Premier League goals trail expected goals as early-season draws mount

The Premier League’s 2026-27 matches have produced 2.70 non-penalty goals per game against 2.88 non-penalty expected goals, according to a Guardian analysis published on 9 October. The gap matters because the share of games ending in draws is higher than in any previous Premier League season so far, although the figures cannot show that a particular missed chance would have changed a result.

Scoring itself has risen: the 2.70-goal average exceeds last season’s and two of the previous four campaigns, the analysis says. The expected-goals average is the second highest in the past six seasons. Taken together, those measures describe a league creating chances at a strong rate while converting fewer of them than the statistical estimate suggests.

How large is the Premier League scoring gap?

The Guardian estimates that finishing at the expected-goals rate would have meant almost two additional goals per week across the competition. That is an aggregate comparison, not a reconstruction of match outcomes. Expected goals assign estimated scoring values to chances; a difference between actual and expected totals does not identify a single cause or prove that a draw should have been a win.

The shortfall extends across clubs rather than resting on one side’s results. Fourteen of the 20 teams have scored fewer goals than expected, compared with 10 last season, according to the Guardian. Across the first five matchweeks, there were 24 instances of a team scoring at least one goal fewer than its expected-goals figure. That was twice the number recorded in the first 50 games of the previous season.

Which clubs and players illustrate the figures?

Manchester United show why shot count and chance value need to be read together. The Guardian says United lead the league for non-penalty shots but are joint second lowest for average expected-goal value per shot. An earlier, separate Opta Analyst review found that United had made a league-leading 182 touches in the opposition box across their first five games. They had scored eight goals, conceded eight and stood 12th at that point. Those Opta figures describe that five-match snapshot, rather than the table on 9 October.

The earlier Opta review offers a contrasting example at Manchester City. Across their opening five matches, City averaged 0.15 expected goals per non-penalty shot, the league’s highest chance value per shot at that stage, and converted 19.1% of their non-penalty shots. These measures concern the quality and conversion of chances in that early sample; they do not establish how either club will finish the season.

The Guardian finds that the leading expected-goal contributor at almost every club is taking a larger share of the team’s chances than the equivalent player did last season. Across the 20 clubs, those leading players have scored 6.8 fewer non-penalty goals than expected, nearly 40% of the division-wide shortfall. The analysis cautions that those shares may change as playing time and opportunities spread through squads.

At Everton, Thierno Barry has accounted for 49.4% of the club’s non-penalty expected goals, up from 17.7% last season, the Guardian reports. He has scored twice from chances valued at 3.5 expected goals. Liverpool’s Alexander Isak also has a larger share of his team’s expected goals than last season, although injuries limited his appearances then and he has played relatively few minutes for Liverpool. At Tottenham, the Guardian identifies centre-back Jan Paul van Hecke as the club’s leading player for non-penalty expected goals so far.

What can the early figures tell us?

The figures establish a scoring shortfall against expected goals at this point in the season, alongside an unusually high proportion of draws. They do not establish why finishing has fallen below the estimate. The Guardian notes that average chance quality and total box touches have not shifted enough to offer a clear explanation, while differences between individual teams are sharper than the league-wide box-touch figure suggests.