premier league

Are Man United more dangerous after a shock defeat?

Posted on by Dan

In this article, the analysts from the excellent Form Lab software analyse over performing teams in major European leagues.

Manchester United didn’t just lose against Man City, they were demolished. They suffered their worst home defeat last weekend since the formation of the Premier League, and in the process drifted out to 2/1 second-favourites to be lifting the trophy in May.

A popular theory is that this makes United more dangerous to their opponents in their next game because they are looking to avenge their loss.

But is this true?

We have examined the data across Europe’s Big 5 leagues to find out if there are any trends that might point to an instant return to form for United this weekend against Everton. Continue reading →

How to spot when teams are over-performing

Posted on by Dan

In this article, the analysts from the excellent Form Lab software analyse over performing teams in major European leagues.

A look at the Big Five European leagues shows that teams that are over-performing in comparison to the previous season tend to fall back during the season and by spotting these sides we have the basis for a successful strategy.

Using a sample that includes almost all Premier League matches since the start of the 1998/99 season we have a total of 920 games where a team has been 4-7 positions higher in the table than they finished the previous season. Had you backed their opponents every time you would have achieved a 5% profit on turnover.  Backing the over-performing side would have yielded a 23% loss and the draw would have been a 15% loss. This is based on average odds, taking the top odds could yield better results.

Narrowing this sample to matches where the teams in question were playing away and were not playing an opponent who also fit the profile we are left with 456 matches and a profit of 13% from backing their opponents (37% loss backing the highlighted teams and 11% loss on the draw). Continue reading →