Title Contenders By The Numbers – Early Days Edition

Logo_EredivisieWith five matches played, we’ll look at some shot numbers across the Eredivisie Title favorites. Yes, it’s early days, and a lot of this may look different when, after another five matches, team numbers will start to settle at levels closer to their true values. Also, casually, this post will touch on shot quality a lot more than I did in the past. We’ll slowly work to a way of combining shot quality and quantity. An improved TSR, so to say.

 

Struggles

So far, over the first five matches, in terms of points won, each one of Twente, Ajax, PSV and Feyenoord has already had its struggles and none has won more than three matches yet. A look at the numbers will reveal where each team has failed to live up to expectations.

There’s only one team that owns the Eredivisie right now. The Dusan Tadic show that Twente is, dominates in terms of shot creation (123 shots for) and prevention (34 against). By definition, you’ll have the highest Total Shot Rate (TSR) then: 0.783. If you’re still not familiar with football analysis’ most significant stat, let me explain by saying that Twente creates over three times more shots than they concede. A simple plot of each team’s shots for (horizontal axis) and against (vertical axis) will help illustrate just how far ahead of the rest of the pack Twente is: nearly off the chart!

 

So, if Twente owns the Eredivisie, they lead the table, right?

Well, no, or at least, not yet. Oddly enough, Twente had trouble scoring in three of their first five matches, leading to two home draws already, and a 1-0 loss at Vitesse. At least they did win the other two games, to make it a 2-2-1 W-D-L record. Twente’s main problem was clutch scoring: 10 of their 11 goals were scored in the two wins. That will always mess up overall ratings like TSR.

 

Shot Quality

Twente’s struggles to score become apparent when we factor in the quality of the 100+ shots that they created. The inclusion of Eredivisie data in Squakwa.com enables us to collect several shot characteristics that reflect shot quality. Shot location is the most important factor here, but also shots and headers need to be separated, as they have different conversion rates.

Overall, we can stratify Twente’s shots for location and shot type in order to compare against a league wide conversion. The average team would have scored around 9.5% of Twente’s 123 shots. With this shot quality for (SQF) of just 0.105, Twente comes in just 15th. By the way, combining shot quality and frequency, the model expects Twente to score 12.9 goals (0.105 * 123), which is somewhat behind their actual 11 goals scored.

 

Misleading TSR

Behind Twente, it’s the usual suspects that complete the TSR top-3: Ajax (0,591) and PSV (0,578). Ajax, however, is one fine example of a misleading TSR! Their 52 shots conceded comes in 2nd lowest in the Eredivisie, but it’s the quality of conceded shots that is a source for major concern. Of 52 shots conceded by Ajax, a worrying 37 (71.1%) have come from inside the box and of those 37, the majority have come from central inside the box positions!

This all leads to a shot quality against (SQA) that is not even close to any other team in the Eredivisie: 0.155. So, despite coming in second in terms of the raw number of shots conceded, Ajax comes in 10th in terms of Expected Goals conceded (8.1), which ties in nicely with their 8 goals conceded!

 

PSV

PSV also deserve a mention in the shot quality column, but for their poor SQF. With an expected conversion of just 0.077 they rank 17th in terms of offensive Shot Quality. They did, however, hide that by significantly outperforming the model in terms of actual goals scored. Despite an expected 6.5 goals scored in the model, they managed 12 in real life.

This chart shows PSV’s shots and goals. At first glance, it’s not too bad, is it? But beware, the golden balls representing goals will soon start to dry up as too many of their attempts are from outside the box and from wide areas within the box. Yes, they often play compact and tight defenses, but the lack of central zone shooting will cost PSV dearly at some point in the season.

PSV attempted 85 shots, of which 38 (44.7%) were from outside the box. Those shots resulted in two goals, while PSV’s 10 remaining goals were scored with their 47 attempts from inside the box. Another reason for PSV’s poor offensive Shot Quality is the fact that from their shots from inside the box, under a quarter were fired in from nice central zones, and the far majority from lateral shooting positions.

 

Feyenoord

Should we mention Feyenoord here? Well, last season’s number three had certainly hoped to be title contenders this year around, but three losses to open the season have lead to a 2-0-3 record now. Let’s look one layer deeper…

Shots created: 64 (13th), shots conceded 73 (7th), for a TSR of 0.467 (11th). Not good.

Shot quality for: 0.088 (16th, ouch), shot quality against 0.110 (12th, ouch again).

We can factor that into the TSR by looking at Expected Goals scored (5.6) and conceded (8.0), which gives and Expected Goals Ratio of 5.6 / 5.6 + 8.0 = 0.412 (15th).

You still there? Good. For Feyenoord’s 12th place 0.492 TSR would is bad already, but a correction for shot quality drops them down, even to 15th. One small side note: Feyenoord played part of the match against Twente with nine men, which may skew the numbers. A bit.

 

In the end

Of the title contenders Twente, Ajax, PSV and Feyenoord, who had the best start over five matches? This in depth look at the numbers makes a firm case for Twente, as clutch scoring and a disappointing offensive shots quality are better problems to have than what the other teams are dealing with. Also, what Twente lack in terms of offensive shot quality, they make up for in terms of raw numbers with over 20 shots created per match.

Ajax have a horribly high quality of shots against, which explains their high amount of goals conceded (8 in 5 matches, versus 31 in 34 matches over last season). PSV have the reverse problem: a very disappointing shot quality for, but for the moment it is concealed behind an impossible conversion rate of nearly twice the model’s expectations. Feyenoord are mainly mentioned here for last year’s 3rd place finish, as their numbers indicate mid-table quality so far. Sure, they will regress to their true level a bit, but their disappointing opening is down to more than just bad luck.

 

TSR = Total Shots Rate

SQF = Shot Quality For

SQA = Shot Quality Against

 

data: squawka.com

2 thoughts on “Title Contenders By The Numbers – Early Days Edition

  1. bkalafat

    Isn’t Twente’s TSR 123/157, which equates to 0.783? You’ve stated its 0.767, which is very close so I wonder if there’s an error in your data somewhere. Thanks, good stuff.

    Reply

Leave a Reply