Brentford 3-0 Chelsea: Fabio Carvalho's Stoppage-Time Goal and the Load-Management Equation

Minute 86 and a Substitution Board Nobody Watched

At the east corner of the Gtech Community Stadium, the fourth official's electronic board went up just past the 85th minute. The number appeared. On the bench, Fabio Carvalho stood, took off his training top, and pulled his socks a little higher above the knee — a habitual gesture of those who have been on an operating table for a cruciate ligament. Nobody in the stands applauded too loudly for an 86th-minute substitution. The game had long been decided; Brentford were ahead and controlling the tempo.

Then the ball reached Kevin Schade on the left flank. One touch, one cross. And at the far post, a player in a Brentford shirt ran into the box from the second line, met the ball with the inside of his right foot, and put it in the net in stoppage time. The score closed at 3-0 to Brentford over Chelsea.

That touch took less than half a second. But to get to it, Carvalho had to travel through 318 days without competitive football.

That is the whole event. The rest of this piece is about how that event is processed — by the coach, by the medical department, by the referee, and by us, the readers.

I watched that clip seven times. Not to find emotion. But to separate three distinct questions that coverage usually collapses into one: where did this player run into the far post from, why did he come on at 86 rather than 60, and through which checks was a stoppage-time goal confirmed.

Those three questions have three different answers, belonging to three different systems. And only one of them is about football proper.

The viewer sees the incident, the referee sees the moment, I see the whole process.

---

Context: a Friday night at the Gtech

Brentford came into this match third in the Premier League table on an unbeaten run stretching back to the opening of the season. For a club of modest financial scale relative to the Big Six, third place early in the season corresponds to a Champions League spot if the season ended right then. That sounds excellent in a bulletin. It is also a claim that needs checking against process data, which no bulletin supplied.

Chelsea arrived at the Gtech having gone three league games without a win, and in this match, for the first time under head coach Xabi Alonso, they failed to score. This is the detail repeated most across post-match headlines. I understand why. It is compact, it has a timestamp, it is easy to quote.

But there is a methodological problem here, and I want to state it plainly: both facts above — Brentford's third place and Chelsea's three-game winless run — sit in the small-sample zone, where conclusions about essence cannot be drawn. Three games. That is the entire evidence base for the "crisis" story on Chelsea's side. Three games is a period in which a player can be out with a tight hamstring.

I learned this lesson the hard way. In 2026, aged 26, I was a field reporter at the European World Cup qualifier between England and Slovakia at Wembley. In the first half I mispronounced defender Kyle Walker's name as "Wall-ker" three times in a row and was laughed at by colleagues. That night I received a reminder email from the editor-in-chief. What I did next was not pronunciation practice. I opened the footage and rewatched phase after phase to understand exactly how a referee determines an offside offence — just to reassure myself I was not the only person on the pitch making mistakes.

The lesson lay elsewhere: I realised I had drawn a conclusion from a single observation. One mispronounced name says nothing about my ability. Three winless games say nothing about Chelsea's system.

---

Friday as a statistical variable

Another data point appeared in almost every post-match bulletin, and it is a perfect example of the kind of data I want to warn about.

Brentford were credited with the best record in Premier League history for a specific matchday: an average of 2.71 points per Friday game, with a minimum threshold of three matches. Broader still, they had won six of their last seven Friday fixtures.

This is the kind of data I call "the merit of the denominator". Take it apart. A minimum threshold of three matches means any club that wins its first three Friday games can top a table nobody knew existed. The 2.71 figure does not measure Brentford's ability on Fridays. It measures how often Brentford were scheduled on Fridays, who they faced in those games, and how many they won.

The Premier League allocates Friday fixtures by broadcast calendar, not by feng shui. A smaller club appearing often on Fridays is usually because it is picked for a slot with less rights competition, or because its opponents come from a differently scheduled group. In Brentford's case, their frequent Friday appearances reflect their position in the broadcast distribution system, and — in part — the fact that they have just come through a favourable run of fixtures.

This is why I object to using this kind of data as an analytical crutch. If Brentford win next Saturday, it does not mean they have "lost their Friday form". If they lose a Friday game in February, it does not mean the record has been broken in any meaningful way.

I once spent four months during the 2026 lockdown rewatching all 380 matches of the 2026-20 season, focusing on Liverpool's tactical fouls under Jurgen Klopp. I counted an average of 10.2 fouls per match, mostly in midfield areas to stop counter-attacks. I sent the analysis to an editor and was rejected on the grounds that "nobody reads when there is no football".

But what I learned from those 380 matches was not the number 10.2. It was that every statistical pattern in football has a minimum sample threshold, and below that threshold you are measuring noise, not signal. Seven Friday games are below the threshold. Three winless games are below the threshold. Two goals after injury are below the threshold.

I am not saying this to diminish achievement. I am saying it to establish precisely how much evidence we have — because if we say Brentford are having an excellent season, we need to know what we are standing on.

---

Chelsea and the ball-conversion question

Chelsea's failure to score in this match, the first under Alonso, is usually read as an attacking problem. I want to break that concept apart, because "attacking problem" in modern football bundles at least four different things: chance creation, chance quality, finishing ability, and the ball-control structure that enables the other three.

Without xG data, without chance-creation counts, without a PPDA figure in the report, we cannot distinguish those four. But there is a reasonable structural inference from the scoreline itself: Brentford won 3-0, the third goal came in stoppage time, and Chelsea scored none. A side beaten 3-0 where the third concession arrives in stoppage time is typically a side that pushed players forward chasing an equaliser and was punished in the space behind.

What does that say about Chelsea? It says they controlled the ball but could not convert it into goals, then lost structure while chasing the game. This is the common pattern for possession-dominant sides early under a new head coach: the ball-control system is installed faster than the penetration system.

The core of the problem is not that Chelsea shot poorly, but the lag between the ball-control structure and the penetration structure — two systems installed at two different speeds, and that lag was exposed in stoppage time.

I have seen this pattern many times. At Euro 2026, I was assigned to analyse the penalty shootouts and wrote a 3,000-word piece citing data from the previous 14 shootouts. The editor cut it to 500 words because it was "too technical". But what I found in that process was not penalty technique. It was decision structure: an 11-metre distance, an average ball flight time of 0.3 seconds, and goalkeeper Gianluigi Donnarumma choosing the correct direction five times out of seven by studying opponents' habits.

Donnarumma was not a better shot-stopper. He prepared in advance. Chelsea did not lose this match through poor finishing. They lost because their system had no prepared answer for the situation of needing to score.

---

The mechanics of the goal: far post and the second-line runner

Carvalho's goal can be described in one sentence: Schade crossed from the wing, Carvalho ran to the far post and tapped it in. That sentence is true and useless.

Let's dissect the mechanism.

The far post — back post in English — is the zone opposite the origin point of the attacking ball. When the ball is crossed from the left, the far post is on the right side of the goal as viewed from the goal being attacked. This is the zone the opposing full-back typically vacates during transition phases, either because he must tuck in to cover the centre-back, or because he has been pulled wide by the crosser himself.

A second-line runner arriving at the far post is a standard attacking template, not an invention. It appears at every level of football from the Premier League down to amateur leagues. What is notable is not the template, but the timing of its activation.

In the second half, when a team is ahead and the opponent must push players forward, the opposing defensive structure typically loses coherence in ways it cannot recover from within a few minutes. In that window, aerial balls into the box from wide areas become more effective than normal, because there are fewer defenders in the box and the distances between them are larger.

That is why an attacking midfielder who comes on at 86 can score in stoppage time. He does not need to do anything extraordinary. He only needs to run into the space the opposing defensive system has already vacated, and to be there when the ball arrives.

This is the point where I want to pause and talk about how we habitually assign meaning to goals of this type. In coverage it is called a "moment of brilliance", an "inspiring return", a "fateful touch". In the language of the analysis room it is called a finish from a standard attacking template, activated by match state.

Every phase of play is a question about intent; data only gives us answers about consequence.

Both descriptions are correct within their own frame of reference. But only one of them can be verified against footage.

---

Fabio Carvalho: the profile of a Portuguese attacking midfielder

Carvalho is a Portuguese attacking midfielder. His natural position is the zone between the penalty area and the centre circle, where he can receive between defensive lines, turn, and create space.

But in this match he was not used in that position. He came on in the 86th minute and scored from the far post. These are two structurally different roles, and that difference matters more than any goal statistic.

In the attacking-midfielder role, a player is an enabler: he receives in midfield, decides the direction of the attack, and converts structure into a specific phase of play. In the second-line finisher role, he is a completer: he receives a structure someone else built and makes the final decision.

For a player returning from 318 days out with an anterior cruciate ligament injury, being used in the second role is a medically sensible decision. A finisher needs to sprint 20 metres at top speed once. An enabler needs to run 11 kilometres at varying speeds, change direction repeatedly, and absorb contact in crowded areas.

But that also means we have no evidence yet about Carvalho's return in his natural role. Two goals in two games is a positive signal about finishing ability. It is not evidence of chance-creation ability.

I emphasise this because it is the kind of confusion I encounter constantly in my work: conflating goals with a return. A player can score twice in two games and still not be fully recovered. A player can score none in five games and be fully recovered.

Goals are an output metric. A return is a process metric. They do not measure the same thing.

---

318 days and a medical protocol nobody gets to read

318 days. That is the number cited in the report, and it is the single most important piece of information in this entire event.

An ACL reconstruction in a professional player has a standard recovery window of six to nine months under ideal conditions. 318 days is roughly ten and a half months. That is the duration of a full reconstruction plus an extended rehabilitation phase, and in many cases it signals a complication or a re-injury during the recovery process.

This is the point where I have to speak plainly about a professional issue.

Professional football clubs do not publish full medical information about their players. This is not speculation; it is operational reality. They publish injury types at a general level, they publish recovery timelines as estimates, and they adjust both according to circumstance.

There are cases where publishing a long recovery is advantageous: it lowers expectations, reduces pressure on the player, and gives the medical department room to work. There are cases where publishing a short recovery is advantageous: it preserves transfer value, reassures shareholders, and keeps the player in the news.

No mechanism obliges clubs to publish a specific diagnosis. No governing body monitors the accuracy of published medical information. Players are bound by contract confidentiality in many cases.

The result is that we — the readers — form judgements about a player's career based on data that his own club has chosen to publish.

I am not saying clubs are lying. I am saying we are reading a summary drafted by a party with an interest in that summary.

With Carvalho, the 318-day figure is the only information we have. We do not know whether it was a first or second operation. We do not know how many MRI scans he underwent in that period. We do not know what his return-to-play protocol included, what load thresholds applied, and who held final sign-off.

These are the questions I always want to ask in press conferences and almost never get answered. Who is authorised to trigger the return-to-play process? Who holds veto? If the medical department and the head coach disagree, who wins?

Nobody answers those. And because nobody answers, we fall back on what is easier to count: two goals in two games.

---

Minute 86 as a load-management signal

There is one detail in this event that I regard as more important than the goal itself: the timing of the substitution.

Carvalho came on in the 86th minute. Not the 60th. Not the 70th. The 86th.

In modern football, when a coach wants to bring a player back from a long injury, a ladder is used. The player begins with individual sessions. Then group training. Then sitting on the bench and entering for 10-15 minutes when the game is settled. Then 30 minutes. Then a half. Then a start.

The 86th minute places Carvalho at the lowest rung of that ladder at which he could still score. This suggests the club's medical department is still managing his minutes in a controlled way.

There is an alternative reading: the coach simply wanted a fresh attacker for the closing phase when the opponent was tired. That is a sensible use of a technical player with the ability to run into space.

But the two readings are not mutually exclusive. And in the case of a player who has just come through 318 days out with a cruciate injury, the first reading always takes priority in analysis.

I have watched enough matches to know that coaches rarely disclose load-management protocols in detail. They talk about "feel", "confidence", "training performance". Those words are not measurable. But minutes are measurable, and minutes are the only thing we can track objectively.

If over the next three matches Carvalho still comes on at 80-plus, we have evidence of a protocol. If he starts the next match, we have evidence the protocol has been relaxed.

This is how you track a player after injury: not through goals, but through minutes.

---

Two goals in two games: the denominator problem

This number appears in every bulletin. Two goals in two matches since returning from 318 days out.

It is a good achievement. It is also a denominator from which no conclusion can be drawn.

Consider the structure of the sample. The first match was a cup tie, in which Carvalho scored against Reading — a lower-division side, in the English League Cup. The second was against Chelsea, in which he came on at 86 and scored in stoppage time.

These two goals were scored in two entirely different contexts, at two levels of resistance, at two different points in a match. There is no statistical basis for combining them into a single index with predictive meaning.

This is a problem I encounter constantly in football data analysis. Small denominators generate indices that sound impressive and carry no predictive value. Two goals in two games could become four in three, or two in eight. There is no way to distinguish those scenarios from the existing data.

What I mean is not that the achievement has no value. It has value — for the player, as a psychological signal that his body can still do what he needs it to do. For the club, as an indication that investment in the recovery process is paying off.

But it is not evidence that Carvalho is back. That evidence will come when he starts three consecutive matches and completes 90 minutes in each. That is the threshold any serious analysis of a player after a cruciate injury needs to reach.

---

The Reading goal and competition continuity

There is a small detail in Carvalho's file I want to separate out: the goal against Reading came in the English League Cup, not the Premier League.

This matters for one simple reason. The League Cup is the competition in which Premier League clubs typically rotate squads and give minutes to young or recovering players. The level of physical resistance, the pace of the game, and the quality of opposition are all lower than in a Premier League fixture.

A League Cup goal is a step on the recovery pathway. It is not a career milestone.

I say this not to diminish the goal. I say it to distinguish two structurally different types of goal. In analysis, we need to weight goals differently based on competition context, opposition quality, and match state.

A stoppage-time goal in a settled match is not the same as a 20th-minute goal in a balanced match. A goal against a lower-division side is not the same as a goal against a European-qualification rival.

This is the kind of distinction that raw metrics like "goals scored" erase. And it is why I always want to read the context before I read the number.

---

The referee's view: added time and late goals

Now to my own line of work.

Carvalho's goal came in stoppage time. In modern football, stoppage time is one of the least analysed and most misunderstood areas of the game.

The fourth official calculates added time from a list of stoppage events in the second half: substitutions, injuries, VAR checks, goal celebrations, extended interruptions. Each standard substitution adds roughly 30 seconds. Each VAR check adds the actual duration of the review.

This means a match with many second-half substitutions will have longer added time. And it means teams that make late substitutions are inadvertently creating additional time for the opponent to attack — unless they themselves exploit that window.

Brentford sent Carvalho on at 86, contributing to the added-time pool. And Carvalho scored inside that window.

This is a mechanism loop I have not seen analysed anywhere in coverage. Using a late substitute increases added time, and longer added time creates more opportunity for players who have just come on. In this case, the person who created the time pool was the person who used it.

Since 2026 I have spent a great deal of time tracking how VAR operates. In the France vs Australia match at the 2026 World Cup, Antoine Griezmann's goal was confirmed after a VAR check on Joshua Risdon's challenge. I sat for hours analysing camera angles, calculating foot speed and contact angle. A colleague told me I was spending too much time on a decision that took 90 seconds.

But what I was studying in that time was not the specific decision. It was the question of who has the right to request a review, at what intervention threshold, and who bears final responsibility.

Applied here: how was Carvalho's stoppage-time goal checked? Was there an automated offside check? If so, how long did it take and was it added to added time?

No information about this appears in the report. The goal was given, and that is the whole story told.

VAR does not fix mistakes; it only reassigns who carries the responsibility.

Since VAR was introduced, responsibility for the accuracy of a decision has been redistributed among the on-field referee, the VAR team, and the technology system. But the mechanism of that distribution is not published to the public. We see outcomes, not processes.

---

The laws are the second match

There is an aspect of this match I regard as more important than the goal, and it concerns how a referee manages a game already decided.

When the score is 2-0 or 3-0 and only a few minutes remain, the referee's role changes. He is no longer managing competition; he is managing closure. This means his intervention threshold shifts — minor contacts are often not penalised, cards are held back unless clear, and time is managed so the game ends safely.

This is a match-management skill that appears in no statistical table. A good referee is one who knows when to withdraw from the game and let it end.

But it is also precisely during that window that tactical fouls tend to increase. The trailing side tries to break the opponent's rhythm with small fouls in midfield. During my 2026 research on Liverpool, I found they averaged 10.2 fouls per match, mostly in midfield to stop counter-attacks. When the opponent is ahead and controlling tempo, that number typically rises, because the trailing side needs to break more phases to regain control.

In the Brentford vs Chelsea match, I have no foul data. But the structure of the match suggests the second half contained more tactical fouls than the first, and the stoppage-time goal was generated precisely from a situation in which Chelsea's defensive structure had dissolved after pushing players forward.

The laws never stand outside the match; they are the second match running in parallel.

That second match has no scoreboard. It has only decisions, and the decisions are not counted.

---

The data gap

I need to be clear about what I do not know in this analysis.

The original report contains certain facts: the 3-0 scoreline, the 86th-minute substitution, the stoppage-time goal, 318 days out with a cruciate injury, two goals in two games, Chelsea's three-game winless run, Chelsea's first blank under Alonso, Brentford's third place, and their unbeaten run.

Absent from the report: passes, possession share, shots, shots on target, xG, xGA, PPDA, fouls, cards, successful tackles, aerial duels won, successful dribbles.

This is a very large data gap. From the available information I can describe the match structure in general terms, but I cannot determine whether Brentford genuinely played better or simply exploited Chelsea's errors more efficiently.

I want to be explicit about this because there is a trend in modern football analysis to fill data gaps with inference and then present that inference as data. That is what I try not to do.

What I can say with high confidence: the goal came from a standard attacking template, in a favourable match state, executed by a player who came on at 86. That is a description of mechanism. It is not a claim about quality.

What I can say with medium confidence: Chelsea have a ball-to-goal conversion problem in the early phase under a new head coach, and that problem was exposed in stoppage time once they had pushed players forward.

What I cannot say: whether Brentford deserve third place.

---

The paradox: the redemption story versus the medical reality

Now to the part I want to spend the most time on.

The story told about Carvalho is a redemption story. A talented young player, a long cruciate injury, a patient recovery, and a stoppage-time goal as an exclamation mark. The coach praises his professional attitude during treatment. Media retell the journey. The crowd applauds.

It is a good story. It is also a story whose structure is designed to generate emotion, not to convey information.

I do not object to that story. I object to it being presented as a conclusion about the player's condition.

Consider two readings of the same event.

First reading: Carvalho is back. He has scored twice in two games. The injury is behind him. He will soon start and become an important part of Brentford's squad.

Second reading: Carvalho has completed the first phase of his recovery pathway. He has been used in matches of low resistance or in short windows at the end of settled games. He scored in structurally favourable situations. He has not yet passed the most important test: 90 minutes at continuous high intensity.

Both readings are compatible with the existing data. But only one is being told.

I understand why. The second reading has no protagonist. It has a process.

And this is the point I want to stress about the nature of analytical work.

When I write about refereeing controversies, I always begin by reconstructing the referee's sequence of actions on the pitch — where he stood, what he saw, how long he took to decide, whether he received support from the VAR team. I do this because the outcome of a decision does not tell you whether that decision was correct. Only the process tells you that.

Apply that principle here: a stoppage-time goal does not tell you the player has recovered. Only the minutes-management process tells you that.

And that process is not published.

Football has no VAR, only shadowed angles waiting to be exposed.

---

The Friday record as a media product

Back to 2.71 points per Friday game.

I believe this fact was included in the report because it is a type of data that spreads easily. It has three properties of a good viral statistic: it is short, it appears counter-intuitive, and it has an absolute comparison anchor — "best in Premier League history".

Those three properties are also three warning signs about analytical quality.

Being short means it cannot carry context. Being counter-intuitive means it draws attention away from ordinary explanatory factors. An absolute comparison anchor means it is presented as a historic peak when in fact it rests on a small sample.

With a minimum threshold of three matches, hundreds of combinations can produce a record. A three-match minimum means any club that wins three consecutive games on a given weekday can top a similar table.

I assume the editorial team behind the original report knew this. Including the fact was not an analytical error. It was an editorial decision to create an engaging hook for readers.

I do not object to that. I object to it being recycled as a claim about capability.

If three months from now somebody says "Brentford are good on Fridays", we need to remember that the basis of that claim is seven football matches allocated by television scheduling.

---

Chelsea's three-game "crisis"

On the other side, the same story runs in the opposite direction.

Chelsea have gone three league games without a win. That is the third-longest such run of the season. In a 38-round Premier League season, three games is 8% of the campaign.

Chelsea's first blank under Alonso is presented as a milestone. Technically it is a milestone — it is the first datum of its kind. But it is also a datum dependent on how many matches Alonso has taken charge of. If he has managed five, failing to score in one of five means something entirely different from failing to score in one of twenty.

The report gives no information on how many matches Alonso has managed. This is an important data gap, because the entire "first time" framing depends on it.

I want to add one more thing about this type of story. In football, pressure on a head coach is a phenomenon with its own structure. It is not determined by match results. It is determined by the gap between results and expectations.

At Chelsea, expectations are set by history, by financial scale, and by position in the stratification of English football. Three winless games at Chelsea generate more pressure than three winless games at Brentford because expectations at Chelsea are higher.

That means the pressure Alonso is under is not a measure of the quality of his work. It is a measure of the gap between results and a club's accumulated expectations over decades.

This is the kind of variable football analysis sometimes overlooks because it is not recorded in any statistical table. But it is part of the mechanism, and if we ignore it, we will misread what is happening.

---

The cultural message in the coach's praise

There is one detail in this story I regard as more important than the goal and more important than the winless run.

Brentford's head coach, Keith Andrews, publicly praised Carvalho's professionalism and work ethic throughout the 318-day treatment period. He spoke about seriousness in training, about how this player maintained professional standards during a period without matches.

This is not a casual remark in a press conference. It is a designed cultural message.

When a head coach publicly praises a player for his attitude during injury recovery, he is doing three things at once. He is rewarding a behaviour he wants repeated in the squad. He is building his own credibility with the group by showing that those who do the right thing are recognised. And he is sending a message to other injured players that the club will value their patience.

From a management perspective, this is one of the most effective forms of intervention a coach can make, because it requires no financial resources and affects a large number of people.

I have seen successful clubs build culture this way. And I have seen failing clubs treat long-term injured players as an accounting problem rather than a human one.

One thing I want to flag is verification context. In reports distributed through multiple editorial layers, coach names and club affiliations are sometimes not cross-checked against official sources before publication. With facts such as a club's current head coach, my working principle is always to cross-check with official club announcements or league organiser data before using it as the basis for any conclusion.

This is not suspicion. It is the second step in a process, and every process needs one.

---

A final paradox: emotion and law pulling in opposite directions

The public reaction to a match event and the rule system's assessment of that same event typically differ in structure, not just in conclusion.

The public assesses an event by its emotional meaning. A stoppage-time goal by a player returning from a long injury carries high emotional meaning, and that meaning is real — it exists in the experience of those watching and in the player's experience.

The rule system assesses the same event by its position within the match. A stoppage-time goal in a settled game is a valid goal if it was scored in a valid state. No other factor is considered.

These two systems do not conflict. They measure different things.

The problem arises when one system is used to answer the other's question. When we ask "what does this goal mean about the player's condition", we are using an emotional event to answer a biomedical question. When we ask "was this goal legal", we are using a rules event to answer a question about meaning.

Neither can answer the question posed. But both create the appearance of an answer.

I think this is why controversies in modern football are often irresolvable. They are not arguments about events. They are arguments about which system has the authority to assess the event.

The referee is the only person on the pitch not permitted to be led by emotion.

He has a single system to reference. Everyone else has several, and can choose the one that suits their feelings.

That is both the privilege and the burden of the job.

And that is why I write about it.

---

Takeaway: what to track

This event generates several trackable signals over the coming weeks.

First, Carvalho's minutes. If he continues to come on after the 80th minute over the next three matches, the load-management protocol remains in force. If he starts, the protocol has been relaxed. If he comes on around the 60-70th minute, the club is moving into the second phase of the pathway.

Second, Brentford's process data. When xG and PPDA figures are published for completed matches, we will have a basis for assessing whether third place reflects actual quality or a favourable run of results.

Third, Chelsea's expected goals. If they create high-quality chances over the next two or three matches without scoring, their problem is finishing. If they create nothing, the problem is attacking structure. These require different solutions.

Fourth, Brentford's fixture list. A run against the top six will be the real test of their current position.

Fifth, and perhaps most important, Portugal's squad announcements. If Carvalho is called up, that is external-system confirmation that he is back at a valued level. If he is not, that is a different signal — and a valuable one too.

---

Conclusion: a question left behind

The international break is coming. It is the season's natural stopping point, and the point at which stories get re-tested.

For Brentford, it is a window to reconcile third place with the reality of a club of modest financial scale. For Chelsea, it is a window to reset their attacking structure before the "crisis" narrative consolidates itself. For Carvalho, it is a window for the medical department to decide the next step in the pathway.

And for us as readers, it is a window to ask a question I consider more important than the match result.

When we celebrate a stoppage-time goal by a player returning from 318 days out, what are we celebrating?

We are celebrating a touch that took half a second. We are celebrating a process that took 318 days. And only one of those two things is visible.

I choose to write about the other one.

Not because it is more beautiful. But because it is the only one that can be checked.