SMRTILYTICS is a bit of a hobby of mine.  For a full explanation of this project, check out the main SMRTILYTICS page.  -OG

Week 22

It is 2016 and Arsenal are still (at time of writing), top of the League.

There are twists and turns left in this League yet, and it has been a fascinating study so far. With 22 of 38 played, we’d be somewhere around the 52nd minute of a 90 minute game and few but the brave would bet on a winner given only a little over half the season has been played. Still, we know a lot more than we knew back in August.

Typically in these posts, I give a summary of how some of the teams are shaping up as well as some of the possible reasons for that, followed by predictions and the Big Chart. This week, I’m going to replace the summary section with a bit more of an in depth look at the Arsenal and what I’m thinking about in the context of the numbers so far.

Here’s a brief snapshot of Arsenal’s numbers to date:

Points: 44 (1st)
Goals Conceded: 21 (3rd)
Total Shot Ratio (on target): 0.58 (5th)
Combined Metric: 0.41 (3rd)
Goals Scored (per game): 1.68 (5th)
Goals Conceded (per game): 0.95 (4th)

Keep in mind that the average goals conceded per game to win the title is 0.85 and 2.09 goals scored (data set back to 1990/91 season).

While I’m less concerned about Arsenal’s goal scoring, I am interested in what the numbers mean in terms of goals conceded. Averaging 0.85 goals per game conceded works out to 32-33 goals in a single season. Arsenal could only allow a maximum of 12 goals in the next 16 games to achieve that and while it isn’t far-fetched, recent goal fests at Southampton and Liverpool suggest that might not be possible. In contrast, Tottenham, who have conceded a league leading 18 goals could allow 1 goal in every game remaining except one and achieve that target. Certainly goals against is one of my personal favourites in terms of predictors, but anything could still happen. What is important at the moment is to win games and keep clean sheets.

Looking at Total Shot Ratio (on target) and I found some interesting information out that was hidden in the data. Dom (@OzGooner49) asked me how the stats have changed since Coquelin was injured. It’s a question that is thrown around podcasts and social media quite a bit and one worth investigating a little further.

Le Coq limped off against West Brom in Week 13. At that point, our Total Shot Ratio (on target) was 0.62 and our goals conceded per game was 0.85. Those numbers are pretty close to where I think we’d need to be to win the League if they were maintained until May (guessing a 0.65 TSR on target since I don’t have historical data on that). Since then, our numbers have notably dropped.

On the surface, most would say that is due to Coquelin’s absence, and that is likely true to some degree, but it struck us that Santi has been gone for the same period and its undeniable what he brought to the midfield. My feelings are well documented on the Spaniard, so I won’t default into the usual Cazorla love-in; but it is fair to assume that it was a combination loss of the two which have altered our overall numbers.

Another interesting point has been our away form since losing that duo. Prior to West Brom we had played 6 away games and lost 1 – winning the remaining 5 on the road. Since (and including West Brom), we’ve only won 1 away game of 6, losing 2 and drawing 3. Yes, we’re top of the table, but our away form has seriously dropped off.

Assuming that Coquelin’s influence is mainly defensive, I took a closer look at shots being allowed since he left and what I found was increasingly interesting. Since the TSR on target has dropped, you’d assume that number has dropped because we are allowing more shots on our goal (remember TSR is a combination of shots taken and conceded and one can offset the other). I was wrong.

In weeks 1 to 12, we averaged 6.5 shots on target per game and allowed 4.25; from weeks 13 to 22, those numbers changed to 4.4 shots taken and 4.1 conceded per game. As you can see, the biggest change was not in shots conceded, but the number of shots we’re putting on goal.

Offensively, it is possible that Coquelin and Santi allowed us to create more quality chances due to their defensive stability. They also were able to distribute quickly to put us into counter-attacking positions. Lastly, we did lose Alexis in the same period which reshaped our offence, putting Giroud up top and sliding Walcott out wide.

Defensively though, we are allowing roughly the same amount of shots but conceding more goals. I think this really speaks to disciplined shield that Coquelin provides our defensive unit and the smart positioning and ball retention of Santi. Without checking the position of those shots both pre and post West Brom, I’d theorize we’re allowing shots closer to goal (they would logically be more likely to score). I’m a big fan of Cech but he can’t influence the number and quality of shots he gets, only the goals conceded stat –and if we’re allowing shots closer to goal, we’re inhibiting his ability to do his job.

Despite all this number crunching, we’re still top of the table. On the downside, we need to get back to clean sheets and significantly improve our away form to stay there. The real positive is that Coquelin and Alexis are back in full training to help us. Given that I believe we will get stronger as more first choice players return from injury, I still have us under my prediction below as the eventual winners despite City besting us in most categories tracked. It will be tight.

Right, on to predictions:

Current Table:
1. Arsenal 44
2. Leicester 44
3. Manchester City 43
4. Tottenham 39

18. Newcastle 21
19. Sunderland 18
20. Aston Villa 12

Table based on Goals Conceded:
1. Tottenham 18
2. Manchester United 20
3. Arsenal 21
4. Manchester City 21

18. Aston Villa 38
19. Newcastle 39
20. Sunderland 45

Table based on Total Shot Ratio (on target):
1. Manchester City 0.67
1. Tottenham 0.67
3. Southampton 0.62
4. Liverpool 0.59

18. Aston Villa 0.40
20. West Brom 0.39
20. Sunderland 0.37

Table based on Combined Metric (see description above)
1. Tottenham 0.27
2. Manchester City 0.32
3. Arsenal 0.41
4. Southampton 0.41

18. Aston Villa 1.04
19. Newcastle 1.06
20. Sunderland 1.30

OG’s Prediction come May 2016 (unchanged this week!):
1. Arsenal
2. Manchester City
3. Tottenham – Literally anyone else except maybe Chelsea
4. Leicester

18. Newcastle
19. Sunderland
20. Aston Villa

Gurgle Gurgle Plip Plop

OG

Click this graphic to see it in full resolution