In honor of the 2019-2020 English Premier League season beginning last weekend, I thought it would be interesting to look at the best finishers in the EPL last season. Just so we are all on the same page, I am defining a finish as the conversion of a shot into a goal. After spending way more time than I would like to admit copying and pasting data from the Premier League website and Understat.com, here is the spreadsheet I ended up with (Chelsea fans, you might want to look away):
This is a lot to digest so let's focus on the Finish % column for now. At a quick glance it's interesting to see which players are elite (dark green), above average (light green), below average (light red), and my goodness you can't finish (dark red). Here is another way to look at the data:
Initial Thoughts:
1. Two of the bottom three finishers in the EPL play for Chelsea. Between them, Barkley and Willian have a combined 6 goals on 100 shots and 9 big chances missed. Ouch.
2. Sergio Agüero leads the league in goals per game (0.81) so I expected him to be an elite finisher. I was very surprised to find that he only converts 17.8% of his chances, 3% above average.
3. Aubameyang missed 23 big chances during the 2018/2019 season. To put that in perspective, the average number of big chances missed among the players included in this analysis was 7.25! Despite that, he still ends up in the elite finishing group for converting 23.4% of his shots into goals. I'm not sure how I feel about that.
4. Keep an eye on Anthony Martial this season.
My Reactions
It's interesting to see how some players in the elite finishing group score a below average number of goals per game, specifically Luka Milivojevic and James Ward-Prowse. While they did take a below average number of shots, they also scored a lot of penalties and free kicks. 3 of Ward-Prowse's 7 goals and 11 of Milivojevic's 12 goals came from penalties or free kicks.
I thought it would be more accurate to only include goals scored from open play so I spent a few more hours copying data from the EPL website and adjusted the finishing table:
First off, Pogba needs to stop shooting. Pierre-Emerick Aubameyang, Glenn Murray, and Jamie Vardy all dropped out of the top 5 due to goals scored from penalty kicks. Milivojevic dropped from the 12th best finisher to the worst finisher, and when you take a look at his shot map for the 2018/2019 season it makes a lot of sense:
Only 3 of his 57 shots were taken inside the penalty area. The guy only wants to score bangers. Full disclosure: I didn't watch a single Crystal Palace game last season, but my first thought after seeing this shooting graph is that Milivojevic wasted 46 Crystal Palace possessions. As the great Stuart Holden once said, you want every possession to end with something: a shot on target, a corner, or a goal. Milivojevic decided to take 46 really low percentage shots, which ended up being blocked or missed, instead of completing an extra pass to try and move the ball further down the pitch or create a better scoring opportunity.
The Top 5 Finishers in the English Premier League
After only taking shots and goals from open play into account, here are the top finishers in the EPL according to Finishing %:
1. Sadio Mané (22 goals, 87 shots, 25.3%)
No surprises here. The man is lethal.
2. Anthony Martial (9 goals, 38 shots, 23.7%)
With Lukaku gone and Pogba (hopefully) taking fewer shots (he only scored 6 goals from open play on 98 shots for a whopping 6.1%) Martial could have 90+ shots this season. He is an excellent shooter when he plays centrally, and I think moving him up top was a great idea.
If he continues to score at the rate he was last year (1 goal every 3.90 shots) I think Martial could have a 20 goal season (he's already off to a perfect start scoring a goal on 4 shots last weekend). However, I will admit it is extremely difficult to predict the number of goals a player will score based on one season of data and shots per goal isn't proven to predict anything (That being said, I'm still going to stand by my 20 goal season prediction because it's fun).
3. Lucas Moura (10 goals, 45 shots, 22.2%)
Another case of if-he-gets-more-shots-this-season-he-could-have-more-goals. However, I doubt it since Harry Kane took twice as many shots as Lucas last season and that trend seems unlikely to change any time soon.
4. Raheem Sterling (17 goals, 77 shots, 22.1%)
After watching a few Man City games I was really surprised at this. I felt like Sterling missed a lot of chances last season and he never struck me as a great finisher, but that just goes to show that I am probably biased. He did score 15 of his 17 within 10 yards of the goal so in terms of finishing he scores on most of the chances that he should and he gets a lot of tap ins.
5. Ayoze Pérez (12 goals, 55 shots, 21.8%)
Pérez was tied with Romelu Lukaku on Finish% but while both players took 55 shots and scored 12 open play goals, Lukaku missed 15 big chances and Pérez only missed 7. Expected goals also backs this up as Lukaku had 13.11xG compared to the 8.29xG of Ayoze Pérez. This means that Lukaku should have scored 13 goals, and Pérez 8, given the quality of their chances. Lukaku scored 1 fewer goal than he should have while Pérez scored 4 more than expected, giving the title of 5th best EPL finisher to Ayoze. Normally I wouldn't use xG to calculate how good someone is at finishing is because xG models aren't great at taking into account the number of defenders between the shooter and the goal (https://bit.ly/2KAy8B7) and xG can get inflated for players with a lot of shots. However, since both Lukaku and Pérez took the exact same number of shots I felt it was relatively safe to use it as a tiebreaker.
Limitations of Finishing %
The biggest limitation here is sample size. I only used data from one EPL season and other analysts (Michael Caley, Sean Steffen) who have done studies on player finishing used four seasons worth of data. Some analysts have dismissed finishing skill as statistical noise, "At the heart of shooting is a law of averages, and, a lot of what most soccer fans perceive as 'good finishing' is simply data noise in a small sample size of a season... While fans may find it hard to accept, finishing skill, at least in MLS, is largely irrelevant." (https://bit.ly/2Mhd9FE) I agree that we can't accurately compute finishing skill with event data, but it's because soccer is fundamentally about space and event data doesn't take that into account. I think we can all agree that a wide open shot 18 yards from goal is easier to score than a highly contested shot 12 yards out. The future of statistically determining a player's finishing ability lies in William Spearman's pitch control model.
This model, besides being really cool, helps visualize how much space each team controls and was the star at last year's OptaPro conference. I believe it could be used to make a finishing metric that determines a player's finishing ability based on how much space they had to take a shot. The less space the player has, the greater the difficulty of the shot, and therefore a better finish is required. The pitch control model combined with xG would take into account the density of defenders between the shooter and the goal and therefore lead to a much more accurate way of determining a player's "finishing skill".
I don't watch many premier league games so if there is something that I missed or some analysis you disagree with I would love to discuss it. Feel free to shoot me an email!