How do Leeds get the best out of Helder Costa?

The signing of Helder Costa from Wolverhampton Wanderers this summer was extremely significant. Not only will he become Leeds United’s most expensive signing since Rio Ferdinand, at a cost of £15,000,000, after his initial loan period is passed. As reported by the Athletic’s Phil Hay, he was also the player that Marcelo Bielsa had personally identified as the one he felt was needed most to provide the extra bit of quality to get over the line for promotion. Having come so close in 2018/19, Bielsa was not prepared to let this one go and, the story goes, he bombarded Victor Orta, Angus Kinnear and owner Andrea Radrizzani with WhatsApp messages to make sure the deal was completed in time for pre-season.

As it happens, they didn’t manage to get it done for the first week, but the club did keep their end of the bargain and Costa was announced with great excitement from the fan base. It’s hard not to drool when your club signs a player capable of scoring goals like this.

However, during pre-season Costa didn’t feature in any of the games that the perceived ‘first team’ played in, until the last one against Calgiari, where he came on as a sub. Instead, Bielsa chose to use him in the group of players who were left behind at Thorp Arch to focus on their fitness.

This shouldn’t come as a surprise, knowing Bielsa’s commitment to his methods. Costa was deemed not ready in terms of fitness levels and so he didn’t join up with the main squad. But as he was a marquee signing, there was a feeling that perhaps the head coach might bend on this one.

Bielsa being Bielsa, he still didn’t compromise on his principles and when the Championship season started, Costa found himself on the bench. Many were hoping to see him start on the right, with Pablo Hernandez shifted inside to the ‘enganche’ role; whilst others thought he’d take the position occupied by Jack Harrison on the left wing. Neither being the case, Costa has had to bide his time and didn’t get his first start in the league until Hernandez picked up an injury, meaning he’d waited 9 games before playing 90 minutes in the Championship.

Whilst Costa had been impressing in his cameos from the bench, picking up an assist for Eddie Nketiah’s match-winning goal against Brentford and constantly showing willingness to run at full-backs, there is a general feeling that he has struggled since coming into the starting lineup. When preparing to write this piece, I put a post on the @AllStatsArentWe twitter account asking for opinions on Costa’s start at Leeds and the response from most was that it was too early to tell and that he’d impressed in patches, but predictably, there were already those that were ready to call him a waste of money.

Whilst it might be fair to say that he hasn’t hit the ground running in the starting 11, one would think this only really applies to two of the three games that he has started and there may be mitigating factors at play here:

First of all, the two games that were lost in that run, coming against Charlton and Millwall, were games that the whole side performed poorly. In the former, Charlton scored early and then dropped deep with 3 centre-backs with Leeds moving the ball far too slowly to do any damage. In the latter criticism of Costa is perhaps more warranted as the performance of the team improved when he was substituted at half-time; though one could argue that was because Leeds changed their formation to control the ball better. In the first half, they had been hit early on by a high-pressing Millwall and struggled to keep the ball at all which was only made worse by the sending off of Gaetano Berardi.

In the game in between those two, Costa was hailed as one of Leeds’ most dangerous players, acting as a major threat on the counter as the Whites found themselves under the most sustained attacking pressure they’ve faced so far this season. Having the pace of Costa, combined with hold up play of Patrick Bamford meant Leeds always had an outlet to occupy West Bromwich Albion when their attacks broke down. Costa would certainly feel aggrieved that he didn’t finish this game with his first league goal for Leeds, having had an effort saved from point-blank range in the first half, added to a good strike that he worked for himself cutting inside after one of the aforementioned counter attacks.

One thing that’s particularly notable about the 3 games that Costa has started so far, is that he has come into the side to replace the injured Hernandez. The Spaniard is the conductor of Bielsa’s attacking orchestra and without him the team have struggled to go forward as a unit, particularly coming up against Millwall and Charlton, who had 10 men behind the ball after taking the lead. With that being the case, Leeds struggled to get Costa into the game, not giving him the ball in situations that he could be effective in.

During his first season at Wolves, Costa was their primary attacking weapon. Whilst they didn’t particularly use a playmaker in that season, Paul Lambert’s tactic was to switch the ball to him as early as possible in order to let him run one-on-one at a fullback. Here is an example of that from their 2–1 win at QPR.

With the ball beginning in Wolves’ left-back channel, Costa stayed high and wide on the right flank, ready to receive the switch of play. In a one-on-one position, Costa backs his excellent dribbling ability to cause havoc and he was always going to drive towards the goal, where he was able to find the finish. One can’t help but draw comparisons to what tends to happen when Leeds switch the play to Jack Harrison. Unfortunately, with the Manchester City loanee, the ball tends to be passed back down the line or crossed in, most often not finding a Leeds United head.

In that 2016–17 season, Costa played 2,786 minutes across 35 appearances, racking up 10 goals and 8 assists. He attempted 6.69 dribbles per 90 with a success rate of 47.3% and his crossing numbers were 4.78 per 90 with a 31.8% accuracy. This was enough to make Costa player of the season in a struggling Wolves side that ultimately finished 15th in the Championship.

He followed that season by not being available for Wolves until mid-November due to an ankle injury. However, after working his way back into the starting lineup by Christmas, he still weighed in with 5 goals and 4 assists, with Wolves now sharing the goals out much more due to the arrivals of Diogo Jota, Ivan Cavaleiro and Leo Bonatini. They also now used a deep-lying playmaker in Ruben Neves, meaning they were not moving the ball towards him so quickly. That said, he dribbled even more often in the 2017/18 season, with 8.9 attempts per 90, being successful with a very impressive 52% of those. Less positively, his crossing attempts did drop down to 2.46 per 90 with a much lower accuracy of 23.1%.

Costa’s time in the first team was cut dramatically in the second half of the 2018/19 season as Wolves changed their formation and stopped playing with wingers. However, he still managed to score in a win against Tottenham Hotspur.

So far in a Leeds United shirt, Costa hasn’t seemed to have a defined role in Leeds’ attack. As he has been filling Hernandez’ position on the right, which the Spaniard plays as a wide playmaker rather than a winger, having Costa hug the touchline has looked alien to the way Leeds usually build their attacks.

When Hernandez returns to fitness it is almost certain that he will return to that position so that Leeds can go back to Bielsa’s favoured game plan to overload the right-hand side with him drifting inside, the right full-back overlapping and Mateusz Klich coming to support. Then with the opposition pulled over to the right, the switch of play would find Jack Harrison positioned high and wide on the left in a one-on-one position.

Here is an example of it that tactic from Leeds’ 1–1 draw at home to Nottingham Forest:

When Adam Forshaw received the ball, Leeds had Stuart Dallas, Patrick Bamford, Mateusz Klich and Pablo Hernandez all situated on the right of the pitch. The tactic of overloading had worked perfectly as all but two of Forest’s outfield players were pulled over to that side to defend the multiple dangers.

With 8 players occupied on the right, Forshaw picked the crossfield pass to Harrison, who now had a one-on-one situation with the full-back. He managed to control the ball perfectly, but rather than take the opportunity to dribble towards goal for a shot or towards the byline for a cutback or ball across the face of goal, he took the option of a quick pass into the area which went behind the arriving Pablo Hernandez and was quickly cleared. Unfortunately, there are many examples of promising positions breaking down like this.

I want to suggest that this is where Costa should be playing for Leeds.

Whilst he has made all of his starts for Leeds on the right-hand side, he is naturally a left-footer, so should be more than comfortable on the left. Some might find this harsh on Harrison as his work-rate is so high and he seems to carry out his tactical instructions to the letter. However, it is hard to argue that his output in terms of take-ons, passing and crossing accuracy and end product, i.e goals and assists, is good enough from somebody who receives the ball in such dangerous areas as often as he does.

To demonstrate this, here are Harrison and Costa’s player report cards compared to all Championship wingers who have played 400 minutes or more so far this season:

As you can see, Harrison is extremely busy, with his passes per 90, key passes per 90 and touches in the box per 90 all rated above the 90th percentile for Championship wingers. However, the metrics relating to quality, such as passing and crossing accuracy are all extremely low, with shooting accuracy below the median as well.

Costa doesn’t look as impressive numbers in terms of number of passes, shots and crosses but the quality of those he attempts are much higher, with the accuracy % being above the 75th percentile for all 3. He does have a lower dribble success % than Harrison, however, as he attempts 8.36 per 90 with a success rate of 54.76. This means he completes an average of 4.51 dribbles per 90, compared to Harrison’s average of 3.86. It could be argued that his number of passes, shots, and crosses would all increase if he was played in this position as well, although, with his high dribble attempts, it’s more likely that shots and crosses would go up than passes, as he always looks to go towards the goal.

Many will point to the off the ball work that Harrison does and his defensive contribution to Bielsa’s system as a reason that he won’t be dropped for the Portuguese. It should be noted though, that Leeds have been trying to make a change on the left wing since January, when they thought they had signed Daniel James. At that point in time, it was clear that an upgrade was needed on the left wing in order to bring more goals and assists. That is still the case now, as it took Harrison until the 11th game of the season get his first assist of the season and he hasn’t scored since the opening game, despite starting every one. For a player that ranks in the top 10 for the entire league (including strikers) for touches in the opposition box, this just isn’t good enough.

It is fair to say that in his appearances on the left-wing, Costa hasn’t performed the same role as Harrison, as he has tended to wander inside looking for the ball, meaning that when Leeds have looked for the switch ball, he has not always been available for it. The Derby game, in particular, saw Bielsa get extremely animated on the touchline as he furiously yelled at Costa to hold his position on the touchline. This seemed to contribute to Leeds losing their shape and retreating backwards in the last 15 minutes, but this also tied into the fact that Klich missed a penalty and the feeling that Leeds needed to hold onto their 1–0 lead rather than add to it seemed to grow amongst the team.

Perhaps the international break might be an occasion for Leeds to work on Costa staying in position on the left if Bielsa does see that as his plan going forwards. If it is simply Costa’s understanding of his tactical instructions in the position that is keeping the manager from using him there, this is surely something that can be solved. It might be frustrating for him to have to wait for the ball rather than get it early, but the rewards could be extremely high as in this particular system he would be given the ball in positions where he would be most likely to cause damage.

Of course, there are other options — as mentioned earlier; Bielsa could decide to continue with Costa on the right-hand side and Jack Harrison on the left and shift Pablo Hernandez to the enganche role. However, Leeds were at their most dominant in the first month of the season with a solid midfield of Phillips, Forshaw and Klich, so putting Hernandez in that role would compromise that ability to dominate. It is notable that since Forshaw has been injured Leeds’ performance levels have dropped, which Bielsa will be very aware of so that option, whilst not impossible, doesn’t seem that likely. He could also decide to return Costa to the bench to be used as an impact sub, but it would be strange for a club to make their biggest signing since 2001 and not plan for them to be a starting player.

Whichever way Bielsa ends up deciding to utilise Costa, he is clearly proven quality in the Championship and it is very likely that he will make a big contribution to Leeds at some point. In a side that’s currently struggling for goals in their quest for promotion, fans will be hoping to see this quality sooner rather than later.

You can follow Josh Hobbs on Twitter @JoshAHobbs.

If you have any interest in contributing to this blog, get in touch with us @AllStatsArentWe.

A Leeds United blog which focuses on the tactical and statistical aspects of the game