Scope and denominator
The comparison covers the 15 completed Premier League seasons from 2011–12 to 2025–26. Each row contains 380 completed matches, so the denominator is stable across the series. The 2010–11 season was left out because the checked record contains 370 completed matches rather than the standard 380; no missing fixtures were treated as scoreless games.
Goals per match is a frequency: total home and away goals divided by completed matches. It is different from total goals, which rewards a season with more fixtures, and different from a team or player scoring rate. Because the match count is constant here, the ordering of total goals and goals per match is identical, but the rate is easier to compare with other competitions or seasons with different coverage.
The season-by-season record
The table shows a league that mostly stayed between 2.5 and 2.9 goals per match before the exceptional 2023–24 spike. The recent line then steps down: 2.934 in 2024–25 and 2.750 in 2025–26. The fall is substantial, but it does not return the competition to its quietest scoring environment.
| Season | Matches | Goals | Goals per match |
|---|---|---|---|
| 2011–12 | 380 | 1,066 | 2.805 |
| 2012–13 | 380 | 1,063 | 2.797 |
| 2013–14 | 380 | 1,052 | 2.768 |
| 2014–15 | 380 | 975 | 2.566 |
| 2015–16 | 380 | 1,026 | 2.700 |
| 2016–17 | 380 | 1,064 | 2.800 |
| 2017–18 | 380 | 1,018 | 2.679 |
| 2018–19 | 380 | 1,072 | 2.821 |
| 2019–20 | 380 | 1,034 | 2.721 |
| 2020–21 | 380 | 1,024 | 2.695 |
| 2021–22 | 380 | 1,071 | 2.818 |
| 2022–23 | 380 | 1,084 | 2.853 |
| 2023–24 | 380 | 1,246 | 3.279 |
| 2024–25 | 380 | 1,115 | 2.934 |
| 2025–26 | 380 | 1,045 | 2.750 |
The middle of the series is not flat, but it is orderly. Six seasons from 2011–12 through 2016–17 sit between 2.566 and 2.805. The next six, 2017–18 through 2022–23, range from 2.679 to 2.853. The difference between those bands is much smaller than the later jump to 3.279.
Reading the table from left to right gives three useful stretches. The opening run is not a boom: 2011–12, 2012–13 and 2013–14 sit at 2.805, 2.797 and 2.768. The 2014–15 trough interrupts that cluster, then 2015–16 and 2016–17 return to 2.700 and 2.800. That sequence is better described as a dip followed by recovery than as a permanent change in the scoring climate.
The middle stretch also resists a simple label. 2017–18 is lower at 2.679, 2018–19 rises to 2.821, 2019–20 eases to 2.721, and 2020–21 is 2.695. The values move, but they remain inside a comparatively narrow band. The next two seasons reach 2.818 and 2.853 before the 2023–24 leap. In other words, the record contains ordinary year-to-year movement before it contains an exceptional year.
Because all of these rows have the same denominator, the total-goal differences can be translated directly into match frequency. The 2023–24 peak produced 162 more goals than 2022–23, which is the same 0.426 increase per match multiplied across 380 fixtures. The table is therefore telling one story in two units: a season total and the average score environment of one fixture.
The low and the peak tell different stories
The low in 2014–15 was 975 goals, 91 fewer than the previous season and 51 fewer than 2015–16. At 2.566 per match, it is the only season in this window below 2.6. The following season returned to 2.700, so the low looks like a trough rather than a new baseline.
The peak in 2023–24 is more striking. Its 1,246 goals are 162 above 2022–23 and 131 above 2024–25. The 3.279 rate is 0.426 higher than the next-best season in the table, 2024–25 at 2.934. That gap is large enough to make the peak a distinct event in this sample, even though the data alone cannot say whether it came from tactics, finishing, officiating, team mix or other conditions.
The stable part of the series is the story; the 2023–24 peak is the exception that needs its own explanation.
What the latest seasons change
The latest row should be read in two directions. From the peak, 2.750 is a fall of 0.529 goals per match, or about 16.1%. From the 2014–15 low, it is an increase of 0.184, or about 7.2%. Calling 2025–26 a collapse would therefore overstate the evidence; calling it a continuation of the peak would ignore the clear reset.
The 2025–26 total of 1,045 goals is also close to the long-run centre. The 15-season unweighted mean is 2.799 goals per match, and the latest season is 0.049 below it. That positioning supports a restrained reading: the peak has receded, while the underlying scoring environment remains within the range seen repeatedly across the period.
That is why the season table matters more than a single headline. A record can be true and still be unrepresentative. The sequence shows whether the record was a one-season jump, a lasting regime or a return to normal levels.
It is also useful to compare the final three rows as a sequence rather than treating 2023–24 as a standalone record. The rate falls from 3.279 to 2.934, then to 2.750. Both steps are downward, but the final value still sits above 2014–15, 2017–18, 2019–20 and 2020–21. The most defensible description is a retreat from an extraordinary high toward a familiar part of the historical range.
The same logic applies to totals. A headline such as “more than a thousand goals” is true for every season in this selected window except the 2014–15 low, but it does not distinguish a 975-goal year from a 1,246-goal year. The rate adds the scale needed to tell those seasons apart without implying that a single season’s total contains an explanation.
For readers comparing competitions, the fixed Premier League schedule is an advantage. It means there is no need to normalise the table for different numbers of fixtures. For readers comparing clubs or players, however, this league-level rate is only context: it says how many goals the competition produced, not how a particular side contributed to the change.
How to read the rate
Use total goals when the question is how much scoring the competition produced. Use goals per match when the question is how frequently goals arrived in a typical fixture. The two measures agree on the ranking here because every completed season has 380 matches, but the distinction remains useful for the Championship and for seasons with incomplete coverage.
The rate also avoids a common overreach. It does not identify the best attack, the most entertaining team or the cause of a change. It is a descriptive league-level signal. A higher value can reflect more goals at both ends, more one-sided matches, or a different balance of results; separating those mechanisms requires additional analysis.
The strongest evergreen takeaway is therefore narrow: Premier League scoring averaged 2.799 per match over these 15 seasons, reached its low in 2014–15, peaked in 2023–24, and moved back toward the long-run centre in 2025–26.
A rate is most useful when the reader can see the observations behind it. The table keeps the denominator, the raw goal total and the normalised frequency together, so a peak is not detached from the number of matches that produced it. It also makes the difference between an isolated extreme and a repeated level visible. Several seasons sit close to one another, while 2014–15 and 2023–24 define the two ends of the window. That combination is why the article describes an evolution rather than a single linear direction.
The same caution applies to the latest row. A fall from the peak is clear, but the latest rate still sits above the low and within the broader band of completed seasons. Readers can therefore use 2025–26 as a current reference point without treating it as a forecast. If a future update adds another completed season, the calculation should retain the same total-goals divided by completed-matches definition and record any coverage difference beside the new row.
Limits and method
Data source: Xtra-Stats. The sample includes completed Premier League matches for seasons 2011–12 through 2025–26. The 2010–11 record was excluded from the main comparison because only 370 completed matches were present. Goals per match is total recorded goals divided by completed matches and is shown to three decimals.
The analysis is descriptive and does not assign causes to the scoring changes. It also does not adjust for team strength, rule changes, penalties, expected goals or the timing of goals. The latest completed season is 2025–26; the following season was not used because it was not complete at the data cut.