Which Teams Are Hardest to Predict?

Eight gameweeks of Predicts10 predictions, team by team.

Based on Gameweeks 1-8: every team has played eight times, with roughly 165 predictions on each. Last updated 27 October 2026.

How we measured it

For each team we counted every prediction made on a fixture they played in (about 165 per team, since each of the 80 fixtures has had around 21 predictions) and checked how many got the right result. We also counted how many of their eight games finished with a result the crowd's favourite didn't see coming. Across all teams, the average prediction scored 2.0 points.

The hardest to call

Hull City: 20.5% of predictions got the right result
Nottingham Forest: 23.3%
Leeds United: 24.7%
Newcastle United: 27.5%
Liverpool: 29.5%

In each of Forest's first six games, the outcome the crowd backed most was wrong: home defeats to Leeds and Coventry and a goalless draw with Spurs, a 2-2 at Liverpool, a 2-1 win at Villa, and a 3-1 defeat at Crystal Palace. The last two went the crowd's way, a 1-0 defeat to Arsenal and a 1-1 draw at Ipswich. Hull went against the crowd in five of their eight games, beating Manchester United 2-0 when every player had backed United and drawing 2-2 with Chelsea when 93% had backed Chelsea. Leeds went against the crowd in six of their eight, including a win at Forest, draws with Brighton and Palace, and a 4-1 win over Newcastle.

Together Forest and Hull's sixteen games produced eleven results the crowd's favourite missed. Liverpool are the most misread of all, with the crowd wrong in seven of their eight, including a 3-0 home defeat to Manchester City, when 45% had backed Liverpool, and a 2-1 win at Brentford, when 54% had backed Brentford.

The easiest to call

Arsenal: 77.9%. The only miss was a 3-0 defeat at Brighton, a result nobody saw coming.
Manchester City: 74.8%. Seven wins from eight, and the crowd's favourite was wrong only twice: the 3-0 at Liverpool and the 2-1 defeat at Aston Villa, when 79% had backed City.
Coventry City: 56.6% (two wins and six defeats)
AFC Bournemouth: 53.9%
Ipswich Town: 53.5%

The pattern is simple: teams whose results fit expectations score well, whether that's Arsenal winning or a struggling side losing.

Exact scores and points

The average exact-score rate across all predictions is 8.6%. Arsenal (23.9%), Bournemouth (17.0%) and Sunderland (15.8%) are the teams whose scorelines players have hit most often. Brighton are the lowest at 1.8%, three exact scores from 165 predictions, followed by Crystal Palace at 2.5% and Tottenham at 3.0%. Brighton's results include 4-0, 3-4, 5-0 twice and 3-0, the sort of heavy scorelines players rarely predict (see which scorelines players overpredict).

By average points per prediction, Arsenal lead on 4.65, then Manchester City (3.13) and Bournemouth (3.04). At the bottom are Hull (1.05), Tottenham (1.14) and Forest (1.28).

What to take from it

Teams the crowd has misread repeatedly may be worth a second look before you follow the majority, particularly Hull, Forest and Leeds.
A team that looks predictable might be mostly predictable because of its recent run, so check whether the crowd is still reading it correctly. City's defeat at Villa came when 79% backed them.
The Gameweek 9 preview covers all seven of the hardest teams to call so far: Hull v Ipswich, Brentford v Forest, Bournemouth v Leeds, Newcastle v Everton, Liverpool v Arsenal and Aston Villa v Fulham.