifying deposit. Min odds, bet and payment method exclusions apply. Returns exclude Bet Credits stake. Time limits and T&Cs apply. 18+ begambleaware.org

CLAIM OFFER HERE

MORE FREE BETS

Today's top sports betting stories

Follow us on Twitter @racingpostsport