Bruno Fernandes finished the 2025/26 Premier League season with 21 assists. It was a new single-season competition record, surpassing the previous mark of 20. Premier League data confirms the scale of the achievement.
But there is something unusual hidden inside that record.
Fernandes did not simply create more goals than everybody else. He produced an extraordinary volume of chances — 136 in total, including 98 from open play and 33 big chances — worth 12.3 expected assists (xA) according to Opta’s season analysis.
Yet 21 of his passes became assists.
That difference matters because an assist contains two separate pieces of football. The creator has to manufacture the chance. Somebody else then has to finish it.
Fernandes excelled at the first part. In 2025/26, he also received an extraordinary return from the second.
Across the Premier League’s leading creators, other players experienced the opposite. Enzo Fernández, Yéremé Pino, Adam Wharton and Declan Rice all generated more underlying creative value than their eventual assist totals suggest. The league-wide data reveals just how differently creators were rewarded for the chances they produced.
Fernandes’ record was therefore not one statistical achievement.
It was two layered on top of each other.
An assist isn’t entirely yours
Assists feel like an individual statistic. A player makes the final pass, his team-mate scores and the creator gets the assist.
But change only the finish and the statistic disappears.
Imagine the same through-ball being played twice. On the first occasion, the striker rolls the shot into the bottom corner. On the second, he shoots wide.
The creative action could be almost identical.
Only one becomes an assist.
Expected assists are useful because they look at the chance produced rather than simply whether the shot went in. A pass leading to a dangerous opportunity carries more xA than one leading to a difficult attempt.
It does not mean that a player with 7.0 xA “should” have exactly seven assists. It gives us another way of judging the quality of the opportunities created.
That distinction produced some striking differences in 2025/26.
Among the 25 highest-xA players in FIVDA’s working dataset, Fernandes recorded 21 assists from 12.3 xA. Jarrod Bowen had 11 from 5.2. At the other extreme, Mathias Jensen recorded one from 6.1, Pino two from 6.1 and Enzo Fernández four from 7.3. The underlying figures show players with similar creative values ending up with dramatically different assist totals.
Figure 1 — Assists vs Expected Assists: The Premier League’s Leading Creators
That is the first lesson from the data.
Assists tell us what happened after a creative pass.
They do not isolate the quality of the pass itself.
Fernandes’ 21 assists contain two different achievements
There would be a simple explanation for Fernandes’ record if his underlying creative numbers were ordinary.
They were anything but.
Fernandes created 136 chances, comfortably the highest total in the Premier League. He also led the competition with 98 chances created from open play and accumulated 12.3 xA. His 33 big chances created provide another indication of how frequently his passing produced dangerous opportunities. Opta’s data makes clear that this was an exceptional creative season even before the assists are counted.
That rules out the simplest alternative explanation.
Fernandes did not break the assist record merely because his team-mates converted an ordinary collection of chances.
He was already producing extraordinary creative volume. That emphasis on process rather than final outcome also develops FIVDA’s early-season analysis of Manchester United’s shot creation, when attacking activity was already telling a different story from the goals column.
But the finishing outcome amplified it.
His 21 assists were 8.7 higher than the 12.3 xA in FIVDA’s league-wide working dataset. Bowen had the next-largest positive difference among the 25 leading xA creators at +5.8, followed by Rayan Cherki at +3.1, Harry Wilson at +2.2 and Mohamed Salah at +1.6. The league-wide comparison puts Fernandes’ separation from the other leading creators into context.
Figure 2 — Who Received the Most and Least Assist Reward?
That 8.7 should not be interpreted as “luck”.
It does not mean Fernandes deserved 12 assists rather than 21. Nor does it give us nine assists that can somehow be removed from his record.
It describes a difference between two things: the quality of the chances his passes produced and the number that ultimately became goals.
In Fernandes’ case, both were exceptional.
Other creators had the opposite experience
The same comparison becomes more revealing when we look below the line rather than above it.
Jensen finished 5.1 assists below his xA in the working dataset. Pino was 4.1 below, Enzo 3.3, Wharton 2.5 and Rice 2.3. The same dataset shows the size of those differences.
Again, this does not tell us how many assists those players deserved.
There are too many things between the pass and the final outcome.
The receiving player still has to finish. The goalkeeper can make a save. Different chances have different characteristics, and the model itself is only an estimate of their value.
But the comparison does tell us something important about how we judge creators.
Two players can create chances of broadly similar underlying value and finish the season with very different assist totals.
That is not a theoretical problem. It happened repeatedly in the 2025/26 Premier League.
And Enzo Fernández provides perhaps the clearest example of why even xA does not tell the whole story.
Enzo’s creativity was missed twice
Enzo finished the season with four Premier League assists from 7.3 xA. He also accumulated 6.0 xA from passes that were not crosses and recorded six secondary assists, the highest total in the league. Opta’s analysis highlights both the underlying creation and his involvement one pass earlier in attacking moves.
A secondary assist is the pass before the assist.
That creates two different ways in which an ordinary assists table can miss creative influence.
Enzo could make the final pass into a dangerous chance, only for the shot not to become a goal. The assist disappears, but xA captures the value of the opportunity.
Alternatively, he could play the pass that opens the defence before a team-mate makes the final delivery. In that case the move ends in a goal, but neither the assist nor the xA generated by the final pass belongs to Enzo.
His wider chance-creation numbers reinforce the point. Enzo created 68 chances, the third-highest total in the Premier League behind Fernandes and Dominik Szoboszlai, according to the completed-season Premier League rankings.
Four assists are still four assists.
But they are not a complete description of the season he had as a creator.
Pino shows what happens when the finish never arrives
Yéremé Pino provides a different example.
He finished with two assists against approximately 6.1 xA in FIVDA’s league-wide dataset, one of the largest negative differences among the 25 highest-xA creators. The season data therefore suggests considerably more underlying creation than his two assists alone would imply.
That pattern was visible much earlier in the campaign.
Pino’s role for Crystal Palace involved moving inside and looking for forward runners rather than simply remaining fixed to the touchline. At one checkpoint, his open-play passes had already produced 1.41 xA without generating an open-play assist. Premier League analysis showed how his inward positioning helped him find runners during attacking transitions.
This is exactly the sort of player an assists table can undersell.
The useful pass happens.
The chance happens.
The goal does not.
Calling that bad luck would go beyond the evidence. Finishing, goalkeeping, the precise chances involved and the limitations of the model can all contribute.
We do not need to know exactly why the gap appeared to recognise that two assists tell only part of the story.
Wharton had the same problem
Adam Wharton finished with five assists while the public models in the Writer Package place his xA at roughly 7.3 to 7.5.
Again, this was not simply an end-of-season statistical curiosity.
Early in the campaign, Wharton had already accumulated 1.34 xA without registering an assist, with his set-piece delivery contributing to his creative role. Premier League analysis identified the discrepancy while the season was still developing.
Pino and Wharton did not necessarily create chances in the same way.
That is precisely the point.
An assists column compresses different players, different tactical roles and different types of chance creation into one number.
xA can separate the creative process from the final result.
Tactical context then helps us understand how that creative value was generated.
Set pieces complicate the comparison — but they still count
Declan Rice is a good example.
Rice finished with five assists and 7.0 xA after creating 63 chances. But 3.3 of that xA came from corners and indirect free-kicks, the highest such figure in the league. Opta’s data shows how significant dead-ball situations were to his creative output.
It would be easy to treat that as a reason to discount part of the number.
It is not.
A dangerous corner still creates a scoring opportunity. Set-piece delivery is a genuine football skill and a genuine source of attacking value.
The useful distinction is not between “real” creativity and set-piece creativity. It is between different types of creation.
Fernandes makes that especially clear because he excelled in both.
Eleven of his assists came from set pieces, equalling the Premier League record, but he also led the competition for open-play chances created. Opta’s season review captures both sides of his output.
Under Michael Carrick, Fernandes was also given freedom to drift into right-sided attacking pockets while remaining a major part of Manchester United’s dead-ball threat. FIVDA’s analysis of Carrick’s Manchester United showed the wider tactical change: Fernandes was moved back towards a more natural No. 10 role, allowing him to receive possession closer to goal.
Premier League tactical analysis documented that positioning and his set-piece influence. His 12.3 xA therefore did not come from one repetitive route.
It was the product of a player creating danger in several different situations.
Why the exact xA number changes depending on where you look
There is one final complication.
Expected assists are produced by models, and different models do not always give the same answer.
Opta has Fernandes at 12.3 xA for 2025/26. StatMuse records 11.75. Enzo is 7.3 in the Opta evidence used here but 7.63 in StatMuse. Wharton is similarly placed between roughly 7.3 and 7.5 depending on the dataset. StatMuse’s figures demonstrate those differences.
FotMob’s model produces slightly different totals again while preserving a broadly similar picture among the selected leading creators.
Figure 3 — Why the Exact xA Number Depends on the Model
That does not make xA useless.
It means the decimal point should not be treated as footballing truth.
Different providers can define and value events differently. Even the terminology is not completely interchangeable: FBref’s passing data distinguishes expected assisted goals from expected assists alongside other passing measures.
For this article, the important findings do not depend on pretending those models are identical.
The broad picture remains.
Fernandes combined exceptional underlying creation with an even more exceptional assist return.
Several other leading creators generated considerably more chance value than their assist totals suggest.
And the gap between those two numbers can tell us something useful — provided we do not pretend it tells us why the gap exists.
Fernandes’ own history provides the biggest warning
There is an obvious temptation after seeing 21 assists from around 12 expected assists.
Perhaps Fernandes possesses some repeatable ability to turn his chances into more assists than their underlying quality would normally produce.
His own recent seasons provide a reason to be cautious.
Using one provider consistently, Fernandes recorded six assists from 6.93 xA in 2021/22, eight from 10.74 in 2022/23 and eight from 10.22 in 2023/24. In 2024/25, he moved the other way, producing ten assists from 8.02 xA. Then came 2025/26: 21 assists from 11.75. His five-season record shows substantial movement in the relationship between his creation and eventual assist return.
Figure 4 — Fernandes’ 2025/26 Assist Gap Was Exceptional Even for Fernandes
That history changes how the latest season should be interpreted.
Fernandes has not spent five consecutive years producing far more assists than his xA.
For three seasons, the opposite happened. In the fourth, his assists moved modestly ahead. In the fifth, the difference became enormous.
That does not tell us what will happen next season.
It tells us that 2025/26 was exceptional even by Fernandes’ own standards.
Treating the gap as a permanent individual ability would therefore require evidence that one season cannot provide.
What assists actually tell us
None of this diminishes Fernandes’ record.
He made the final pass for 21 Premier League goals, breaking the competition’s single-season assist record. Premier League records confirm an achievement that belongs entirely in the history books.
The mistake would be asking that number to answer a different question.
If we want to know who supplied the final pass before the most goals, assists are exactly the statistic we need.
If we want to know who consistently created dangerous opportunities, we need more context.
Fernandes happens to dominate both conversations.
His 136 chances, 33 big chances and 12.3 Opta xA show that the creative process behind the record was extraordinary in its own right. Opta’s underlying numbers make that clear.
Then came the second achievement.
Those chances became goals often enough for 12.3 xA to turn into 21 assists.
Elsewhere, Enzo, Pino, Wharton and Rice demonstrate the other side of the equation. A creator can produce dangerous opportunities without receiving the same reward in the assists column.
The same principle applies when evaluating broader player output: FIVDA’s James Garner analysis showed why a season-level headline can become more informative once the underlying role is separated out. Enzo goes one step further by showing how important creative actions can even occur before the pass that produces the shot.
That leaves us with a more useful way to read an assists table.
An assist tells us who supplied the final pass before a goal.
Expected assists help tell us what those final passes were creating.
Neither answers every question.
But in Bruno Fernandes’ record-breaking season, putting them together reveals something the number 21 cannot show on its own.
He did not just create at a record-breaking level.
He combined one of the league’s great creative seasons with an extraordinary finishing return from the chances he supplied.
That is why the record is more interesting once we look beyond the record itself.
Stay tuned for future posts and please send us a message if there is a specific topic you would like to see covered.
As always, thanks for taking the time to read these posts!
JC
Sources for the data presented are referenced within the article.
The main image is edited but originally from TNT Sports Football.