Just a quick post looking at the title race. I’m focusing on the two Manchester teams – there’s a non-zero percent chance that Arsenal or Spurs win the league but it’s very close to zero.
First up the underlying numbers for the two teams, total shots ratio (TSR) – the best indicator I have of a teams future performance and PDO – a great indicator of the amount of luck a team is having.
Two points jump out
1) City control the play much more than United, taking 64% of the shots in their matches, compared to a pretty pedestrian 52% for United (a big drop from 63% each of their last three seasons)
2) Both teams benefit from a bloated PDO (shooting % + save %). In fact these would be the two highest values since the 2000-01 season, when my data begins.
To predict how the teams finish the season I’ve gone back and looked at the team seasons with the closest TSR to those of United and City and compared how they performed over each ten game segment during the season (ie games 1-10 is segment one, games 2-11 is segment two…). The first plot shows the number of points picked up in each of the segments
It’s a bit of a mess (290 data points is a lot but it also leaves a lot of noise and the 3-1-0 points system makes some combinations more likely to appear than others) but the overall message is clear, City are expected to score more points than United from this point onwards. These results are then taken and fitted as a normal distribution, and added to the points the teams have already scored to give the distribution of points we’d expect each team to have at the end of the season.
(If anything this is a bit disingenuous to Man City as their expected points aren’t really normally distributed)
|Expected points||Chance of winning the title|
Two final notes, firstly this approach makes no allowance for the remaining fixtures for each team. Using my measure of quality of competition it looks like City have the tougher games to finish the season (their opponents take ~50% of the shots in their matches compared to ~48% for United’s opponents) so their chance of winning the league can probably be knocked down a bit (probably somewhere between two and five percent) but they’re still heavy favourites.
Secondly the derby is shaping up to be a massive game. If United were to win that it then we’re basically looking at a 50:50 split.