Ouch, 1:6. Clearly it was not a good weekend for Man Utd fans, especially as no one saw it coming. As an Arsenal fan, I could argue that the 8:2 result wasn’t so bad as Man Utd were flying high and Arsenal were in turbulent times. But it was a lie. It hurt just as bad. After the match, Wenger jumped forward to offer his sympathy for Sir Alex. I doubt Sir Alex cares. The defeat gives City now a 30% chance to win the title, with Man Utd having 27%. 5 points from the top of the table and with more than 3 quarters of league still to go, he and Man Utd fans will be anxious to see just how the team will respond to that result.
So do teams start to underperform after a shocking defeat? I decided to take a look.
One can say that a team loses a game badly if the goal margin is like 4 or more. However for a good team, to lose against a weak team at home can sometimes also be pretty bad. So just looking at the goal margins would not be enough; instead I looked at the probability of such a GD or more to happen. If the probability was below 1%, we could consider it a bad defeat for the losing side. Searching through the Premier League data between 06/07 season and 10/11 season (5 season in total), I found 19 such matches. They are listed below:
The Derby vs. Villa and Middlesbrough vs. Man City match in 07/08 season were of less interest as they were approaching end of season. That left us with 17 matches.
I examined up to 10 Premier League match results which were immediately after the bad defeat. As usual, I compared each team’s average points gained against expectation (Derived from our Dectech Model). Below is what I found:
There appears to have no pattern of underperformance. Good news. Wait a minute, you might say, only Chelsea out of the big 4 has suffered big defeats. How have they coped?
It looks inconclusive to me. What do you think?
Hi Jin,
unusually poor,but isolated results don’t really impact on a team’s subsequent performance,imo.
Regarding the Big Four.
About once every three seasons one of the Big four has a run of 5 consecutive EPL games where they fail to win a game,(a poor run by their standards) and the average success rate over that 5 game period was 40%.The average success rate for the Big Four prior to this poor sequence over a representative number of games was 76% and their average success rate after the poor run was also 76%……The poor blip wasn’t a sign of bad things to come.
Can you explain the last Chelsea graph ?
http://thepowerofgoals.blogspot.com/
Mark.
Hi Mark,
Thanks for your comment. Regarding the Chelsea chart: the blue line represents accumulated actual minus expected points for up to 10 matches after their 3:0 defeat to Middlesbrough in 2006. And the blue line represents similar stats but after their home 3:0 loss to Sunderland last year.
Cheers,
Jin
Cheers .Jin.
Yes, inconclusive. I think that most premiership teams have a bad result at least once a season and often this makes them play better the next game because they dont want a repeat performance.