As you can tell, I am not Tim who wrote this column before and did an excellent job and was one of my inspirations for soccer football(sorry I am American) stats writing. Before I jump into the post I wanted to take a chance to introduce myself, my name is Scott, I hope to continue with some of the same things done in the past and add some new stuff along the way. Please let me know what you like and don’t like.
Pre-match I knew that this was going to be a tough challenge for Arsenal. Drawing conclusions from single matches is never a good idea but this match up in particular is going to be so unlike any other that Arsenal have this season (until they travel to Manchester for the return fixture).
An answer to a question that I get often: how can the shot on target xG be higher than the regular xG? That happens because they are separate models where with the regular every shot is assigned a value regardless of where the shot ends up. With the shot on target version, if a shot is blocked or misses it gets 0 xG but if it is on target it will have a higher corresponding value compared to the regular model. Over the full season the two end up very close to each other but over a single match the shooting accuracy can vary quite a bit and this helps to illustrate that.
The simulated odds for the match are based on all of the shots for the match. It is run through a Monte Carlo simulation, that simulates the match 10,000 times. This should give a rough estimate of the “fairness” of the outcome based on the shots for both teams.
22 – the number of tackles that Arsenal had today, including 8 in the attacking half.
14 – the number of interceptions Arsenal made, including 2 in the attacking half.
The defense was a work in progress today but there were certainly flashes of Emery’s new pressing system paying dividends.
38 – The number of passes completed, led Arsenal
44.8% – The percentage of passes attempted that went forward
7/11 – Final third passing
1/1 – Passes into the box
1/1 – Dribbles
1 – Time Dispossessed
4/6 – Tackles (2nd on Arsenal)
4 – Interceptions (2nd on Arsenal)
0.1 – xG chain
7 – Times he lost possession (6 bad touches, 1 time dispossessed)
These are the stats for Matteo Guendouzi. He was shaky to start the match off (understandably so) but after the first 15 to 20 minutes grew into the match to be one of Arsenal’s better players. It is waaaaay to early but this looks like a talented kid who will challenge Arsenal’s other midfielders for a starting position.
1 – Shot
7/11 – Passing
1 – Key Pass
1/1 – Dribbles
2/3 – Tackles
These were Aaron Ramsey’s stats in the match. Emery lined him up in the number 10 role, with what looked like instructions to push forward. From watching the match, it really looked like he struggled to influence the game. Finding space between the lines to operate isn’t Ramsey’s forte and in this match he didn’t create many chances for his teammates to find him consistently further up the pitch.
Emery made a sub on 54 minuets (its crazy to see a sub before 70 minuets after that being the default for so long. It’s a new era!) bringing on Alexandre Lacazette. With this change Arsenal seemed to have much better connection between the midfield and the attacking line.
It’s hard to separate things from the effects of the game state but looking at the average positions of the Arsenal players helps illustrate this as well
Arsenal player average touch locations pre and post Lacazette sub.
After sub, Arsenal really seemed to have more thrust and control of the match. Also slight score effect of being down a goal. I still think the Ramsey pushed forward didn’t work but Emery did well to change pic.twitter.com/VzhlNcEldT
— Scott Willis (@oh_that_crab) August 12, 2018
Starting the season with a loss sucks, but I think that there are still positives to be drawn on from this match. Arsenal have another tricky fixture next weekend away to Chelsea to try to make sure that they don’t have too big a deficit in the top 4 race to over come.
Sources: Opta, StatsZone, my own xG models