A red card in the fifth minute and a red card in the 90+13th minute occupy exactly the same space in a disciplinary table.
They do not create the same football problem.
The Premier League recorded 44 red cards involving 40 players in 2025/26, with Chelsea receiving eight and Everton and Tottenham four each, according to My Football Facts. But counting dismissals tells us little about how long teams actually played short-handed, what the score was when the card arrived or what happened afterwards.
FIVDA reconstructed all 44 dismissals from the season’s OpenFootball match timelines. Twenty-five were followed by no change in the dismissed team’s provisional points position. Fifteen were followed by deterioration. Four were followed by improvement.
The temptation is to turn those outcomes into a simple estimate of what red cards “cost”.
The evidence does not allow that.
What it does show is that red cards create radically different levels of exposure depending on when they happen and what is already happening in the match.
Forty-four red cards — but forty-four very different events
Consider two Chelsea dismissals.
Robert Sánchez was sent off after five minutes against Manchester United with the match goalless. The Premier League match report records that United were subsequently reduced to ten themselves before half-time, but Sánchez’s dismissal initially left Chelsea facing almost an entire match with a numerical disadvantage.
Malo Gusto’s red against Nottingham Forest arrived in the 87th minute with Chelsea already 3-0 ahead. Nothing subsequently changed in the score.
Both count as one red card.
So does Dominik Szoboszlai’s dismissal for Liverpool at 90+13 while his team were already losing to Manchester City.
That is why a red-card leaderboard is primarily a disciplinary statistic. It does not measure the amount of football a team actually played at a numerical disadvantage, or the match situation placed at risk by the dismissal.
FIVDA’s reconstruction assigns each card the score state immediately before the dismissal, the final result and the goals scored afterwards. For regulation-time dismissals, it also records nominal exposure as 90 minus the dismissal minute, while stoppage-time cards are assigned zero remaining regulation minutes.
Across the season, that produces 1,434 nominal short-handed regulation minutes, an average of approximately 32.6 per dismissal.
Figure 1 — A Sample of 4 Premier League Dismissal in 2025/26: The FIVDA Reconstruction
That measure is deliberately limited. It does not pretend to reconstruct the exact amount of stoppage time played short-handed, and matches in which the opposition were later reduced to ten require additional qualification.
It does, however, expose how much information a simple card count leaves out.
Most dismissals did not change the points category
The central finding is initially surprising.
Of the 44 dismissals, 25 — 56.8% — were followed by no change in the team’s provisional points position. Fifteen, or 34.1%, were followed by deterioration, while four teams improved their position after going down to ten.
Figure 2 — What Happened to Teams’ Match Position After a Red Card?
That does not mean 25 red cards had no effect.
Seventeen dismissals occurred when the team was already losing. Every one of those teams ultimately lost. Their points position therefore could not deteriorate: they had zero provisional points at the moment of the card and finished with zero.
A team losing 1-0 when reduced to ten and eventually losing 3-0 registers no points-position change, even though the match itself has clearly deteriorated.
The goal totals reinforce that distinction. Across the reconstructed events, dismissed teams scored 14 subsequent goals and conceded 37, a post-dismissal goal difference of -23.
Points movement therefore captures something useful, but it cannot be interpreted without score state. That same distinction between an eventual result and the path taken to reach it also appears in FIVDA’s analysis of how Premier League teams responded after falling behind.
Late cards had little time to change anything
Timing creates a second problem for conventional red-card totals.
Eleven of the season’s 44 dismissals occurred from the 81st minute onwards. Only one was followed by deterioration in the dismissed team’s points position.
That exception was Burnley’s Lesley Ugochukwu, sent off in the 84th minute of a goalless match against Liverpool. Liverpool subsequently scored in the 90+5th minute to win.
The other ten late dismissals produced no change in points category. Gusto’s 87th-minute red came with Chelsea already three goals ahead. Lewis Cook was dismissed at 90+6 with Bournemouth already losing 3-2. Sunderland’s Luke O’Nien went at 90+5 with his team 3-0 down, while Szoboszlai’s red arrived at 90+13 with Liverpool already behind.
Figure 3 — The Later the Red Card, the Less Time It Had to Change the Match
The contrast with earlier dismissals is striking. Of five cards shown in the opening 30 minutes, four were followed by deterioration and one by improvement. Between minutes 31 and 60, six of 19 were followed by deterioration, ten by no points-position change and three by improvement.
Those groups are too small to claim that an early red card is a fixed amount more damaging than a late one. Opponent quality, score state and the circumstances of individual matches all matter.
The narrower conclusion is much stronger: very late dismissals contributed substantially to the league’s red-card count while creating relatively little opportunity for the result to change afterwards.
Being ahead when the red arrived put the most points at risk
Score state changes the interpretation again.
Nine teams were winning when a player was dismissed. Only two preserved the victory. Four eventually drew and three lost.
Eighteen were drawing: four won, six drew and eight lost.
Seventeen were already losing, and all 17 lost.
Figure 4 — Score State at the Red Card Changed What Was at Risk
Chelsea’s home match against Brighton provides one of the clearest examples of a favourable position disappearing. Chelsea were leading 1-0 when Trevoh Chalobah was dismissed in the 53rd minute, before Brighton scored three times to win 3-1, as the Premier League match report records.
Tottenham experienced a similarly severe reversal against Crystal Palace. Spurs led 1-0 when Micky van de Ven was dismissed after 38 minutes and ultimately lost 3-1. Crystal Palace also went from leading Manchester United 1-0 at Maxence Lacroix’s dismissal to losing 2-1.
These are reasonably described as costly dismissals in an observational sense: the teams had three provisional points when the cards arrived and finished with none.
What we cannot say is that the cards cost them those three points. That would require knowing how the matches would have finished had they remained 11 against 11.
The dataset cannot provide that counterfactual.
Four teams actually improved after a dismissal
The opposite outcomes make the same point.
Four teams were drawing when they had a player dismissed and subsequently won.
Everton produced perhaps the season’s most striking example. Idrissa Gueye was sent off after 13 minutes at Manchester United with the match goalless. Everton subsequently scored through Kiernan Dewsbury-Hall and won 1-0.
Brentford also won 1-0 at Aston Villa after Kevin Schade’s 42nd-minute dismissal, with Dango Ouattara scoring in first-half stoppage time. Newcastle beat Manchester United 2-1 after Jacob Ramsey was dismissed in first-half stoppage time while the match was goalless.
Bournemouth’s victory at Fulham needs more care. Ryan Christie was dismissed after 41 minutes at 0-0, but Fulham’s Joachim Andersen was then sent off in first-half stoppage time. Bournemouth eventually won 1-0, but the game had ceased to be a conventional 10-v-11 contest.
None of these results proves that the dismissed teams benefited from losing a player.
They demonstrate something more useful: numerical inferiority changes the tactical problem without determining the outcome.
A ten-man side can lower its pressing height, defend more narrowly, sacrifice an attacker, preserve a counter-attacking outlet or concede possession while protecting central areas. The opponent still has to convert its spare player into territorial control and better chances.
Everton and Brentford show that attacking ambition does not necessarily disappear with ten men. Chelsea-Brighton and Tottenham-Palace show how quickly a favourable score can disappear when the opponent successfully exploits the numerical advantage.
There is no single ten-man strategy hidden in these results.
Why the season’s “minus 17 points” needs a warning label
There is one number that appears to summarise the entire season.
If provisional points at the moment of dismissal are assigned as three for winning, one for drawing and zero for losing, then compared with the eventual result, the 44 dismissals produced an aggregate swing of -17 provisional points.
The OpenFootball reconstruction breaks that distribution down further: three events finished three points worse, four finished two worse, eight finished one worse, 25 produced no movement and four finished two points better.
But saying “red cards cost Premier League clubs 17 points” would turn a descriptive statistic into a causal claim.
The accurate conclusion is that teams finished an aggregate 17 provisional points worse off than their position at the moment of dismissal.
Some of those teams might have conceded anyway. Some might have recovered without the card. Everton might have beaten Manchester United with 11 players. Chelsea might still have failed to beat Brighton.
There is no missing 11-v-11 version of those matches available for comparison.
That limitation is not a reason to ignore the statistic. It is the reason to describe it correctly.
A red-card table measures discipline, not impact
Chelsea finished with a club-record eight Premier League dismissals, according to BBC Sport. That tells us something meaningful about the frequency with which Chelsea received red cards.
It does not tell us how damaging each one was. That matters in a season when FIVDA’s wider results analysis also found Premier League victories becoming unusually narrow.
The season contained cards in the opening minutes and deep into stoppage time; dismissals while teams were winning, drawing and losing; matches spent for long periods at 10 against 11 and others in which the opposition subsequently lost a player too.
Treating those events as equivalent strips away the information that determines how much of the match was actually exposed.
The 2025/26 evidence points towards a better way to read red cards. Start with the dismissal count, but then ask when it happened, what the score was, how long the numerical disadvantage lasted and what changed afterwards.
Twenty-five of the 44 dismissals were followed by no movement in provisional points position. That does not make them harmless. Fifteen were followed by deterioration, but that does not prove the cards caused every lost point. Four were followed by improvement, but that does not mean playing with ten was advantageous.
The disciplinary table counts the cards. It is another example of why a Premier League ranking only becomes useful once the context behind the number is understood.
The match context tells us what those cards actually put at risk.
Main image is by BOOM Photography on Pexels.
