It’s a strange phenomenon in football when the ball goes behind for a corner. The crowd gets up on their feet with a surge of noise, excitement and anticipation in equal measure. But why is there so much expectation? On average this season in the Premier League, just 14.63% of corners result in a shot, let alone a great goal-scoring opportunity. Clearly, the tradition of centre-backs marching forwards and the penalty box filling up with players gives a sense that a goal is imminent, when in fact that isn’t the case. If the crowd level were to respond in line with the chance of scoring it would be nothing more than a ripple of applause!
Some teams, however, are better than others; choosing to spend their time on the training field practising clever routines in the hope of increasing their chances of scoring from the corner flag.
In this tactical analysis, we review the offensive corners taken by Manchester United in the 2019/20 Premier League season, looking at where they’ve succeeded and underperformed. We also provide an analysis of the statistics looking at how they compare with their league rivals.
League Comparison
Below is a chart showing the typology of goals scored by each Premier League team this season. The chart specifically lists the goals scored following a corner, with the league average set at 6.5 goals in total.
We can see that Man United fall below the league average, scoring just five goals this season, which makes up 11% of their total league goals. On first glance, this looks disappointing for United in comparison to their rivals. Liverpool, for example, have scored double the amount, netting 10 goals in total from their corners. We can also see that sides like Arsenal, Everton and Bournemouth have also outperformed the North-West club, which is unexpected for a side that dominates possession and have the likes of Harry Maguire to aim at.
Digging into the numbers a bit deeper, I wanted to explore what the statistics didn’t show us. Firstly they don’t show how many corners each side has taken. If United have scored five from five that tells a different story to if they’ve scored five from 500. Below is a chart showing the percentage likelihood of each team scoring as a result of a corner, listed in order of highest frequency of corners.
This chart shows a slightly different picture. Liverpool who led the number of goals scored from a corner now rank 8th best in terms of effectiveness. They average a 5.85% chance of scoring from each corner taken. The standout team in the league are Newcastle. The Magpies have scored nine from 103 corners, giving them an 8.74% chance with each corner taken – the highest in the league. Something for teams to work on before they head to St. James Park.
For a side like Man United, the chart doesn’t make for great reading, particularly with the quality of players they have. They rank 12th for potency, with just a 3.57% chance of scoring each time the ball goes behind.
So where’s it going wrong?
There could be a number of factors affecting Man United’s ability to score frequently from corner kicks. I thought the best place to start was looking into the areas they prefer to distribute their crosses and their respective effectiveness in each area. Below is a breakdown of how Man United have taken corners and the distribution levels into each zone. The amount of Expected Goals (xG) they’ve earned from each region has also been calculated.
What’s immediately striking is the ineffectiveness from Man United’s near-post deliveries. Ole Gunnar Solskjær’s men have attempted 27 crosses to this area, with just one resulting in a shot at goal earning them 0.09xG.
This is well below the league average of 0.42xG in this area and immediately highlights a lack of ingenuity when delivering to this zone, in terms of off the ball movements and delivery type. Or it shows that a high percentage of Man United’s deliveries are intended to land elsewhere and famously cannot beat the first man due to poor execution. Other than for Aston Villa, when delivering to the near-post, Man United are the least dangerous in the league.
The next thing to note is the preferred tactics of Man United. From their 140 corners, 41 were aimed at the GK zone, which is 29.29%. We can see from the darker blue shading in the graphic above that this is the case. Of those 41 deliveries, just four (9.76%) resulted in an attempt at goal earning them a total of 0.83xG this season, as we can see from the first chart.
However, digging deeper we can calculate that each delivery into this area generates on average just 0.02xG, which isn’t particularly dangerous – equivalent to a long-range effort on your weak foot.
If we compare that to the deliveries to the penalty area, which make up just 9.29% of their corner distribution, we identify something interesting. In direct comparison, we can see that the deliveries to the GK are most rewarding in terms of total xG, however, if we look closer we can see that Man United might be missing a trick. Of the 13 deliveries to the penalty area, five resulted in an attempt at goal. This 38.46% success rate, earning them 0.43xG is streets ahead of the effectiveness in any other area. Doing the same calculation we can see that each delivery to the penalty area yields Man United 0.036xG per corner – 0.016xG higher per delivery than those to the GK zone.
On this basis, if Man United had attempted as many deliveries to the penalty area than they did to the GK zone (41), they would have expected to earn an xG of around 1.48, on the assumption they continued their shot success rate to this area. Though that is unlikely as a 38.46% corner/shot conversion is high, the point remains the same; Man United are not distributing their corners to the optimal zones to earn the highest amount of xG possible. According to the numbers, an increased frequency of delivery away from the near post and into the penalty area zone will see increased goals for the 20 times league champions.
If we submerge even deeper into the numbers we come to an interesting conundrum. Based on the point above, if Man United delivered more corners to the penalty area they would see an increase in shots – but what about the quality of chances each delivery creates? Below is a chart showing the league average for each corner distribution zone. I have removed the short corners as these always result in zero xG due it being a short pass.
Firstly we notice that in general Man United are below the league average in most areas, other than to the penalty area as we’ve already discussed. The interesting point is the significant increase in the league average set by the corners to the GK zone (1.34), over double the next best of 0.57xG. This shows us that deliveries to this zone are clearly creating higher quality chances, which is logical due to area on the pitch; a header from the 6-yard box is easier than an edge of the box volley, and therefore earns a higher xG.
So if we continue with that thought, let’s look at the numbers for Man United. As mentioned they’ve had 41 corners to the GK zone, resulting in 4 shots – earning 0.83xG. Despite the previous point showing that they would earn 0.016 more xG per corner if they distributed to the penalty instead, there’s a clear caveat.
If the four shots amassed 0.83xG, that means each shot averaged 0.21xG. Compare this to the penalty area deliveries, where United had five shots earning 0.47xG – giving it an average of 0.09 per shot. This shows us that even though shots will come more rarely if deliveries are placed in the GK zone, when they do arrive, the quality is 2.2 times higher on average.
This throws up an interesting conundrum, more frequent lower chances or higher quality chances less frequently. This will no doubt be down to manager preference and the routines each coach prefers to run from their black book, however, I looked to see if there was an actual correct answer. This is where I turned to poker for assistance.
Expected Value
The conundrum is similar to a decision that crops up on a poker table. Is my chance of winning a hand outweighed by the amount I have to pay? If it is, you should generally call, as in an infinite amount of hands you will be profitable.
To make sense of this I’ll refer you to an easy example. Let’s say you flip a coin with a friend where you bet £1 on heads each time. If it’s head you’ll win £1, if it’s tails you lose £1. Naturally, you know your chances are 50:50, but I’ll follow it through. To work out whether you should play you follow the simple formula below.
(Chance of winning x Amount) – (Chance of Losing x Amount)
So if we plug our game into this we get the following:
(50% x £1) – (50% x £1) = 0
This zero means that other than for variance if you played an infinite amount of hands you would win no money but also lose no money – a pointless bet. If we alter the bet to £2 for heads and £1 for tails, the bet becomes profitable.
(50% x £2) – (50% x £1) = 50p
If you are getting these odds you should play every time as you’ll expect to make 50p per bet on average – money in the bank!
So how does this apply, you are no doubt asking. Well, I plugged the numbers calculated from our corner distribution odds to see which corner is most profitable. Obviously you can’t lose xG from a corner, but we can calculate which one stands you in the best stead statistically.
GK Zone – (9.67% x 0.21) – (90.24% x 0) = 0.02xG
Penalty Area – (38.46% x 0.09) – (61.54% x 0) = 0.04xG
We can see that based on the numbers posted by Man United this season, the results show (albeit marginally) that they are better off delivering to the penalty area than to the GK zone. This is because on average the Red Devils are yielding 0.02xG more per corner – a higher expected value.
The test obviously has a few points to note; firstly, it is based on Man United’s ability to date. It shows us that based on their performance so far, they are 0.02xG better off by delivering to the penalty area. This, however, does not mean Man United can’t improve their deliveries to this zone and therefore change the expected value calculation. We’ve seen that they have an unusually high shot/corner rate (38.65%) to the penalty area, which is unlikely to continue indefinitely and will, therefore, drop the expected value.
We’ve also seen that Solskjær’s men are 0.51xG below the league average for deliveries to the GK Zone. This could be from a number of factors, including delivery type, delivery quality, movement off the ball and numbers committed to these corners – enough for another article to review. If United identified these areas that are preventing them from being as successful as their rivals with corners distributed to this zone, they will no doubt alter the expected value calculation.
So what have we learned?
Firstly from this set-piece analysis, we’ve learned that in general, Man United have been underperforming from their corner kicks this season. They are under the average in terms of goals scored and rank lowly for average chances expected per corner. Whether it’s something Solskjær is yet to focus his attention on, or Man United have deliberately committed fewer players to prevent being counter-attacked against is something that requires further analysis in a future article. Either way, the numbers tell us they are not reliant on corners to score their goals.
Secondly, we’ve learned that on current performance, Man United are hugely underwhelming in their front-post corners. With the creative players they possess and the aerial threats in the squad – even without air-tight, synchronised routines – you’d expect a better return than second-worst in the league.
Finally, based on their current ability we’ve solved the conundrum of which corner distribution is highest yielding for the Red Devils. We’ve proved using the statistics that the optimum delivery type based on their current routines is the delivery to the penalty area. By understanding the numbers Solskjær and his men might just pick up a few vital goals in tight games from these set-piece deliveries, something they’ve been failing to do up until now.
