Why Wolves’ Decision To Revert To 3-4-3 Is A Good One

From the moment he was appointed, Nuno Espírito Santo’s Wolves side lined up in a 3-4-3 shape. It led them to the Championship title and a respectable first half-season in the Premier League, however, after some poor results, the shape was changed to a 3-5-2, with Leander Dendoncker coming into the line-up at the expense of one of the attackers.

Since then, the 3-5-2 had become a mainstay but, after a slow start to the 2019/20 season, Wolves have since reverted to the 3-4-3. It started slowly, often switching to it when chasing the game (which has been quite often this season, as they’ve regularly conceded first) but, having started with it in the previous two games, it seems as though it may be here to stay again – which is good news for Wolves fans.

Before continuing, @WolvesAnalytics has posted a few threads talking about similar topics as I’ve talked about here and there may be a decent amount of overlap. While I haven’t used anything from them specifically, I do feel like seeing these threads has influenced this piece and some of the things I looked at, so I’d recommend giving them a follow.

A look back at 2018/19

After finishing 7th, qualifying for the Europa League and taking 4 points from Chelsea, Manchester United and Arsenal, it became easy to forget Wolves had a pretty torrid run after the first few games in 2018/19.

Wolves returned from the October international break having won four of their previous six, with the other two games being draws against the Manchester clubs. Instead of building on this, however, the West Midlands side went on a six-game winless run, which included losses to struggling sides like Huddersfield, Cardiff and Brighton.

There wasn’t a huge concern for Wolves. Following the defeat to Cardiff, they were 13th in the table and 7 points clear of the relegation zone, but they did make some alterations.

In the game immediately after Cardiff, Wolves recorded a 2-1 home win against Chelsea, where Morgan Gibbs-White was brought into the side.

Teenager Gibbs-White was played centrally, with Jiménez and Jota either side of him. It looked as though he was tasked with stopping the supply to Chelsea’s #6, which was Cesc Fàbregas rather than the expected Jorginho for this game.

While Wolves got the result thanks to a couple of good counter-attacking goals, it was somewhat fortunate with the xG for the game 1.4 – 1.0 in Chelsea’s favour.

Despite the win against Chelsea, the next game, away to Newcastle, saw Wolves revert to the 3-4-3 but this time with Adama Traoré playing centrally and Jiménez rested. Salomón Rondón may have caused some problems, but Wolves put in a much more dominant performance, particularly after Newcastle went down to 10-men just before the hour mark.

Following this, Wolves opted for an extra midfielder in the next three games, with one of Gibbs-White or João Moutinho playing just behind the front two. I’d argue with Gibbs-White playing behind the front two, it was still closer to the 3-4-3 than the 3-5-2 that became prominent in the New Year, with the teenager often being closer to the front two and likely to make runs into space, whereas Moutinho would play closer to the midfield.

However, it still wasn’t time for the 3-5-2 that became prominent. Wolves went back to 3-4-3 for the away win against Tottenham but played Leander Dendoncker in midfield with Rúben Neves, giving the Belgian international his first Premier League start. Wolves followed up this game with a dismal home loss to Crystal Palace, playing 3-4-3 again, before going playing 3-5-2 with the Neves, Moutinho and Dendoncker midfield against Manchester City away on the 14th of January 2019.

This version of the 3-5-2, the one that Wolves eventually stuck with, got off to a false start as they lost 3-0 to Manchester City. However, it wasn’t a fair test as Willy Boly got sent off after just 19 minutes.

After the City game, Wolves stuck with the 3-5-2 and went on to win their next three games, against Leicester, West Ham and Everton, scoring 10 goals in the process. The shift to the 3-5-2 was praised and Dendoncker was seen as a good addition to the line-up. He not only bolstered the midfield, he also got himself his first goal in the away win against Everton.

From this point, it’s a ‘and the rest is history’ moment. Wolves went on to finish 7th, the 3-5-2 was in and everyone was happy.

The question worth asking though is: Did Wolves improve because of the 3-5-2?

This is where things (hopefully) get a bit more interesting, after giving what was probably too much background.

If we count the game against Manchester City as the first showing for the 3-5-2, the underlying numbers before and after the change make for some interesting reading.

Before the game against Manchester City, Wolves had an xGD of +0.42 per game (penalties are included for all of of this). While I can’t go back and check where this placed them at the time if they maintained this throughout the season, it would have put them in 4th place for xGD. The problem was Wolves were hugely underperforming these numbers. They scored 23 from 27.5 xG For and conceded 25 from an xG Against of 18.7, placing their actual GD at -0.1.

So, Wolves were doing fine with the 3-4-3 but just (presumably) had some bad luck at either end of the pitch. Did they improve with the 3-5-2 then? Or was it just a case of their actual numbers coming into line with their expected numbers, while the 3-5-2 got all the praise.

Well, in terms of xG, it looks like the latter. Looking at the second half of the season, after (and including) the Manchester City game, Wolves’ xG For took a small dip from 1.3 to 1.2 per game, while their xG Against rose from 0.9 to 1.1 per game.

Combining these, their xGD dropped to just +0.10 per game. Looking at the final table, if they had this across the course of the season, they’d have had the 8th best xGD in the league. Which is by no means bad, particularly for a newly-promoted side, but it’s not quite the heights they were performing at before.

During this run, they still conceded more than expected, with 21 goals conceded from an expected 18.9, but over-performed at the other end of the pitch, scoring 24 from 20.6 expected.

Back to the Future

Moving back into the current season, Wolves’ xGD stands at +0.16 per game. They’ve flip-flopped between the 3-4-3 and 3-5-2 but I haven’t split it due to the small number of games played so far. Regardless, it’s much closer to their record in the second half of last season than the first half.

Granted, this season they’ve had the Europa League to contend with, which can have a considerable effect given their small squad, but it’s still disappointing after how strong they looked last season.

However, Wolves have put in two strong performances with the 3-4-3 in their previous two games, making it seem as though they’ll look to stick with it once more. This raises a couple of interesting points.

What does this mean for Leander Dendoncker?

My worry when writing this was that it’d come off as me bashing on Dendoncker, which isn’t my intention, it’s more about his role within the system than his individual qualities. While I quite like Dendoncker, it is questionable just what he added to this Wolves side when looking at 2018/19 numbers.

On the ball, Neves, Moutinho, Boly and Conor Coady were the main sources of progression for Wolves, with Neves leading in passes to the final third by some distance. Neves attempted 11.1 of these passes per 90, Moutinho attempted 8.3 per 90, whereas Dendoncker attempted just 3.0 of the passes per 90.

But that’s okay, being in a midfield with the likes of Neves and Moutinho, there wasn’t a huge emphasis on him to progress the ball. He was brought in to bolster the midfield and add some solidity. But even then, he was involved in 8.39 defensive duels per 90, sandwiched between Moutinho’s 9.6 and Neves’ 7.53. Then, both Neves and Moutinho recorded over 8 ball recoveries per 90, while Dendoncker recorded 6.91 per 90.

This plus the fact that Wolves’ xG Against also rose with the 3-5-2, makes it seem Wolves didn’t tighten up the defence by having him as an extra midfielder.

Okay, so he wasn’t hugely progressive on the ball or particularly active without it, but maybe he added something in attack, I mean he scored a couple of goals.

Well… His non-penalty xG + xA of 0.24 per 90 wasn’t terrible, it was above both Neves and Moutinho and behind only attackers and Matt Doherty, but I’m not sure if it was impressive enough to warrant his place in the side.

So, where does that leave Dendoncker if Wolves do revert to the 3-4-3? Neves and Moutinho seem to be the best pairing in midfield, but Dendoncker could be an improvement in the back three for Wolves.

An improvement on Ryan Bennett in the right centre-back spot is something I mentioned when looking at possible transfers for Wolves in the summer. With Matt Doherty getting forward, often making himself an option in the box, the right centre-back tends to have a lot of space to defend, which isn’t a good match for a defender who isn’t mobile. It’s likely part of the reason Bennett collected so many yellow cards last season.

Dendoncker played in that role in the second half of the Newcastle game before starting there against Arsenal and Aston Villa. It’s the position I thought he was initially brought in to play and while it may be early, he does look like an upgrade on Bennett for that role. I’ve mentioned before I’m not very good at assessing centre-backs, but Dendoncker looks both more mobile than Bennett and more comfortable on the ball.

There’s also not a huge emphasis on him to progress the ball in this role. He’s got Coady to his left and Neves in front of him, though he is still able to pull out a switch of play or a ball into an attacker now and then.

Now Ryan Bennett has returned from injury and will have two weeks recovery over the international period, it’ll be interesting to see whether Nuno continues with Dendoncker as the right centre-back or give Bennett his spot back.

The form of Adama Traoré

Adama Traoré has received a lot of attention and praise for Wolves this season, particularly after his two goals to win the game against Manchester City.

I’ve heard lots of commentator’s talk about how well he’s developed or how well Nuno has done to help his progression. However, my lukewarm take on Traoré is that he’s the same player as last season, Wolves are just utilising him better.

His underlying numbers are improved but not by as big an amount as you might think given he’s ‘added end product to his game this season’. Last season he had 0.12 non-penalty xG per 90 and 0.16 xA per 90, while this season he has 0.15 non-penalty xG per 90 and 0.23 xA per 90.

Looking at his role within the team, last season Wolves experimented with him central or as one of the attackers in the 3-4-3, which doesn’t feel a natural fit. The role as one of the players either side of the striker feels like a bit of a hybrid between a winger and a #10 with the wing-backs providing the width.

This suits someone like Diogo Jota, but not so much Traoré, who feels much more like a traditional winger. He likes to pick the ball up, beat his marker and look to pick someone out in the box, rather than pop up in pockets of space and link up with other players in halfspaces.

With Matt Doherty being injured at the start of the season, Traoré started at right wing-back in the opening few games of the Europa League and Premier League and played well. There’s going to be some compromise given he’s never played in defence before (as far as I know), but it essentially meant he had the right flank to himself and at times he felt like the only source of creativity for Wolves.

With this in mind, once Doherty returned to fitness, it became difficult to drop him. Even if he was dropped, he was often turned to when Wolves needed a change, both in the Europa League and Premier League. However, Doherty is also an important part of Wolves’ play, had a strong 2018/19 and is defensively more solid than Traoré, making it hard to drop him too.

The 3-4-3 allows for them both to play, while also helping to bring out the best in both of them. Rather than have Traoré tuck in, he can stay wide in the final third, while Doherty can roam inwards and even move into scoring positions.

One example against Arsenal can be seen below. Traoré received a cross-field ball out wide, with Doherty moving inwards and leaving Traoré 1v1 against Kieran Tierney.

It’s not to say Doherty never goes wide or Traoré never gets the ball inside, but it does feel as though there is an emphasis on Traoré staying wide and getting him 1v1 against the full-back. Which feels like it’d be a silly thing not to try and do given his dribbling ability.


Wolves had stronger xG numbers with the 3-4-3 than the 3-5-2 but were underperforming these numbers. The switch to the 3-5-2 coincided with Wolves’ better form but it doesn’t look as though the 3-5-2 was responsible for it, which the narrative seemed to be at the time (admittedly I hadn’t looked at the numbers before this and went along with the narrative). Wolves’ xG numbers took a dip with the 3-5-2 and these numbers continued into 2019/20, eventually leading them back to the 3-4-3.

For what it’s worth, the idea of the 3-5-2 wasn’t a bad one. There were concerns about Wolves getting overrun in midfield, which it helped to quell, while it also produced a good partnership between Raúl Jiménez and Diogo Jota. However, at times it often felt like the team were too reliant on the front two. There was little goal threat elsewhere in the team and it sometimes felt like nine defenders and the front two. If neither of those two could come up with anything, it felt like the team couldn’t.

There’s also an argument that the 3-5-2 allowed Neves to collect the ball from the defence and have more time to progress the ball, while also giving Moutinho a bit more license to influence in the final third, with Dendoncker there as extra protection.

Neither of these look true in the numbers though, Moutinho didn’t have a significant change in passes to the final third or the penalty box per 90 in each system, despite playing slightly more passes per game in the 3-5-2. Neves meanwhile attempted 2 more passes to the final third per 90 in the 3-4-3 compared with the 3-5-2.

With all things considered, Wolves’ decision to revert to 3-4-3 looks to be a good one. They’ve had stronger xG numbers with the system, have good combinations out wide with Traoré and Doherty on the right and Jota and Jonny on the left, while it also means they can play Leander Dendoncker in place of Ryan Bennett, which looks to be an upgrade.

The hope for Wolves is that they keep the 3-4-3 and a slightly more front-footed approach from this point onwards, which was seen in the games against Arsenal and Aston Villa. The next four games see them travel to Bournemouth and Brighton while hosting Sheffield United and West Ham. All games in which you’d hope Wolves will continue with a more positive approach and look to push themselves into the top six – which they still only trail by one point despite just three wins this season.

xG Data from fbref, which uses StatsBomb values.

Leave a Reply

Your email address will not be published. Required fields are marked *