Premier League goalkeeper workload: why more saves do not mean better defending

Metric and coverage

What the workload figure measures

The central measure is saves per match: a team’s total goalkeeper saves divided by its number of covered matches. It is presented to two decimal places. This is a team-level rate, not a ranking of the individual goalkeepers who played in those fixtures.

The comparison also reports goals conceded in the same covered matches. That matching matters: the goals figure belongs to the same set of fixtures used for the save total and save rate. Final league rank and points are then included as context rather than as components of the workload calculation.

Why the denominator matters

The teams do not all have the same amount of coverage. The table runs from 34 covered matches for Burnley, Liverpool and Wolves to 37 for Chelsea, Tottenham and Manchester United. A team with more covered matches has more opportunities to accumulate saves, so total saves alone would favour teams with larger denominators. Saves per match reduces that problem by putting the teams on a common match-based scale.

That adjustment does not make the measure a complete assessment of defending. It tells us how often saves were recorded on average in the covered fixtures. It does not, by itself, establish why the workload occurred, how difficult the saves were, or whether a higher workload represented superior goalkeeping.

Full 20-team table

The full comparison, ordered by saves per match, is shown below. The Premier League competition page provides the relevant league context. Every goals-conceded figure refers to the same covered matches shown for that team.

Workload versus goals conceded

High workload does not automatically mean a poor result

Burnley sits at the top of the workload ordering with 3.94 saves per match, followed by West Ham at 3.44. Their goals-conceded totals in the covered fixtures are also high at 67 and 60 respectively. That is a strong descriptive pairing, but it should not be turned into a claim that saves caused the goals or that every high-workload team defended badly.

The middle of the table shows why a simple ranking is inadequate. Sunderland recorded 3.11 saves per match and finished seventh with 54 points, while Aston Villa recorded 3.09 and finished fourth with 65 points. Their workload rates are close, but their final rank and points are different. Brentford had 2.94 saves per match, 46 goals conceded, ninth place and 53 points; Bournemouth had a slightly lower rate of 2.92, but 53 goals conceded, sixth place and 57 points.

There are also teams with the same displayed workload rate but very different outcomes. Liverpool and Wolves are both listed at 2.74 saves per match, and both covered 34 matches with 93 total saves. Liverpool conceded 49 goals, finished fifth and collected 60 points; Wolves conceded 65, finished twentieth and collected 20 points. The identical rate does not make their defensive or league outcomes identical.

Similar rates can conceal different coverage

Brighton and Chelsea are both listed at 2.78 saves per match, but Brighton’s coverage is 36 matches with 100 saves, while Chelsea’s is 37 matches with 103 saves. Brighton conceded 45 goals and finished eighth with 53 points; Chelsea conceded 51, finished tenth with 52 points. Since the displayed rate is rounded, and the denominators differ, the two rows should be read as similar reported workload rates rather than as proof of exactly the same underlying performance.

The reverse pattern is visible too. Aston Villa’s 3.09 saves per match is among the higher figures in the table, yet the team finished fourth with 65 points. Sunderland’s 3.11 was slightly higher, alongside seventh place and 54 points. These observations support a careful interpretation: saves per match describes how much goalkeeper intervention was recorded, while goals conceded and league results describe outcomes that cannot be explained by this one measure alone.

Burnley and Arsenal: a useful contrast

The largest workload gap

Burnley and Arsenal represent the clearest contrast in the table. Burnley recorded 134 saves across 34 covered matches, equivalent to 3.94 saves per match. Arsenal recorded 56 saves across 35 covered matches, or 1.60 per match. The difference in the reported rates is 2.34 saves per match, and Burnley made 78 more total saves despite having one fewer covered match.

The goals-conceded figures point in the same broad direction for this comparison. Burnley conceded 67 goals in its 34 covered fixtures, while Arsenal conceded 23 in its 35. Burnley finished 19th with 22 points; Arsenal finished first with 85 points. That is a gap of 44 goals conceded, 18 positions in the final ranking and 63 points.

What the contrast can—and cannot—show

As a descriptive comparison, Burnley versus Arsenal is consistent with the idea that a heavier goalkeeper workload can accompany a worse defensive and league outcome, while a lower workload can accompany a better one. It is useful because the two teams sit at opposite ends of both the saves-per-match ordering and the final table.

It is not proof that reducing saves caused Arsenal’s success, that Burnley’s goalkeeper was weaker, or that the same relationship must hold for every team. The metric is a record of team saves, not an individual goalkeeper assessment. The safest reading is narrower: in these covered fixtures, Arsenal required fewer saves per match and conceded fewer goals than Burnley, while also finishing higher and taking more points.

Rank and points context

The strongest final records are not ordered by workload

The final standings add useful context but do not turn saves per match into a league-quality measure. Arsenal finished first with 85 points at 1.60 saves per match. Manchester City finished second with 78 points at 2.31, and Manchester United finished third with 71 points at 2.38. Aston Villa was fourth with 65 points at 3.09, while Liverpool was fifth with 60 points at 2.74.

This top-five sequence is not monotonic. Aston Villa’s workload rate was higher than Liverpool’s, yet Villa finished one place higher and collected five more points. Manchester United’s 2.38 saves per match was lower than Liverpool’s 2.74, but United finished third with 71 points. The results show why a low or high workload cannot be treated as a standalone league ranking.

The lower-ranked teams also resist a simple rule

Burnley finished 19th with 22 points and West Ham finished 18th with 39 points, matching the two highest save rates in the table. Wolves finished 20th with 20 points, but its 2.74 saves per match was below the rates of 12 teams. Tottenham finished 17th with 41 points at 2.46, lower than the workload figures for several teams above it.

There are positive results among teams with relatively high workload as well. Sunderland’s 3.11 saves per match came with seventh place and 54 points. Bournemouth’s 2.92 came with sixth place and 57 points. These are not exceptions that overturn the broader comparison; they are reminders that rank and points provide outcome context, while saves per match captures only one part of the team’s defensive picture.

Limitations and interpretation

What can be compared safely

The safest comparison is between the reported saves-per-match rate and goals conceded over the same covered fixtures. This preserves the matching fixture window and avoids treating a larger raw save total as automatically meaningful. Burnley’s 134 saves in 34 matches and Arsenal’s 56 in 35 can be contrasted, but their raw totals should not be read without those denominators.

The table can also describe how the workload rates sit alongside final league rank and points. It cannot establish that one variable caused another. A descriptive association between more saves and more goals conceded in particular rows is not causal evidence, and the Burnley-Arsenal contrast is explicitly an illustration rather than a universal law.

What the metric does not establish

This is a team metric, not an individual goalkeeper ranking. It does not isolate the contribution of a particular goalkeeper, and it does not by itself measure the quality of the defending or the difficulty of each save. The provided coverage also varies from 34 to 37 matches, while the displayed rates are rounded to two decimals. Those limitations require the figures to be read as a structured comparison of workload and outcomes, not as a complete evaluation of goalkeeping.

Frequently asked questions

What does saves per match measure?

It measures total team goalkeeper saves divided by the number of covered matches. In this table, it is reported to two decimal places and is used to compare workload despite different coverage totals.

Why not compare total saves directly?

Coverage ranges from 34 to 37 matches. Teams with more covered matches have more opportunities to record saves, so the per-match rate is the appropriate primary comparison. Total saves remain useful only when their denominators are shown alongside them.

Does making more saves mean a goalkeeper is better?

No. The figures describe team goalkeeper workload, not an individual goalkeeper ranking. A higher save rate means more saves were recorded per covered match; it does not prove better goalkeeping or better defending.

What do Burnley and Arsenal show?

Burnley recorded 3.94 saves per match, 67 goals conceded, 19th place and 22 points. Arsenal recorded 1.60 saves per match, 23 goals conceded, first place and 85 points. They provide a clear descriptive contrast, but they do not prove a universal causal relationship.

Can saves per match explain league rank or points?

No. The table places the workload rate beside final rank and points, but it does not establish that saves caused either outcome. Aston Villa finished fourth with 3.09 saves per match, while Wolves finished twentieth with 2.74, showing why the rate cannot be used alone to explain the final standings.