Premier League Red Cards Across Seasons: How to Read Event Totals

What the event series actually measures

The counting rule is deliberately narrow: the series includes only event-feed entries labelled “Card” and “Red Card”. Each qualifying entry is preserved as a raw event. Separately, a match enters the incidence count once if it contains at least one qualifying event. That distinction produces 462 raw event rows in 429 matches across the full 3,800-match coverage.

A sending-off label, not a severity scale

Under IFAB Law 12, a red card communicates a sending-off. Sending-off offences include serious foul play, violent conduct, denying an obvious goal-scoring opportunity and a second caution. The label therefore covers more than one kind of offence; a total of labelled events cannot be read as a scale of foul severity.

Two denominators answer different questions

The raw-event rate measures qualifying event rows relative to matches: 462 across 3,800 matches equals 12.16 raw events per 100 matches. The match-incidence measure asks how many matches had at least one such event: 429 of 3,800, or 11.29%. The first rate can exceed the second because a match can contain multiple counted events. Keeping both measures visible prevents an event count from being mistaken for the number of affected matches.

Season-by-season red-card results

All ten seasons contain 380 completed matches, so each seasonal rate uses the same match denominator. The table reports the raw event total alongside matches with an event, then gives the two corresponding per-100 measures. Rates are shown to two decimal places.

The high and low

The high is 2015: 59 raw events and 15.53 per 100 matches, with 55 matches containing an event (14.47%). The low is 2022: 30 events and 7.89 per 100, with 27 affected matches (7.11%). The high-to-low difference is 29 events, and the high/low ratio is 1.97.

The sequence is not a one-way movement

Totals fell from 59 in 2015 to 41 in 2016 and 39 in 2017. They then stood at 47, 45, 48 and 43 in 2018, 2019, 2020 and 2021, before reaching the ten-season low of 30 in 2022. The total rose to 58 in 2023 and was 52 in 2024. This sequence supports a description of year-to-year variation; it does not, by itself, explain why the figures moved.

How to compare seasons without overclaiming

The same competition and fixed season size make comparisons cleaner, but a clean comparison is not the same as a broad explanation.

Read raw events beside match incidence

In 2015, 59 events appeared in 55 matches, meaning 14.47% of matches had at least one counted event. In 2022, 30 events appeared in 27 matches, or 7.11% of matches. A total of 59 therefore does not mean 59 separate matches, just as 30 does not mean 30 separate matches. Use the raw-event column when the question is event-feed volume; use the match percentage when the question is how widely the event appeared across matches.

Keep description separate from explanation

Because every season has the same 380-match denominator, raw totals and per-100 rates identify the same high and low here. That makes the comparison straightforward. It does not identify the reason for any change. If this series moved alongside another measure, that would show correlation or co-occurrence, not proof that one factor caused the other.

Team volume is context, not a disciplinary ranking

The team view adds attribution, not adjustment. It shows raw event volume and the number of matches containing an event for each listed team in the covered feed rows. These are not adjusted rates and should not be treated as a disciplinary-quality table.

Arsenal’s 33 is the largest raw total among these five listed teams, but that is only a statement about supplied volume. Southampton and Brighton do not have the same ten-season match exposure as the competition-wide sample. The rows therefore must not be read as equal-exposure comparisons or as a fair disciplinary ranking.

Method and coverage

Coverage and counting convention

The competition is the Premier League in England. Coverage runs over ten completed seasons labelled 2015 through 2024 inclusive. Each season contributes 380 completed matches, for 3,800 completed matches overall. Only event-feed entries carrying the labels “Card” and “Red Card” are included. A match with one or more included entries is counted once in the match-incidence figures.

The overall event rate uses 462 raw events over 3,800 matches, while the overall match percentage uses 429 matches with at least one event over the same 3,800. Seasonal event rates and percentages use 380 as their denominator. This fixed structure lets the table show both volume and incidence without mixing denominators.

Documented reconciliation difference

An independent team-statistics cross-check totals 461 rather than 462. The difference comes from one 2019 fixture/team record with one raw event but a zero team-stat red-card value. It is reported as a reconciliation difference; the main Card/Red Card series remains 462.

Cancelled labels are not subtracted

Seven separate rows labelled “Red card cancelled” were observed in 2021–2024. They are not subtracted from the main Card/Red Card count. The series therefore follows its stated event-label convention rather than applying a separate cancellation adjustment.

Limitations and official context

What the totals cannot establish

A raw event total cannot establish referee consistency, foul severity, causation, team quality or suspension length. It is also not a universal disciplinary taxonomy: the event label is a feed convention. The safest claim is descriptive—for example, that one season recorded more qualifying event rows than another—not that its football was more severe or its officials less consistent.

This is the correlation-versus-causation boundary in practical terms. Even if the red-card series moved alongside another measure, the figures alone would show co-occurrence, not that one factor caused the other.

Rules and VAR context

The Premier League’s red-card decisions explanation states that awarded red cards are automatically checked by VAR. Its published Premier League article and further Premier League article also cover the possibility that red-card incidents with a yellow or no card can be checked, while the final decision remains with the referee. VAR review does not turn this event count into a severity score.

Frequently asked questions

What exactly is counted?

Only event-feed entries labelled “Card” and “Red Card” enter the main series. That produces 462 raw event rows across the covered seasons.

What denominators are used for the rates?

The overall denominator is 3,800 completed matches. The raw-event rate is 462 events per 3,800 matches, expressed as 12.16 per 100; the match percentage is 429 matches with at least one event out of 3,800, or 11.29%. Each season uses 380 matches.

Which season is the high, and which is the low?

2015 is the high with 59 events and 15.53 per 100 matches. 2022 is the low with 30 events and 7.89 per 100. The difference is 29 events and the high/low ratio is 1.97.

Why are there 462 events but only 429 matches?

The event total preserves each qualifying event row, while the match figure counts a match once when it contains at least one such event. A match can therefore contribute multiple raw events but only one match to the incidence measure.

Can these figures rank disciplinary quality?

No. The team figures are raw volumes rather than adjusted rates, and the season totals cannot establish foul severity, referee consistency, team quality, suspension length or causation. They are most useful for describing event volume within the stated coverage and counting convention.