Manchester United have forked out £70million on Casemiro’s transfer fee alone.
A lot of money has already been spent this summer, with 108 signings completed so far by the 20 Premier League clubs.
Per Transfermarkt.com, Chelsea have been the biggest spenders with £181million in transfer fees. Next are Nottingham Forest, whose promotion-winning squad has had a massive overhaul with £133million spent on 16 new players — with nearly a week still to go until Thursday night’s deadline.
The question is… where big money has been spent, could it have been spent more wisely?
Data has become increasingly important for player recruitment among professional clubs.
While it should rarely be used as a crutch, advanced data modelling can be a crucial tool in the decision-making process, as Dr Ryan Beal, chief executive and co-founder of SentientSports Recruitment Analytics, explains to The Athletic.
“We look to be forward-looking analytics — trying to find players that might be the right fit for a club. You’ve got a lot of players who are good enough for the Premier League, but who is right for which team?,” Beal says. “We use artificial intelligence not to find players and make decisions, but to add a layer of risk assurance to the decision makers at clubs.”
So if we were to put our faith solely in the data, which alternatives could some of the biggest moves have gone for?
Let’s start at Old Trafford.
Manchester United moved very quickly for Casemiro after that 4-0 thrashing by Brentford, particularly considering their long-standing interest in Barcelona’s Frenkie de Jong — also a midfielder, but one with a wholly different playing profile. Considering the fee and the high wages that Casemiro will be on at United, who might have been more cost-efficient?
Using SentientSports’ analytics model, we can look at some alternatives to the 30-year-old, 63-cap Brazil international — filtering by age, transfer value and Premier League level to create a shortlist of players.
Of course, some players will be off the table having moved clubs this summer already, but the analysis shows the options that were available when the transfer window officially opened in June.
As you can see below, Casemiro does profile as a suitable fit for United, but there might have been some younger, cheaper alternative options in the market.
Specifically, Atalanta’s much younger Teun Koopmeiners (24) and Inter Milan’s marginally younger pass-master Marcelo Brozovic (he will turn 30 in November) profiled well as players who would suit United in terms of their tactical fit for new manager Erik ten Hag’s system, and their predicted chemistry with their new team-mates.
While Koopmeiners only made the move to Italy from AZ Alkmaar last summer, his age profile and tactical suitability might have been a longer-term, cost-efficient option for former Ajax coach Ten Hag — especially as a midfielder schooled in the Dutch philosophy of playing.
We know what you’re thinking… How exactly do you quantify chemistry?
“Chemistry aims to assess how the new transfer will link up with new team-mates, which then correlates nicely with how a player is involved in build-up, and their pass percentage, and involvement in the team,” Beal says. “For tactical fit, it’s more of a red flag to say that a player may take more time to adapt to the new team’s tactical system, and then it’s up to the human scout to say, ‘Can they make those adaptations?’.”
Put simply, player chemistry is a great way of looking at how they are going to fit in the long term, but tactical fit can reflect how well a signing might adapt in the short term to hit the ground running.
So, now we’ve nailed the definitions, let’s have fun with some more options.
Given his position, many have questioned Martinez’s height within the Premier League, but the 5ft 9in (175cm) Argentina international is understandably deemed to be a highly suitable fit for Ten Hag’s tactical setup, having played under him in his three seasons at Ajax.
While Pau Torres of Villarreal and Inter Milan’s Alessandro Bastoni may have been strong candidates for United’s shortlist, Martinez is predicted to prove a strong piece of business for Ten Hag in the long run.
Of course, the model predicts that Ten Hag has to get United playing like his Ajax teams to get the most out of Martinez — which will take time.
“You’ve got to look at what the manager was doing in his old club, and currently United are ‘between two managers’. They’re trying to play like Ten Hag’s Ajax, but they’ve still got that hangover from old managers,” Beal says. “Their own style is somewhere in the middle at the moment, so we’ve used how well players will fit once they play like Ten Hag’s Ajax — and it’s that transition that’s really hard to model in data.”
At the other end of the pitch, we can look at United’s need to reinforce their attacking options — a topic that has been one of the worst-kept secrets of the summer.
The Athletic understands that United are confident of securing a deal for another Ajax player, Antony, and SentientSports’ model suggests the fit would again be highly suitable. Of course, the Brazil international would assimilate well with Ten Hag’s tactical setup but he is also predicted to combine well with his prospective team-mates.
While Rennes’ Martin Terrier could have been a good, more cost-efficient option after his 21-goal haul in Ligue 1 last season, United’s biggest transfer dealings look to make a lot of sense in the data with the signings of Casemiro, Martinez and (potentially) Antony.
Staying in the attacking third, Chelsea’s need for a clinical finisher has been laid bare in the early part of the season, with head coach Thomas Tuchel’s side wasting chances that have cost them points in their past two games.
Pierre-Emerick Aubameyang has been tipped as the answer to Chelsea’s finishing struggles, but is the former Arsenal forward a suitable fit? Per SentientSports’ model, the data would suggest the transition would not be quite as seamless as people might think.
Leading the line supported by two forwards was something that Aubameyang adapted to well after moving to Barcelona from Arsenal halfway through last season, so the tactical fit may well allow him to assimilate within Tuchel’s system. He is proven in the Premier League but naturally might take time to adapt to Chelsea’s attacking play if he were to make the move back to London.
In the chart below, the 33-year-old is compared to attacking alternatives, who are under 30.
For good measure, we can also see that Sterling’s adaptation to his new side is fairly strong from a tactical perspective, but might need a little more time to build that relationship with his Stamford Bridge team-mates.
Aubameyang’s Camp Nou team-mate Memphis Depay has been linked with a return to Manchester United, his home for 18 months from summer 2015, but the suitability of the Dutchman could also be worth a late swoop from Chelsea — especially given the positional versatility that the Netherlands forward could offer.
Meanwhile, the risk of signing Anthony Gordon is reflected in SentientSports’ model.
Naturally, Gordon will be asked to play a different style at Chelsea compared with current club Everton, having to adapt to playing with different types of players. The 20-year-old’s ceiling is high in terms of ability, but he would simply need longer to integrate. The model predicts he would combine well with Mason Mount and Sterling but may require time to link with wing-backs in a new system.
In defence, Chelsea’s quest for centre-back reinforcement has been clear, having lost Antonio Rudiger and Andreas Christensen on free transfers this summer.
SentientSports’ model accounts for the difference between the two clubs’ styles when assessing Fofana’s suitability — specifically in relation to the height of their defensive line and their pressing approach.
Fofana is capable of playing in a back three for Leicester, but his adaptation to a Tuchel system is taken into account within the model — particularly in comparison with other centre-back options who are already used to playing in a top European-level side.
While Koulibaly’s red card against Leeds last weekend blighted his start to life in a Chelsea shirt, there is confidence that the 31-year-old will settle well in his new surroundings. If Chelsea hadn’t been able to agree a deal for Fofana, the model suggests Borussia Dortmund’s Manuel Akanji would have been a suitable alternative. Interestingly, Leicester have been linked with a move for Akanji should Fofana leave…
At 27 years old, either club would be getting a player in his prime, with Champions League experience, who is able to play on the left or right side of defence.
Chelsea appear locked in with Fofana now, but, given the fees involved, might it have been sensible for the club’s new co-owner Todd Boehly and company to pivot away from Fofana?
Attacking options have been the order of the day for a lot of Premier League clubs this summer.
The 23-year-old Italian is yet to start in the Premier League, so time will tell on how well he integrates into David Moyes’ system, but he has looked lively in the Europa Conference League qualifiers. The model suggests that Scamacca is in the bracket of a suitable fit for his new side, but it is interesting to see Amine Gouiri profile as the outstanding candidate for West Ham’s style — with Moyes believed to be an admirer of the Nice centre-forward.
As a potentially more cost-efficient option per SentientSports’ model, Alkmaar’s Vangelis Pavlidis profiles well after scoring 16 goals last season, while Joshua Zirkzee had a breakout season with 15 goals on loan at Anderlecht, before returning to parent club Bayern Munich. All are academic options following West Ham’s signing of Scamacca, but an interesting data exploration nonetheless.
Finally, Frank Lampard has made clear how Everton are in desperate need of attacking reinforcements. Dominic Calvert-Lewin’s ongoing injury struggles have been compounded by the sale of Richarlison to Tottenham and the potential transfer of Gordon to Chelsea.
Everton signed Brighton’s Neal Maupay on Friday but that’s unlikely to be the end of their efforts to strengthen their attack — so who could be the most suitable to get through the Goodison Park door in the final week of the window?
In terms of style, Brentford’s Ivan Toney profiles as the outstanding candidate on the shortlist, but any feasible conversation would have to begin with the acceptance of having to part with a large sum of money. Southampton’s Che Adams looks a suitable fit for Lampard’s style, and is a player The Athletic understands Everton have held an interest in.
A late swoop for 6ft 7in Sasa Kalajdzic could be a very sensible option. The Stuttgart striker has been strongly linked with a move to the Premier League, with Wolves believed to have shown interest. Kalajdzic would very likely be a more cost-effective option, and with his profile similar to that of Calvert-Lewin, the transition could be seamless.
Interestingly, Maupay profiles ‘okay’ when you look at the graphic below and how likely he is to gel within Lampard’s system and squad.
A player’s value is ultimately whatever the buying club are willing to pay, and it is no secret that Premier League sides often pay a premium compared with leagues across continental Europe, given their eye-watering financial clout.
Without fail, there will always be some hits and some misses in the market, but predicting where they may occur is where the data can help.
“At clubs, you’ve got the use of data and artificial intelligence to improve decision making through risk analytics,” Beal says. “But it’s got to be that hybrid approach where you can filter players down using data, scouts then go and watch them, and then you can combine it all at the end to make that final decision.
“That’s really where data will be used most within recruitment in football.”
(Photos: Getty Images/Design: Eamonn Dalton)