05.10.18

BLOG: Shots, pressure, clarity and the Premier League

Article by Tom Worville

Points of interest

– 82% of shots taken under low clarity are either off target or are blocked

– Compared with the Premier League average from 2017/18, Arsenal took 8% fewer shots from high pressure situations whereas Burnley took nearly 9% more

– Kevin De Bruyne created 28% of Manchester City’s high clarity shots last season

 

In the Premier League last season, 9328 shots were taken. Removing penalties and direct free-kicks, that leaves us with 8909 shots.

Using new pressure and clarity qualifiers, which were first launched and discussed in this blog earlier this year, we can add further context to these shots. Let’s begin with a quick recap on the levels of these two qualifiers, which were collected from the beginning of the 2017/18 season.

Clarity

Clarity level Description Referred to in this analysis
0 Open goal (collected from the 18/19 season) N/A
1 One person (including goalkeeper) between shooter and goal High Clarity
2 One-two players between the shooter and goal Moderate Clarity
3 Over two players between the shooter and goal Low Clarity

As the data used here is from 2017/18, Clarity 0 attempts aren’t included.

Pressure

Pressure level Description Referred to in this analysis
1 No player within 4 metres or any movement towards the shooter Low Pressure
2 Players between 2-4 metres away Moderate Pressure
3 Players are less than two metres away and in tackling distance when a shot occurs High Pressure

What proportion of pressure shot types result in a goal?

Goal xG Goal – xG
Low Pressure 0.15 0.12 + 0.03
Moderate Pressure 0.08 0.08 = 0.00
High Pressure 0.09 0.11 – 0.02

When considering only pressure on shots, we see an interesting (albeit obvious) trend. High and moderate pressure chances are converted at a rate of 8% and 9%, whereas low pressure chances are converted at 15%, a rate of 3% above what they perhaps should be converted. If we consider the xG/shot of these chances – something we can interpret as an expected conversion rate in this context – we can see that on average, low pressure shots are of the highest quality.

It should be noted that these qualifiers are not currently integrated within our expected goals model due to the current sample size limitation. However, analysing these shots within the current framework allows us to better understand the impact that these qualifiers are having on shots.

What proportion of clarity shot types result in a goal?

Goal xG Goal – xG
Low Clarity 0.04 0.05 – 0.01
Moderate Clarity 0.08 0.08 = 0.00
High Clarity 0.26 0.23 + 0.03

The above table shows us that just 4% of low clarity shots are scored, whereas the average xG/shot of these chances suggests that 5% of these should be scored. High clarity attempts are converted over 3% more than xG/shot suggests.

By comparing both clarity and pressure separately, there appears to be obvious impacts on conversion rates under the differing levels. While high clarity chances are converted above xG, and have a high expected conversion rate, these are likely to be down to being from one-on-one situations. Furthermore, low pressure chances are also converted above the level we’d expect – likely down to these situations being from tap-in situations, which further analysis could confirm.

The relationship between clarity and pressure (Goal | xG)

Low Pressure Moderate Pressure High Pressure
Low Clarity 0.05 | 0.04 0.04 | 0.04 0.04 | 0.07
Moderate Clarity 0.10 | 0.07 0.07 | 0.07 0.08 | 0.1
High Clarity 0.33 | 0.28 0.22 | 0.2 0.19 | 0.19

What proportion of clarity shot types are on target/off target/blocked?

Now that we’ve seen the various conversion rates for different chance types, it’d be good to understand the impact of clarity on a player’s ability to direct the shot towards goal.

Low Clarity Moderate Clarity High Clarity
Off target 40 39 40
On target 18 31 59
Blocked 42 30 1

A high proportion of low clarity efforts are blocked (42%) or off target (40%), reinforcing the notion that having a ‘clear sight of goal’ should be high up the list when a player is considering whether to shoot. There’s also an upward trend with on target shots – the higher the clarity, the higher number of these are on target.

Shots with moderate levels of clarity have a more even spread of outcomes, leaning slightly more towards them being off target. High clarity attempts find the target 61% of the time, and are very rarely blocked – and when they are it is usually down to good recovery work from a defender to block the shot or clear it off the line.

Team analysis

With this information outlined, we can now explore specific teams’ shooting patterns, analysing the proportion of shots that teams take at different pressure and clarity levels.

To visualise this data, we’ve devised a coloured matrix to allow for easy comparisons to the league average. This Leicester City example below helps illustrate how to interpret these visualisations.

Arsenal

Compared to the PL average, Arsenal take 8% fewer shots from high pressure situations and 2% more shots from high clarity situations. For the north London side, Alexandre Lacazette took the most shots in high clarity situations (22), ahead of Aaron Ramsey (13) and Pierre-Emerick Aubameyang (12). Of those shots that were assisted, Mesut Özil created 19 of the chances (17%), closely followed by Aaron Ramsey with 14 (13%). Again, of those shots that were assisted just 13% were from crosses (13).

Manchester City

Manchester City take 7% more chances under low pressure, and 6% more chances with high clarity – showing how their style of play looks to maximise situations where the final shooter is free. 24% of their shots come from situations with high clarity (just one player between the shooter and goal), 4% higher than Arsenal’s tally. 85% of Manchester City’s high clarity chances came from open play situations (which includes fast breaks). Kevin De Bruyne was the chief chance creator here, generating 35 of Manchester City’s 127 high clarity chances (28%).

Burnley

Burnley take 8.8% more shots under high pressure than the average Premier League team, matching the eye test in terms of their chance generation. Of those assisted, just over half came from crosses (51%) and 47% of all the high pressure chances came from set piece situations. Of these 149 chances, 50 were assisted by Johann Berg Gudmundsson (33) and Robbie Brady (17) combined, showing the importance of these two when creating chances. Of Gudmundsson and Brady’s chances, almost all of them came from crosses (26 and 14 respectively). Burnley’s strike force of Ashley Barnes, Sam Vokes and Chris Wood were responsible for taking 74 of these chances.

In defence

To get a better understanding of how teams defend shots, we can look at the pressure applied and clarity given up by teams.

Burnley

Burnley conceded 6.3% more shots at low clarity compared to the average team, highlighting Sean Dyche’s side’s tendency to get in front of the shooter. On the flip side, they don’t look to apply high pressure to shots that often – having 5.4% fewer of their shots being conceded under high pressure circumstances compared to league average.

Given the way that Burnley look to defend – not by pressuring the ball which can open them up to the opponent taking a player out of the game – it’s probably not surprising that plenty of these chances conceded are of low clarity.

Brighton

Brighton are another side with a tendency to defend in a low block, which is reflected in the types of shots they concede. Their matrix displays a large increase over league average in terms of low clarity shots conceded.

Manchester City

Manchester City conceded 3% more of their shots in low pressure, high clarity situations – logically the best situations for a given chance. It’s likely that their increased proportion of high clarity and low pressure shots is due to their press being broken and getting countered.

By assessing both pressure and clarity together, a matrix of conversion rates allows us to compare to the xG/shot of the shots in question.

Without question, low pressure x high clarity shots are the best quality chances, converted at a rate of 33%. All high clarity attempts were converted at double digit rates, showing the superior quality of chances of these types.

Follwoing this focus on team analysis, the next article will focus on a detailed analysis of particular players and their decisions these situations.

Back to analysis