The model makes Manchester City the title favourite, first in 49.1% of 20,000 simulated seasons, ahead of Arsenal on 46.4% and Brighton on 1.5%. On expected points that is Manchester City 80.7, Arsenal 80.5, Brighton 63.2. Two clubs finish in the top four more often than not: Manchester City 97.8%, Arsenal 98.0%.

At the other end, Hull City carries the highest relegation risk at 62.1%, then Coventry on 53.2% and Tottenham on 44.3%. The bottom club averages 34.7 points to Manchester City's 80.7.

This simulation gives you a spread of outcomes, not a single, definitive answer. It's useful for seeing just how wide a club's range of possible finishes is. But remember, it's blind to the chaos a season actually throws at you: injuries, fixture pile-ups, managerial changes, VAR drama. It doesn't know who's crocked or who's suspended next week. What it does show is how the picture's shifted since August, which clubs the model's changed its mind about the most. Use it to gauge the width of the playing field, not to pinpoint the goal. The data's there, clear as day. What you do with it is up to you.

What has changed since August

The pre-season projection was run before a ball was kicked. 50 matches later, these are the clubs the model has most changed its mind about. Movement is in expected finishing position, which is measured against the rest of the division, so it reflects a changed view rather than simply games being played.

Rising

Falling

Compare with the full pre-season table, exactly as the model had it in August.

Every club, every finishing position

Club1234567891011121314151617181920xPts
Manchester City49.0737.828.032.841.110.630.290.080.080.030.010.010.010.0180.7
Arsenal46.4340.368.402.781.120.470.250.130.020.030.0180.5
Brighton1.526.4421.9317.3913.8910.078.356.314.473.312.371.680.970.610.400.200.060.010.010.0163.2
Brentford0.874.4414.3215.5013.7811.409.667.936.434.843.702.651.801.250.730.320.260.090.030.0161.0
Leeds0.773.1612.3714.0413.2912.2510.408.496.815.564.183.032.331.340.900.530.290.170.060.0260.1
Liverpool0.753.4712.7813.4613.0812.1910.258.486.695.504.303.302.211.580.940.520.290.160.060.0160.1
Manchester United0.211.506.829.399.6010.7410.199.508.838.076.705.534.383.142.261.340.920.540.270.0656.4
Chelsea0.080.974.756.698.098.549.509.799.398.577.616.675.734.303.312.261.751.100.600.3054.6
Everton0.130.523.285.307.078.399.709.9310.309.718.637.476.074.703.342.431.740.760.370.1453.8
Newcastle0.090.703.175.056.788.178.439.799.729.208.488.006.475.223.962.911.911.200.550.2253.4
Aston Villa0.040.402.313.555.286.577.698.499.259.369.178.487.726.495.413.732.801.841.020.4251.6
Bournemouth0.020.130.891.622.683.765.056.387.598.799.319.549.528.627.536.325.003.662.251.3748.1
Nottingham Forest0.010.090.511.061.902.723.775.176.347.809.079.359.889.949.117.726.124.713.071.6346.7
Sunderland0.010.010.290.691.001.952.673.654.806.367.658.559.6410.2210.129.287.876.725.243.2844.6
Crystal Palace0.010.060.230.500.731.392.072.893.705.036.577.859.2610.8711.4811.2010.398.916.8441.5
Fulham0.010.050.160.340.591.031.312.153.154.375.837.419.0010.4011.4011.7011.8310.698.6140.2
Ipswich0.010.130.190.270.491.011.602.263.184.455.677.578.7511.0012.1213.2414.4613.5838.6
Tottenham0.010.010.060.120.370.480.781.291.712.914.115.396.798.4410.5412.6814.1414.8215.3237.8
Coventry0.010.040.120.150.310.500.941.322.162.954.075.607.469.5511.6414.3417.3921.4536.2
Hull City0.010.030.030.110.210.410.721.151.852.884.346.098.4711.6415.0820.2126.7634.7

Each cell is the percentage of 20,000 simulated seasons in which that club finished in that position, so every row adds up to 100%. A blank cell is under 0.005%, not impossible. The xPts column is the club's average points total across those seasons. Club names link to that club's full model profile.

View the same grid as a shaded heatmap, which is easier to scan than the table above.

How it's built

The table above is where the division actually stands after 50 matches, finished off 20,000 times over. Points already won are treated as certain, and only the 330 fixtures left to play are simulated, from each club's attacking and defending strength: a blend of shots-adjusted Elo, last season's goal rates and how the club has actually played this season, the last of those counting for more the longer the season runs. Scorelines are drawn from the same negative-binomial and Dixon-Coles model the live BetSignals signals use. It knows nothing about upcoming fixture difficulty, injuries or suspensions.

_Generated from the BetSignals season simulation, 20,000 runs._