Helder Costa: Important Cog or Expensive Mistake?

In this piece, Tom Alderson assesses Helder Costa’s time at Leeds through the lens of his two seasons in the Championship with Wolves.

Helder Costa came to Leeds last summer with high expectations from fans after becoming the club’s biggest signing since relegation from the Premier League. Much was expected of him by the Leeds faithful largely because of his role in Wolves’ rise and promotion. However, some fans feel he has underperformed this season due to his lack of goals and assists compared to his standout 2016–17 season.

But just how good has he been? I want to interrogate this question more fully, taking a look at all three of the Portuguese’s three seasons in the Championship.

The Stats

Costa’s first season with Wolves — the 2016–17 season — was his best in the Championship, scoring ten goals and providing a further eight assists.

SmarterScout uses algorithms to quantify a player’s actions on the pitch and tries to evaluate how this contributes to their team’s expected goals and assists. This is adjusted for the league the player competes in.

In the graphic above, we can see that Costa is given a value of 84 for attacking output, backing up how good Costa’s season was and how important he was to Wolves. He also ranked highly in defending quantity but this is due to Wolves as a team rather than Costa as an individual.

Stylistically, Costa ranked highly for disrupting opponent’s play, passes towards goal, dribbling and, unexpectedly, aerial duels (although the duel ratings show Costa was not successful in many of these). Considering how many goals Costa scored that year, it was unusual to find a shooting style rating of 57. Costa accrued 7.62 Expected Goals (xG) and 8.83 Expected Assists (xA) in the 2016–17 season, suggesting that he overperformed in front of goal and slightly underperformed on the assists front. This may explain his relatively lower style rating for shooting and higher score for passes towards goal.

The visuals below show the contributors towards Wolves’ Expected Goals and Chances Created in the 2016–17 season. The larger the size of the block, the greater the contribution.

It can be seen that, whilst Costa was one of the biggest goal threats in the team, the responsibility was shared around. In terms of chance creation, Costa was the top producer for Wolves with only Jon Bodvarsson averaging more than a chance created per game.

This changed the following season with Wolves adding players such as Diogo Jota and Ruben Neves amongst other players. This meant that Costa was no longer required to handle the responsibility of being the club’s focus of attack. This saw a drop off in production with Costa only producing five goals and four assists, half of what he managed the season before.

If we focus on the change in Costa’s xG and Chances Created compared to the season before, there was no change in xG per 90 minutes and only a 0.12 drop in Chances Created per 90 minutes. The reduction in the size of Costa’s boxes is therefore due to the improvement of the squad and also due to a change in tactics with the appointment of Nuno Espirito Santo.

Looking at Costa’s SmarterScout Style Ratings for the 2017–18 season, we can start to get an idea for how his role changed from the previous season. The most notable changes were in his ‘Link Up’ performance which increased from 8 to 94, and ‘Pass Towards Goal’, which dropped from 75 to 3. The increase in ‘Link Up’ is likely due to Wolves moving towards a possession style of football due to the improvement in the squad. This also contributed to his decrease in overall chance creation and scoring chances and, therefore, the reduction in ‘Pass towards Goal’, shooting style rating and attacking output.

The more subtle changes were found in the ‘Receive in the Box’ and ‘Dribbling Style’ ratings. In his second season at Wolves, Costa was receiving the ball in the box much more often, leading to an increase in rating from 31 to 60. There was also an increase in Costa’s dribbling rating from 71 to 84. This will be analysed further when we look at footage of Costa.

Finally, we move on to this season at Leeds. The main criticism of Costa has been his lack of end product, with his three goals and three assists representing his worst season in the Championship (although hopefully there is a chance for him to fix that yet).

The stats above back up the claim that this is Costa’s worst season output wise, although he can consider himself unlucky to not have scored more with an xG of 5.53. It was not solely goals and assists that made Costa a target for Leeds, however, but his ability to dribble that was lacking in the Leeds squad during Bielsa’s first season.

If we look at Costa’s SmarterScout visual at Leeds we can see many similarities to his second season at Wolves, with the ratings for ‘Attacking Output’, ‘Link Up’, ‘Pass Towards Goal’ and ‘Dribble’ being nearly identical.

Costa’s main style ratings this season, according to SmarterScout are ‘Shoot’, with a rating of 96, and ‘Receive in the Box’, with a rating of 99. Both of these have seen a large increase on his last season in the Championship with Wolves. This is likely due to opposition teams sitting deep in their own half, meaning Costa is receiving the ball much deeper compared to previous seasons.

Looking at Costa’s progressive running stats per season, we can see more similarities between his second season at Wolves and this season at Leeds. Looking at the similarities and differences between Costa’s statistical profile for each of his Championship seasons, it is starting to become clear why Leeds were interested in buying Costa last summer and why the 2016–17 season was his best for goals and assists.

The Tactics

Wolves set up mainly in a 4–2–3–1 in the 2016–17 season with Costa playing on the right wing. This can be seen in the heatmap below:

If we focus on the colours of the heatmap blocks, we can see that green and yellow features most predominantly, signifying medium- and long-range passes. Costa also found himself in the penalty box a lot that season, mostly dribbling or taking shots on goal. We’ll be looking at this in more detail in the clips later.

Another thing to notice is how much of the pitch SmarterScout registered Costa playing in. It was obvious from watching footage that Costa sometimes switched to the left-hand side. This may also be due to Wolves playing on the break that season: Costa would have been free to move around the pitch during breakaways.

In the following season, Wolves changed to a 3–4–3 formation with Costa playing on the right of the front three. Costa’s heatmap for that season can be seen below:

The first thing to look at is the change in position compared to the previous season. Costa moved to a more central position as the width was now provided by the wing back — usually Matt Doherty. Costa also found himself in the box less that season, despite playing more as a forward. This is likely due to Leo Bonatini and Diogo Jota being the main focuses for goalscoring.

The second changes are in the colours of the blocks. Costa played more shorter passes that season, as can be seen by the light blue/green blocks. As previously mentioned, this may be because of the improvement in the squad meaning a move to more possession and, therefore, more interplay.

A smaller but noticeable change can be seen near the halfway line. The colour of the blocks are a darker blue/purple, meaning Costa was dribbling in this area of the pitch. We will look at the reasons for this in more detail in the next section.

The final heat map is for Costa’s current season at Leeds, where Costa plays on the right wing in either a fluid 4–1–4–1, or a 3–3–1–3.

What is most noticeable with this heatmap is how focused Costa’s play is on the right wing with very little play in other areas of the pitch unlike at Wolves where he was allowed more freedom.

The colours of the blocks outside of the penalty box are similar to the 2017–18 season: dark blue/purple near the halfway line to represent dribbling and light blue/green representing short passes in the opponent’s half.

Another difference is in the penalty box where Costa found himself more compared to his second season at Wolves. The colours of the blocks are a mix of yellow/green and white, signifying Costa crossing the ball or taking shots. This change in positioning and style of play is probably as a result of opponents defending deeper against Leeds compared to when opponents played Wolves.

The Eye Test

From watching footage of the 2016–17 season, a tactic that was commonly used by Wolves was to switch the play to the right-hand side of the field to Costa, either in the final third or early in the play as can be seen in the image below:

This tactic may explain why Costa had a high aerial style rating that season as poor long balls would leave him in an aerial duel with the opposition left back.

The aim of this tactic was to leave the left back isolated against Costa:

Once in this position, Costa would either run to the byline and put in a cross or cut inside and take a shot on goal.

In this case, he took the latter option which led to Wolves second goal against QPR.

Costa was Wolves’ best attacking player that season; using this tactic allowed him as much time and space as possible to pick the best option and it is easy to see why he was effective at creating chances in 2016–17.

The visual below is taken from a match against Sheffield United during the 2017–18 season. Costa is highlighted in yellow with Matt Doherty, the right wing-back, highlighted in pink.

We can see in this image Doherty is providing width on the right for Wolves meaning Costa is free to play nearer to the centre of the pitch. The reduction in time and space because of this may explain Costa’s reduction in goals and assists compared to the previous season.

Costa’s main role from watching footage of the 2017–18 season and this season at Leeds was in transition. In the graphic below, Costa receives the ball inside his own half surrounded by four opposition players.

Instead of passing the ball, Costa begins to dribble into the space in front of him:

Once he has dribbled past one player and beaten another Costa finds himself dribbling into a large area of empty space.

This puts the opposition defenders under a lot of pressure as they are now running towards their own goal with Costa running directly at them. If one of the defenders comes to engage Costa, this leaves room for him to pass the ball to a teammate in space. If they don’t engage then Costa is left to run at the defence.

Finally, as a result of his role in dropping deeper to collect the ball, Costa often found himself in the situation below with himself facing his own goal with a defender on his back:

Costa spins the defender, leaving him behind and is free to run into the space behind or to pass to a teammate.

His ability to get past players in this way helps Costa in carrying the ball from deep. He doesn’t need the ball to be given to him with no defender in front of him for him to carry out his role. This strategy has helped Leeds this season to attack teams in a different way to last season and helps negate teams defending deep in their own half.

The Final Word

It is clear to see from looking at stats and footage of Costa’s second season at Wolves and this season how similar his roles in both teams are and why he was a target for Leeds last summer. He has improved Leeds this season in transition as he has an ability to run with the ball which was lacking in the squad last year.

With Pablo Hernandez being the main creator of chances at Leeds from the right channel; Leeds often switch the ball to Jack Harrison on the left wing as he is often in more space. Leeds may be able to get more out of Costa by employing the tactics used by Wolves during his first season in the Championship. This is more likely to happen once Hernandez leaves, or starts playing less, but may mean Costa could help fill the void in creativity left by Hernandez.

Whilst he is unlikely to reproduce the output of the 2016–17 season in this current Leeds team with Bielsa, we may see improvement next season as he has slightly underperformed according to his expected goals this season also and due to the development we’ve seen from other players at Leeds after a full season with Bielsa. Either way, Leeds fans should look forward to Costa’s time at Leeds with quiet positivity.

You can follow Tom Alderson on Twitter @tomalderson97.

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.

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