Unless you’ve been living under a rock, you’ll know that Leeds United are on their second-best run of form of the 2019/20 Championship Season. During that run, they have won their last five games without conceding a goal.
Despite only picking up an assist each during this time, Leeds’ wingers, Jack Harrison and Helder Costa have been on excellent form throughout the run and Harrison, in particular, has been getting rave reviews.
I thought it would be worth looking at their seasons in a little more detail to assess how each of them has contributed to Leeds’ promotion push. They’ve both blown hot and cold at times and, as is the case with all wingers, they can be very frustrating and wasteful in good positions. Nonetheless, they have both been very important to Leeds’ attacking process.
As the radars above show, Harrison has been one of the most dangerous wingers in the league with very high non-penalty xG per 90 (0.21) — at an xG per shot of 0.14 — meaning he’s regularly getting in good positions to score, rather than just by taking a high volume of shots. Costa isn’t far away from those numbers with 0.18 and 0.1 respectively.
On top of this, they are both well above average for successful dribbles and have some of the highest numbers in the league when it comes to touches in the opposition box: they are a constant threat in behind fullbacks and have plenty of opportunities for pull-backs, crosses and shots.
Harrison is above average for expected assists per 90 with 0.21 and is one of the most creative in the league, averaging over a key pass per 90. Meanwhile, Costa might not create as many chances as Harrison, but he retains the ball better, with his passing percentage at 77.03% and his higher dribble success meaning he is dispossessed less often than his teammate.
For reference, here is a radar of West Bromwich Albion’s Grady Diangana, who many believe to be the best winger in the league.
As you can see, Diangana doesn’t get into the opposition penalty area as often as Leeds’ wingers. However, he is much more decisive with the opportunities he gets, matching Harrison’s figures for goals and surpassing Costa for goals and assists despite having played far fewer minutes.
So far, we’ve established the two Leeds wingers haven’t been up to Diangana’s level. But had he not been injured for long periods of the season, he’d be a very strong contender for Championship Player of the Year. Not being up to his level doesn’t mean they aren’t some of the best in the league in their positions.
There is one obvious weakness in both of their radars though, which we’ll look into more next.
In the scatter graph above, we see both Leeds players make well above the average for crosses per 90. Unfortunately, though, they both sit on just above 25% accuracy for their crossing which is significantly below the average of 31.61%.
This graph matches the eye test and proves why many have found Costa and Harrison so frustrating: despite the fact that they often look so dangerous, too often their final ball is lacking.
However, crossing accuracy doesn’t tell the full story: it doesn’t measure the quality of the cross, merely whether a teammate made connection with it. For example, a cross meant for the six-yard box that sails over everybody’s heads but is controlled on the other side of the box before going out of play would still be considered ‘accurate’ whilst one that flashed across the six-yard box with the striker missing it by an inch would not, despite the latter being the more dangerous ball in.
In Leeds’ case, despite putting in such a high volume of crosses, they operate a one striker system and their attacking midfielder, Pablo Hernandez, tends to arrive late in the box or hover around the edge of the penalty area. Thus, in terms of targets to cross to, it can often be Patrick Bamford in the middle and the winger from the opposite side having to make a run to the back post.
The example above would be one of those considered inaccurate, but Harrison essentially does as well as he can with it by quickly getting the ball out his feet, looking up and delivering to the back post. Unfortunately, Costa couldn’t quite manage to get to it before Boro’s left-back. But that doesn’t make it any less of a good cross.
Looking again at the scatter graph: of the players in the top right sector of the graph, the vast majority play in teams featuring two strikers or play with a target man in the box. For example, the player with the highest accuracy on the graph is Shane Ferguson of Millwall, who is either crossing for Tom Bradshaw or Matt Smith, two of the best aerial presences in the league.
Also featured in that sector of the graph are Kadeem Harris crossing for two strikers in Sheffield Wednesday’s 4–4–2; and Jeremie Bela, who is again playing in a 4–4–2 for Birmingham, with Lukas Jutkiewicz and Scott Hogan waiting in the box. The stand out player on the graph is Bristol City’s Niclas Eliasson, another winger operating in a 4–4–2 system and another man crossing to a tall centre forward in Famara Diedhiou.
With all that being considered, perhaps Harrison and Costa’s crossing accuracy stats can be looked on a little more positively.
Whilst crosses rely on a teammate getting on the end of the ball to be considered completed, dribbling is very much down to the individual.
In 2018/19, Leeds were lacking in dribblers: Samu Saiz was the only player to feature in the top 30 for dribbles attempted (minimum 1000 mins played) with 6.1 per 90. This season both Costa and Harrison feature in that top 30 with 6.61 and 6.18 per 90 dribbles respectively.
On this occasion, they both sit above average for dribbles attempted and for success rate (you’ll have to trust me on this: Harrison’s name has been omitted from the graph due to crossover with Callum O’Dowda).
Again, this graph feels like it matches up with the eye test. Leeds’ wingers don’t dribble as often as Diangana or QPR’s Bright Osayi-Samuel, for example, but when they do, they do it effectively. In recent weeks both of them have caused major headaches for opposition full-backs with Harrison, in particular, completing lots of take ons in the final third. Meanwhile, Costa has been effective in beating players in breakaway situations and carrying the ball rapidly over longer distances before making a pass or driving towards goal for a shot.
Having two confident dribblers in the team has brought a different element to Leeds’ attacking play compared to the 18/19 season where Leeds would often move the ball out wide, only to the pass it back inside or try to force a cross. During Jack Clarke’s short run in the side, they briefly had a player looking to get behind full-backs. Now they have that from both wings in every game.
Don’t we want more goal involvements from our wingers?
One criticism of both Harrison and Costa — particularly Costa, considering the fee that will be paid for him — is that they don’t have enough goals or assists.
Costa’s career-best figures came in the Championship in 16/17 where he managed 10 goals and 8 assists. He looks highly unlikely to get anywhere near those this year. On the other hand, Harrison has 5 goals and 7 assists, which is already better than last year’s 4 goals and 2 assists. That said, many fans would still feel that he should have more of both considering the positions he gets into.
The question is — why haven’t they contributed more?
On the goals front, whilst we’ve spoken about Patrick Bamford’s underachievement in front of goal many times, Harrison and Costa are the second- and third- biggest culprits when it comes to Leeds’ chronic xG underperformance. Harrison is underperforming by 2.92 and Costa by 2.45. If they were simply matching their xG, it’s unlikely that many would be questioning their output.
It’s fair to say that whilst both of them will know they should have scored more goals. Harrison, in particular, missed a number of ‘big chances’ during Leeds’ most recent poor spell. They have also both been incredibly unlucky: Harrison has hit the inside of the post with curling efforts from the edge of the box in consecutive games, whilst Costa was denied by a wonder save against Bristol City just as he was about to tap the ball into the net.
Looking at their shot maps below, we can see that Harrison and Costa have both failed to score from the six-yard box, despite both having excellent chances there. (Note: Harrison’s goal vs Reading isn’t included as these maps don’t include headers.)
In Harrison’s case, having hit the target so often from six yards, he would have expected to have scored more. However, he has been denied by a mixture of good goalkeeping and poor finishing. Costa on the other hand would hope to increase his shooting accuracy from 30% before the end of the season. He has missed the target completely on two occasions from a position that it would seem harder to miss than score.
I wanted to include the clip above in the piece because although it didn’t end up in a goal, or even require a particularly brilliant save to stop it, it highlights the quality of both players as Harrison carried the ball inside from the left before giving the ball to Costa who used great skill to fashion the shooting opportunity. Unfortunately for him he wasn’t able to get it into the corner, but if he can do things like this more often he will certainly score plenty more goals in a Leeds shirt.
As for assists: in the case of Harrison, 7 assists is not a poor return by any means. In fact, only four players in the whole league have more. Considering he averages more than one chance created per game, it’s not beyond the realms of possibility that he could break 10 before the end of the season. This is the same amount Hernandez managed last season in a career-best season. As with crosses, assists rely on teammates finishing the chances created for them.
When it comes to Costa, he would certainly be hoping for more assists than three at this stage. One can’t help but wonder, though, if he’d have made more had Eddie Nketiah stayed. Costa favours going to the byline for the cut-back or the low ball across the six-yard box, a spot where Nketiah particularly thrived. They combined in almost identical fashion for the Nketiah’s first two goals in a Leeds shirt.
Patrick Bamford gets a lot of chances as we know, but he makes very different runs, which perhaps don’t dovetail so well with Costa.
Also, it’s important to remember that Helder is still only in his first season under Marcelo Bielsa’s tutelage. Nobody would argue the fact that Harrison has massively improved in his second season and one would expect it would be similar with Costa. His role is different to what it was at Wolves, and Leeds’ tactics are not simply to switch the ball early for him to run at the full-back, which is what Wolves did in his first season where he was at his best. There will certainly be more to come from him.
It would be remiss to conclude this article without mentioning Costa and Harrison’s contribution to Leeds’ press which is enormous.
I don’t want to say a lot about it as I wrote about pressing recently and this article has gone on for long enough, so I will point you to this piece here, where I mentioned their role in challenges attempted in the final third and include a video clip of them both pressing high up the pitch forcing turnovers.
Two of the league’s outstanding wingers that we haven’t mentioned thus far are Brentford’s Said Benrahma and Brian Mbeumo, both of whom operate more as wide forwards in a front three rather than wingers in a midfield four like Costa and Harrison. They have 10 goals and 5 assists, and 14 goals and 6 assists respectively. However, Brentford are 16th in the league for PPDA, Wyscout’s pressing metric, whilst Leeds are 1st, meaning that Brentford’s wingers don’t have to put anywhere near the defensive work that Harrison and Costa do.
In context, with their remarkable contributions to Leeds’ pressing, their attacking output is all the more impressive.
You can follow Josh Hobbs on Twitter @JoshAHobbs.
If you enjoy this content and want to help us create more and better pieces, why not consider signing up to our Patreon page to get bonus material?
If you have any interest in contributing to this blog, get in touch with us @AllStatsArentWe.