Huddersfield and Newcastle are threatening to break one of the greatest records of the Premier League: the lowest ever season points total, currently held by Derby (11 points, 2007-08). Both sides have home fixtures to try and break their wins duck this weekend. But the Model predicts no joy, with forecasts of draws for Huddersfield against Fulham (1-1)  and Newcastle against Watford (0-0).
Who would bet against Manchester City this weekend? The Model wouldn’t, forecasting a 68% win chance for the Sky Blues against Southampton, and a scoreline of 2-0.
Our relatively new two-step forecasting model approach has thrown up a strange forecast for Arsenal against Liverpool. The first-step model (an ordered logit model for the result) predicts a 39% win chance for Arsenal. However, from the second-step model (a bivariate poisson model of the scoreline), the most likely scoreline consistent with Arsenal being the most likely winner is 1-0, and has an estimated probability of occurring of just 1%. In other words, the first-step model is in disagreement with the second-step about the teams’ relative abilities, with the latter model giving Arsenal almost no chance of victory against Liverpool.
The rest of our Premier League forecasts are in the the table below:

  • The Forecast scoreline, with the % chance of that happening
  • The % chance of a win by the Home team, P(H), or Away team, P(A) (with one hundred minus those two numbers giving the % chance of a draw
Forecast Win (%)
Score Pr (%) P(H) P(A)
Arsenal Liverpool 1-0 1 39 32
Bournemouth Manchester Utd 4-3 3 39 32
Cardiff Leicester 1-2 9 32 39
Chelsea Crystal Palace 2-0 11 61 16
Everton Brighton 2-1 9 46 26
Huddersfield Fulham 1-1 8 38 34
Manchester City Southampton 2-0 13 68 12
Newcastle Watford 0-0 26 36 35
West Ham Burnley 2-0 9 48 25
Wolves Spurs 0-1 19 30 41