Here’s a funny pattern you probably didn’t notice: Mesut Özil has had three half-seasons where he averaged 0.33 goals per game. That 0.33 goals per game includes both goals scored and assists and is a bit of a low return for a #10 at a club like Arsenal. In comparison, this season in the Premier League de Bruyne has scored or assisted 11 in 19 apps (0.47 goals per game), Lallana has 14 in 18 (0.78), Alexis has 20 in 20.. wait.. what???
Ok, Alexis’ amazing form aside this season (he actually has 25 total goals and assists in 26 apps already), Özil has a weird history of playing well below par in certain half-seasons. This isn’t a knock on his his overall production for the year and I’m not one of those Özil haters just looking for a reason to dislike him. Arsenal need Özil if they are going to compete in the League, the FA Cup, and the Champions League this year.
But when you look at the season bifurcated by January (which is the halfway point in the Premier League) his record is odd:
1st half of 2016/17: 23 apps, 9 goals, 6 assists, 0.65 goals per app
2nd half of 2015/16: 21 apps, 3 goals, 4 assists, 0.33 goals per app
1st half of 2015/16: 28 apps, 6 goals, 16 assists, 0.79 goals per app
2nd half of 2014/15: 22 apps, 4 goals, 6 assists, 0.46 goals per app
(INJURY) 1st half of 2014/15: 9 apps, 1 goal, 2 assists, 0.33 goals per app
2nd half of 2013/14: 19 apps, 2 goals, 4 assist, 0.32 goals per app
1st half of 2013/14: 22 apps, 5 goals, 9 assists, 0.64 goals per app
(Real Madrid) 2nd half of 2012/13: 30 apps, 5 goals, 12 assists, 0.57 goals per app
(Real Madrid) 1st half of 2012/13: 29 apps, 5 goals, 8 assists, 0.45 goals per app
Özil has had just three poor half-seasons out of 7 half-seasons at Arsenal. That’s not terrible. And if you look at his production in terms of calendar years, he has only had one bad year:
2013: 31 goals in 52 apps = 0.57 goals per app
2014: 9 goals in 28 apps = 0.32 goals per app
2015: 30 total goals in 52 apps = 0.58 goals per app
2016: 22 total goals in 44 apps = 0.50 goals per app
And it can be expressed in terms of seasons in which he had just one off season:
2016/17 season: 15 total goals in 23 appearances = 0.65 goals per app
2015/16 season: 29 total goals in 49 appearances = 0.59 goals per app
2014/15 season: 13 total goals in 31 appearances = 0.43 goals per app
2013/14 season: 21 total goals in 43 appearances = 0.48 goals per app
2012/13 season: 30 total goals in 59 appearances = 0.51 goals per app
But what worries me is that when Alexis does revert to the mean and starts scoring at a 10% rate (which is low for a striker but average for the League) he will still probably score more than 20 goals in the League (alone) this season (assuming he takes the same number of shots (68) that he took in the first half of the season). Meanwhile, Özil is also on a hot streak. Not one as amazing as last season but well above his four year average. What happens if Özil starts reverting to the mean at the exact same time? Or worse, what if Özil drops to one of his historically poor 0.33 goals per app rates?