Is targeting xG overperformers like Cunha and Mbeumo a sensible or risky transfer strategy?

New York Times

19 hours ago

Now more than ever, Manchester United need players who transcend expectations, and their recruitment strategy so far this summer reflects this.
United have already reached an agreement to sign Matheus Cunha from Wolves after activating his £62.5million release clause. Brentford's Bryan Mbeumo could follow, with The Athletic's David Ornstein reporting that United have made an initial £55million offer after the player made clear his desire to join.
Advertisement
Cunha and Mbeumo have impressed since arriving in English football with a catalogue of dazzling performances. But they've also stood out in a colder, more technical sense as the Premier League's most ruthless finishers this season, both comfortably beating their expected goal tallies.
Mbeumo scored nearly twice as many goals as expected, netting 15 non-penalty goals from 7.6 xG. Cunha followed closely behind, converting 8.7 xG into 15 goals — the league's second-highest overperformance in 2024-25.
Their razor-sharp finishing obviously deserves praise, and it's a trait United sorely lacked this season. But by targeting the league's biggest statistical outliers, is their transfer strategy betting on unsustainable hot streaks rather than sound underlying processes?
The Athletic explores whether xG overperformance should be something to pursue or a recruitment red flag.
Let's start with Mbeumo. His range of goals this season covered the full spectrum: smart volleys, curled finishes, thumping near-post drives, poacher's tap-ins, and yes, a healthy dollop of fortune, including an overhit cross against Ipswich Town that floated in.
Mbeumo the match-winner back in October 🤯
Last time out against Ipswich ⏪ pic.twitter.com/VedWXrcOFK
— Brentford FC (@BrentfordFC) May 9, 2025
His golden touch more than filled the goalscoring void left by Ivan Toney's departure to Al Hilal, propelling Brentford to 56 points, their highest-ever Premier League total.
But how repeatable was that output, given the quality of chances he had?
To estimate this, The Athletic simulated Mbeumo's season 100,000 times using the individual xG values of his 79 non-penalty shots. An average finisher, given the same chances, would be expected to match his tally just once every 200 seasons or so, with only 0.4 per cent of simulations producing 15 goals or more. According to the model, it was more likely he'd score two. Cunha, on the other hand, had a 1.7 per cent chance of achieving his tally or more, due to his smaller level of overperformance.
It does seem extremely unlikely that either player could repeat their 2024-25 output (from a similar quality of chances) in future seasons. But xG is just a model. Finishing is influenced by a messy cocktail of factors including distance, angle, pressure, goalkeeper positioning, shot technique, individual skill, and a garnish of luck — no metric can perfectly capture them all.
Advertisement
Football is a low-scoring sport, and shot data, particularly on the player-specific level where sample sizes are small, is subject to large amounts of variance. If the model underestimates Cunha's chance conversion by 25 per cent, his 15-goal tally would be reached in a more plausible 10 per cent of simulations.
Still, there's one pattern that does hold firm: regularly outpacing your xG isn't something most players can sustain. The graphic below compares Premier League players' xG overperformance from one season to the next, and… there's virtually no correlation.
Of course, there are exceptions. Some players' finishing ability is such that they consistently outperform their expected goals. The most prominent example is Tottenham's Son Heung-min, who has scored 97 goals from an xG of 72 since the 2018-19 Premier League season. Other notable outperformers include Harry Kane, Phil Foden and Kevin De Bruyne.
Mbeumo isn't in that bracket, at least not yet. This was the first time in his four Premier League seasons he's outperformed his xG. Could it be that his finishing has just drastically improved this season? Probably not, but there has been a change in his shooting patterns this season.
In previous seasons, Mbeumo's shots were spread fairly evenly across the box. This year, they clustered tightly around the right edge of the penalty area. Cutting inside and curling efforts into the far corner, as he did in Brentford's 4-0 away victory to Leicester, has become his signature finish. It's possible he's genuinely better than the average player at this particular shot type. Whether he can get into those same positions as often at a team like United is a different question.
For Cunha, the story is a little different, as he finished above expectations in both of his two full Premier League seasons. But digging into his shotmap for this season and a pattern emerges. Five of his goals have come from outside the box, the joint most with new team-mate Bruno Fernandes, including one against United direct from a corner. Either Cunha is an elite ball-striker outside the box, or more likely, this ability to score from range will stabilise in the future
The danger with overperformance is overpaying for a scoring streak that doesn't last. Brentford manager Thomas Frank has already referenced Mbeumo's '20 goals for a mid-table' club as a reason why 'if he is sold, it must be for a nice, nice amount'. It's easier to market a player's realised actual goal return rather than a less flattering statistical estimate.
One notable cautionary tale is Timo Werner's £47.5m transfer to Chelsea from RB Leipzig in 2020. He was signed off the back of a season in which he massively overperformed, scoring 28 goals from just 21.2 xG. But the German international's shooting boots never made it to London and he has underperformed his xG by 11 goals during his Premier League stint. Liverpool signed Christian Benteke after a similar purple patch in his final season for Aston Villa, and like Werner he failed to replicate this at a bigger club.
A more recent counterexample is Omar Marmoush, whose clinical striking has persisted after a move from the Bundesliga to Manchester City, scoring seven league goals since joining in January.
In the case of Mbeumo and Cunha, concerns about overperformance obscure a more important truth: both post strong underlying xG numbers. They're overperforming relative to a high base, which is especially impressive for players who don't operate as out-and-out strikers in elite, chance-heavy sides.
Advertisement
Finishing can fluctuate but xG creation tends to hold. The ability to consistently get into high-quality scoring positions is what should really be valued, rather than finishing purple patches.
Focusing too narrowly on finishing obscures the other aspects of play that make both players attractive to United. For example, both are skilled at laying on chances, something sorely needed for struggling striker Rasmus Hojlund, who recorded the lowest xG per 90 of all Premier League strikers this season.
The scatter graph below shows each Premier League player's percentage contribution to their team's total xG and xA, with Cunha and Mbeumo standing out on both fronts. But they also bring other dimensions to United's attack not reflected in these statistics, such as Cunha's ability to drive the ball up the pitch and Mbeumo's physicality in aerial duels.
As Liverpool's former head of research Ian Graham wrote in his book How to Win the Premier League, one job of data analysis is 'to separate long-term underlying performance from short-term unpredictable performance' and that 'most teams cannot tell the difference between signal and noise'.
Mbeumo and Cunha's finishing overperformance? Likely unsustainable noise, particularly in Mbeumo's case. But the signal — their ability to consistently get into high-quality scoring positions and create for others — is very real. Mbeumo's 9.3 expected assists led the entire Premier League, while Cunha's non-penalty xG surpassed every player in United's current squad.
United may be paying a premium for the noise, but in this case, they'll still be getting the signal. And for now, that should be good enough for Ruben Amorim.
(Header photos: Getty Images)