This is more of a thought exercise than something I’m going to take seriously – I doubt I’ll revisit it too often in the future. Predicting single matches seems pretty silly to me, there’s simply too much random in a game (case in point, 11% of premiership goals are scored due to deflections).
That being said this is one of the occasions where there’ll be plenty of predictions around so I’ll add mine to the mix.
The methodology is as follows. I’ve picked the twenty teams that as a group have a shot ratio* average equal to that of both City and United. As such:
*Shots ratio = total shots for/(total shots for + total shots against)
I then looked at every game where one of City’s comparable’s played at home against one of United’s comparable’s. In total there were 38 games – a seasons worth of data. First lets see how the games break down in terms of attacking metrics:
The home teams are actually even more dominant than their average (their shot ratio rises from 0.644 to 0.649) whereas the away teams – despite being comfortably above average normally – are completely dominated (their shot ratio drops from 0.554 to 0.351)
So how does this translate to results?
Over the course of a season the home team scores 80 points, as opposed to a mere 20 for the away team. That’s a pretty remarkable return for the home team. As previously mentioned, their opponents are comfortably above average. The sample size is still pretty small but the numbers don’t look horrendously out of whack at first sight.
Using this and the last post it’s pretty easy to weight the title race. To recap the post-derby title odds from last time
And thus apparently I have City as favourites to win the title. The derby still decides it though. It’s going to be a great game to watch – I imagine there are few games with this much riding on them.