Title Pace: 2026-27 against the 10 best Premier League seasons

Pick any club. We line up its first matches of 2026-27 with the 10 best Premier League seasons since 2002-03, through the same number of matches.

Every 2026-27 Premier League club is measured against the 10 highest-scoring seasons in our archive (2002-03 on), match for match: a club that has played nine games is set beside those ten seasons' first nine games. A pace score of 100 means level with the average of the ten. The 2026-27 data covers 89 of the season's 380 matches, from 21 August 2026 to 1 November 2026. This page reports results; it is not a forecast.

Title pace leaderboard

Every current club, ordered by pace score. Choose a club to see its matches beside each of the ten seasons and a points-by-match chart.

2026-27 after 89 of 380 matches (to 1 November 2026). Pace 100 = level with the average of the 10 best Premier League seasons since 2002-03 through the same number of matches.
#ClubPW-D-LGDPtsTop-10 avgPaceVerdict
1Arsenal98-0-1+92422.4107At or above the pace
2Manchester City97-0-2+102122.494Within striking distance
3Brighton & Hove Albion96-1-2+171922.485Off the pace
4Chelsea95-1-3+41622.471Well behind the pace
5Liverpool94-3-2+21522.467Well behind the pace
6Brentford93-5-1+71422.463Well behind the pace
7Newcastle United83-2-3−21119.955Well behind the pace
8AFC Bournemouth93-3-3+11222.454Well behind the pace
9Tottenham Hotspur93-3-301222.454Well behind the pace
10Aston Villa93-3-3−31222.454Well behind the pace
11Everton82-4-2−11019.950Well behind the pace
12Hull City92-5-2+11122.449Well behind the pace
13Leeds United92-5-2+11122.449Well behind the pace
14Ipswich Town93-2-4−41122.449Well behind the pace
15Fulham92-3-4−2922.440Well behind the pace
16Coventry City93-0-6−7922.440Well behind the pace
17Manchester United91-4-4−5722.431Well behind the pace
18Crystal Palace92-1-6−12722.431Well behind the pace
19Nottingham Forest91-3-5−6622.427Well behind the pace
20Sunderland91-2-6−10522.422Well behind the pace

The 10 best seasons since 2002-03

The benchmark is derived from the data by one rule: the 480 complete team-seasons from 2002-03 to 2025-26, ordered by final points, then goal difference, then goals scored. The next two below the cut: Arsenal 2003-04 (90 points); Manchester United 2008-09 (90 points).

The 10 best Premier League seasons since 2002-03 by final points (ties: goal difference, then goals scored), from the results in the archive.
#SeasonClubW-D-LGFGAGDPts
12017-18Manchester City32-4-210627+79100
22019-20Liverpool32-3-38533+5299
32018-19Manchester City32-2-49523+7298
42018-19Liverpool30-7-18922+6797
52004-05Chelsea29-8-17215+5795
62021-22Manchester City29-6-39926+7393
72016-17Chelsea30-3-58533+5293
82021-22Liverpool28-8-29426+6892
92023-24Manchester City28-7-39634+6291
102005-06Chelsea29-4-57222+5091

How it works

The comparison

  • Same number of matches. A club that has played N league matches of 2026-27 is compared with the first N league matches of each of the 10 best seasons. N is the club's own count, so a club with a game in hand is measured at its own N.
  • First N matches means the club's league matches of that season in kickoff order. The x-axis of the chart is this match number; it can differ from the official matchweek when fixtures were moved.
  • The 10 best seasons are chosen by a stated rule, not by hand: every complete team-season from 2002-03 to 2025-26 (24 seasons of 20 clubs = 480) ordered by final points, then goal difference, then goals scored, then the earlier season. The current season is incomplete, so it can never be on the list. A club's own earlier seasons can be on it.
  • The archive starts in 2002-03, so the list is the best since 2002-03, not of all time. Seasons before 2002-03 are not in our data and are not compared.

The numbers

  • Points are 3 for a win and 1 for a draw, counted from the results. Points deductions are not in the data and are not applied, so a final position here can differ from the official table where a club was docked points.
  • We store counts only (wins, draws, defeats, goals, shots) and compute every rate from the summed counts: points per match = points ÷ matches. Seasons are never averaged rate by rate.
  • The top-10 average is the mean of the ten seasons' totals through N; its points per match comes from those means.
  • Ranks (“3rd of 11”) place the club among itself and the ten seasons. Higher is better for points, wins, goals scored, goal difference and shots; lower is better for defeats, goals conceded and shots conceded. Tied values share a rank.
  • Shots and shots on target come from ESPN's box scores, which our archive holds for 2025-26 and 2026-27 only. All ten benchmark seasons are older than that, so their shot columns stay empty: we never fill them. For context the comparison also shows the club's own 2025-26 season through N, with shots, as a reference row that is not part of the average.

Pace score

  • Pace score = 100 × the club's points through N ÷ the top-10 average points through N, rounded half up. 100 means level with the average of the ten best seasons at the same point.
  • 100 and up: at or above the pace · 90 to 99: within striking distance · 75 to 89: off the pace · below 75: well behind the pace.
  • The leaderboard orders clubs by the unrounded pace score, then goal difference, then goals scored, then name. “Ahead of” counts the ten seasons the club has more points than through N.

Club pages and the luck rankings

  • A club page lists every season the club has in the archive: record, final points, final position and points against goals-based expected points. Positions come from the results-only table (points, then goal difference, then goals scored); first place in that table is marked as the champion.
  • Goals-based expected points treat a club's goals scored and conceded per game over a season as the means of two independent Poisson distributions. The win and draw chances of one such game give 3 × P(win) + P(draw) points, times the games played. Points minus that figure is the luck measure: how many more (or fewer) points a club took than its goals usually earn.
  • The luck rankings place all 480 complete team-seasons by points minus expected points, 1 = most points above what the goals predicted. It is the same model as the 10 most statistically improbable seasons, and the figures match that article. 2026-27 is shown separately on games played so far and is not ranked.

Data

  • Source: ESPN's public Premier League scoreboard and match summary feeds (soccer/eng.1), as recorded in each file. Results 2002-03 to 2025-26: data_layer/epl_goal_times_2002_2027.json, pulled 5 October 2026. Shots for 2025-26: data_layer/epl_2025_26_boxscores.json, pulled 7 September 2026. 2026-27 results and shots: data_layer/epl_2026_27_results.json, pulled 1 November 2026.
  • The archive loader gives 9,170 de-duplicated Premier League matches: every 380-match season from 2002-03 to 2025-26 plus the first 50 of 2026-27. ESPN listed 2009-10 twice; the duplicate rows are dropped. Its 50 early 2026-27 rows are replaced by the newer 2026-27 file, which agrees with all 50 and runs to 1 November 2026.
  • The figures are a snapshot. They move only when the data files are pulled again.

Data as of: 2026-27 to 1 November 2026 (89 matches).