After buying Wolves in 2016 and investing roughly £30m in players (~£15m each window) Fosun were disappointed with a 15th place finish. Their late decision to remove Kenny Jackett and replace him with Walter Zenga proved to be a misstep while replacement Paul Lambert didn’t have the desired effect of propelling the club up the table. Rather than giving Lambert his first pre-season with the squad and allowing him to make necessary changes Fosun acted again, replacing the Scotsman with Portuguese manager Nuno Espirito Santo. Nuno comes with a fairly impressive CV but did Lambert deserve more time?
Wolves under Lambert
In some ways Wolves were unlucky last season. Their Expected Goal Difference (xGD) per game was 0.22, the 7th highest in the division. The Championship can be strange at times (think about how Huddersfield went up with a negative goal difference) but this is encouraging. The six teams above them are Newcastle, Sheffield Wednesday, Brighton, Bristol City, Fulham and Hudderfield. Of these six Bristol City are the only ones who didn’t actually finish in the top six come the end of the season.
Wolves had the bad luck of under-performing their xG in front of goal and clubs over-performing theirs when attacking Wolves. Their xG Rating of 0.898 was the 5th worst in the division. Only Burton, Rotherham, Birmingham and Bristol City had worst luck in front of goal this season, while their xG Against Rating was 1.164 meaning they conceding 16.4% more goals than ‘expected’ – the 8th worst in the divison.
Looking at the differences in Lambert’s term compared to Zenga’s brings up some interesting facts too. Wolves’ xGD per game took a fairly big turn for the better with Lambert compared to Zenga with Lambert’s Wolves having an xGD of 0.28 and Zenga’s having 0.16 per game. Lambert’s xGD puts Wolves just ahead of Huddersfield as the 6th best xGD in the division while Zenga’s Wolves stay in 7th. Interestingly though the bad luck in front of goal didn’t really come in to play until Lambert took charge.
Under Zenga Wolves marginally outperformed their xG with a rating of 1.012 yet under Lambert this took a huge plunge to 0.846 despite Wolves on average creating better chances, 1.379 xG per game compared to Zenga’s 1.173. Was Lambert’s under-performance just down to bad luck?
If Wolves had finished at the rate of the average shot in the Championship under Lambert they ‘should have’ scored 15% more goals than they did while at the opposite end conceded 10% less than they ‘should have’. Had Lambert been given more time and the chance to make a few alterations could Wolves have reverted to the mean and starting scoring the goals they were ‘expected’ to score?
However, you’ve heard it time and time again, football is a results business. It seems unlikely that when the ‘footballing review’ that led to Lambert’s dismissal came around they were discussing how they’d be unlucky given their expected goal values. Lambert took Wolves from 19th to 15th while Fosun were hoping to make a push towards the play-offs.
A big part of Lambert’s downfall was his record at Molineux. Looking at the Championship table since Lambert took charge and Wolves have the 6th best away record, taking only 6 points less than Fulham and Newcastle who attained 27 points in the time period. The problem is that Wolves had the 6th worst home record in this time, taking just 20 points and losing 7 from 15.
Wolves could have played exactly the same as they did this season but managed to score at an average rate and possibly found themselves chasing the play-offs. Bad luck at either end of the pitch has effectively cost Lambert his job. Though it would have been nice to see Lambert given more time and the chance to make some alterations to the squad it won’t be happening now. Wolves have hired their 4th manager in a year with Nuno Espirito Santo.
As soon as Nuno became available Fosun were said to want to bring him in to replace Lambert and looking past all the business with Jorge Mendes it’s not hard to see why. The former Porto man comes to Molineux having taken Rio Ave into the Europa League and attaining a 4th place finish with Valenica. Cynics will be quick to point out that he hasn’t lasted long in any of his jobs and it is worth bringing up. Do Fosun see Nuno as the long-term solution to their plan at Wolves or is it going to be Watford 2.0 where each year a new manager is brought in?
At Rio Ave he performed well by getting them into Europe and two cup finals then Valencia came calling. Valencia are a bigger club than Rio Ave and offered a step-up so the move makes sense. It all looks good as he performs well, finishes 4th and signs a 3-year contract extension. However after a bad start to the following season he resigns before taking the reigns at Porto in 2016. At Porto it’s not as though he did terribly, losing only 2 games in the league and going out in the Champions League Round of 16 to eventual finalists Juventus, but at a club like Porto no silverware in a season is seen as a disaster.
The only concerning part about the above is how he resigned after the start of the season didn’t go his way at Valencia. Whether this was due to something behind the scenes I don’t know but it doesn’t fill me with confidence that pretty much the only time in his short managerial career things didn’t go as planned he walks rather than reacting and turning things around.
I only have detailed data on his Valencia side in 2014-15 so that’ll be the main point of focus here. Straight off the bat though, what’s encouraging is that when he finished 4th with Valencia it wasn’t a fluke. Their xGD in 2014-15 was the 4th highest in the league, with the top 4 in xGD being the top 4 in the table; Barcelona, Real Madrid, Atletico Madrid and Valencia. They had a good portion of luck at either end of the pitch with an xG for rating of 1.2 and xG against rating of 0.73 but it’s more a case of this reinforced their 4th place position rather than luckily handed them it.
Defensively they had the third best xG against per game behind only Barcelona and Atletico Madrid and only conceded 15 open play goals all season. Looking at Valencia’s defensive set-up when compared to Championship teams this season is interesting. Their average position of winning the ball back is below average yet they allow significantly less passes in the opposition half per defensive action than any Championship teams. Idea for pressing graph from @FootballFactMan, his original version is here.
Their directness in the final third was also similar to Wolves’. On average Wolves have 6.543 passes in the final third per shot, Nuno’s Valencia had 6.495. If his Wolves side line up anything like his Valencia side then it could mean there’ll be lots of work when not in possession, looking to win it back quickly and get chances on goal.
Watching a few clips of his Valencia side reinforced this. When they lost the ball they aimed to get it back quickly and when in possession they generally built using combinations trying to make overloads in wide areas with both full-backs pushing forward and the deepest midfielder dropping between the two centre-backs.
This is an approach that can suit Wolves. Wolves have exceptional wide players in Helder Costa, Ivan Cavaleiro, Jordan Graham and Ben Marshall while also possessing attacking full-backs Dominic Iorfa, Matt Doherty, Silvio and Conor Coady and have both Romain Saiss and Jack Price who are comfortable collecting the ball between the two centre-backs and orchestrating play.
It’s only a few short clips from Valencia’s game against Celta Vigo but below you can see how they aim to combine out wide and win the ball back quickly:
If he can replicate the home form he had with Valencia and Porto (1 loss with Valencia in 2014-15, none with Porto last season) and get the team playing how he wants Nuno could be a great hire. With Fosun flip-flopping between a Championship experienced manager and a continental one initially backing Jackett only to hire Zenga, to then revert back to Lambert and now go for Nuno, they’ll be more hopeful than ever that Nuno will work out.
Squad and Transfers
After the spending spree last year, Jorge Mendes’ involvement and no one really knowing exactly who is bringing players in Wolves transfers are under the spotlight more so than ever before. Now with Fosun having experienced a year of being in charge at a club they’ll be looking to make sure any money spent this year is money spent to return Wolves to the Premier League.
Defence is the only area where Wolves have recruited so far this summer bringing in Portuguese centre-back Roderick Miranda from Rio Ave, right-back Phil Ofosu-Ayeh and another centre-back in Ryan Bennett. Despite such quick and heavy recruitment Wolves’ defence didn’t do too badly last season with the 10th best goals against per game and the 9th best xGA per game. However with Mike Williamson leaving and the Richard Stearman loan deal not being made permanent some reinforcement was to be expected.
With the arrivals being aged 25, 26 and 27 you’d imagine they’re coming to make an instant impact. If Silvio signs a new deal Wolves now have two player for each position in the back line, even if Conor Coady moves back to his more natural midfield spot.
At right back there’s the aforementioned Coady, plus youngster Dominic Iorfa who fell out of favour from Lambert making only 3 starts after the turn of the year and finally new boy Phil Ofosu-Ayeh. Left-back has right-footed Matt Doherty and Silvio while centre-back has Danny Batth, Kortney Hause, Roderick Miranda and Ryan Bennett. It’ll be interesting to see who Nuno settles on as his first choice back four.
Given the choices Nuno now has after these signings Wolves don’t really need another defensive signing this summer unless someone else is to leave.
Miranda looks like he could be a good signing for Wolves this season with the clips below showing him as a centre-back comfortable bringing the ball out from the back:
Midfield is where most of Wolves’ highlights came from last season. Their two top scorers last season were Dave Edwards and Helder Costa both with 10 goals each while the latter also chipped in with 8 assists. Costa’s signing proved to be the most successful of last summer, initially arriving on a loan before making a £13m move in January Costa has been the most exciting Wolves player in recent seasons.
As previously mentioned Wolves’ wide options are probably the strongest area in their team. Helder Costa had the 6th highest xG + xA for players born after 1/6/1992 last season (24 and younger during the season) with the top 5 containing 2 Newcastle players and 3 Chelsea loanees, while Ivan Cavaleiro occupied 16th on the same list in a season interrupted with injuries. Jordan Graham is fit after returning from a long-term injury to play on the final day of the season, Graham managed 0.5 assists p90 in 2015/16 before getting injured, then January signing Ben Marshall was 5th for successful passes plus crosses into the box this season.
The centre-midfield line-up will depend a lot on how Nuno lines up, whether it’ll be a 4-3-3 with someone in front of the defence or a 4-2-3-1 with someone just behind the forward.
Should it be the former then a new midfield signing could be signed while for the latter Wolves could probably cope. In the 4-3-3 shape both Romain Saiss and Jack Price offer good options to sit in front on the defence. Price had the second most passes to the final third last season with 13.65 p90, only 0.03 behind 1st place Jonjo Shelvey, while Saiss wasn’t far behind with 11.25 p90. Saiss also had the 13th most possession adjusted interceptions + tackles last season.
In front of one of Price or Saiss could be Edwards and Coady, Edwards’ goal contribution from midfield is vital to Wolves while the signing of a new right-back could see Coady move back into midfield. This is where Wolves could do with a new signing, with George Saville on his way out Wolves would only have Lee Evans as back-up for the central midfield spots.
Burton’s Matty Palmer and Nottingham Forest’s Ben Osborn could both be useful for Wolves, though the majority would have been off set-pieces Palmer had the 3rd highest xA p90 of midfielders u24 in the league last season while Osborn had the 6th highest and can play pretty much anywhere in midfield. Palmer also has the 4th highest xP Rating of u24 midfielders this season and is entering the last year of his contract at Burton.
If a 4-2-3-1 is picked Wolves would have Price, Saiss, Edwards, Coady and Evans for the deeper two roles while younger players Bright Enobakhare, Conor Ronan and Morgan Gibbs-White can all occupy the number 10 role as well as Ivan Cavaleiro and Edwards. Enobakhare in particular put in a great display in the number 10 role on the last day of the season and could possibly feature more next season.
Without new signings this gives Wolves the following midfield options this upcoming season, where you can see not much recruitment is needed in this area:
So far I’ve said the defence and midfield are all in an okay state, so what’s the problem? Well, last season Wolves forwards didn’t have the best of years in front of goal. The graph below shows the bad form of the Wolves strikers last season:
Not one Wolves forward managed to outperform or even come close to performing their xG in front of goal last season. The worst offender was Nouha Dicko with an xG Rating of 0.486 while Andreas Weimann was the best despite only managing a rating of 0.749. Running a Monte Carlo simulation on the shots taken by these four forwards gives us the following histogram:
Not only was the 10 goals managed by the four forwards the second worst score in 1000 iterations of the Monte Carlo simulation it only had a 0.4% chance of happening, the same probability that they ended up with 28 goals! The most probable output was 19 or 2o goals, giving Wolves around an extra 10 goals.
Across the course of the season 10 goals may not sound like a huge amount but given that only 6 of Wolves’ 20 losses this season were by more than a 1 goal margin these 10 goals could have been crucial in turning losses to draws or draws to wins – not to mention this is only if Wolves were average in front of goal. If they had some luck in front of goal they could have had somewhere between an extra 10-20 goals this season.
To compare Monaco scored 90 goals this season excluding penalties and own goals yet a Monte Carlo simulation on their shots this season gives the following:
The 90 goals they actually scored didn’t pop up once during 1000 iterations and in the end the most probable results would have seen them with ~30 goals less than this 90.
Now there could be some reasons behind the under-performance of Dicko and Jon Dadi Bodvarsson. Dicko returned from a long-term injury part way through the season without having a proper pre-season. It may take a pre-season and bit of time before he regains his confidence in front of goal and can revert to mean by scoring around the value of his expected goals.
It’s also worth noting Dicko’s movement in the channels and link-up play help get the best out of players like Costa and Cavaleiro giving them space to move into.
Then this is Bodvarsson’s first season in a new country it may take a while for him to find his feet, if his performance in front of goal next season is the same as this season that’s when it may be time to move him on, especially as even his xG p90 numbers were below average for forwards this season.
A new striker does seem to be on Wolves’ agenda however with Britt Assombalonga recently being linked with a move to Molineux. The Nottingham Forest striker managed 0.593 non-penalty goals p90 last season, 5th highest in the division. His xG p90 of 0.324 is 0.02 lower than that of Dicko’s though which puts the pair at opposite ends of the spectrum.
Assombalonga is out-performing his xG by just under double while Dicko under-performs his by just under half, if they both revert to mean they’ll be scoring the same amount of goals. Given Forest paid around £6m for Assombalonga and he still has 4 years of contract left there’s a good chance the club could demand around £10m for the forward, this means Wolves would be paying £10m for a striker the same age as Dicko who’s ‘expected’ to score the same amount as Dicko.
The only way this would work is if Assombalonga is able to dramatically out-perform his xG two seasons in a row or at Wolves he is given higher probability chances on a more regular basis than he was at Forest. Both are a possibility (Harry Kane has sustained over performance of his xG) but it could be an expensive gamble to take.
Take for example ex-Wolves man Sam Winnall’s January move from Barnsley to Sheffield Wednesday. At Barnsley he was on fire with 0.616 non-penalty goals p90 but he was also hugely out-performing his xG with an xG Rating 1.964. At Wednesday this evened out and his xG Rating dipped to 0.724 and despite his xG p90 increasing by 0.126 p90 his goals p90 was just over half of his Barnsley number at 0.318.
This wasn’t a costly mistake given the TransferMarkt fee of £500k and his xG p90 was higher than Wednesday’s other forwards so he still may come good for the Owls, but he wasn’t the big upgrade they were hoping for. Steven Fletcher and Fernando Forestieri had xG p90’s of 0.403 and 0.430, both only marginally less than Winnall’s 0.440 p90. Given they already had four forwards and his xG p90 was below their current front line Wednesday could have invested this money elsewhere.
If this reverting to mean happens to Assombalonga at Wolves it would have much greater effect given the fee that would most likely be involved.
Another option for Wolves’ forward line could be Osasuna’s Kenan Kodro. I’ve previously talked about Kodro here though since then I have changed my xG method. His performance remains impressive however with the 8th highest xG p90 in Europe’s top five leagues for players u24 last season and second highest in La Liga.
Playing for relegated Osasuna and in his last year of contract it’s a player Wolves could realistically sign for a reduced fee.
Kodro’s xG p90 of 0.402 is not only higher than that of all Wolves’ current forwards it came in the Spanish top division for a struggling side. His xG Rating was 1.069 means he’s only marginally out-performed his xG making it more likely he can sustain his level of goal scoring from last season – 0.429 goals p90. Kodro is not only a year younger than Assombalonga he comes with more impressive stats playing in a tougher division and it’s likely he’d cost less. Having not played in England there may be more of a settling in period than Assombalonga but he could be the better long term buy while offering better value for money.
Overall, Wolves’ strikers (Dicko and Mason in particular) did well at getting into good scoring positions but couldn’t capitalise once there. Wolves may be better off hoping their current strikers revert to mean and bring in a good value signing in Kodro rather than going out and spending big money on a player who’s over performance may not be able to be sustained.
Given Wolves’ heavy recruitment last season, talented young players coming through and impressive xG numbers the Midlands side don’t need a huge amount of change this summer. After the quick dealings in defence centre-midfield and striker are the only areas where Wolves really need to be looking for recruitment.
Nuno has inherited a big squad made even bigger with the youngsters on the brink of breaking into the first team, if Nuno can implement his style of football Wolves have a great base to build from without additional signings. Should Nuno be able to bring in (literally) a couple quality players to enhance the side Wolves could enjoy a good season. It’s just a case of how well the squad, manager and league all take to each other and if Wolves can turn around their bad luck in front of goal from last season.