The big boys start their seasons this weekend. The transfer window closes today, and although it has been a relatively quiet summer in England, the anticipation of fans all around the country is hyped up to boiling point. Fans of the traditional top 6 (or is it a top 2 now…?) will dream of Premier League glory and stunning nights of Champions League action in the spring. Other Premier League fans will simply whisper to themselves: “Maybe… Leicester did it”.

The University of Reading’s forecasting mega machine, aka RED, has been rebooted and updated for the new season. This year, it is producing more detailed predictions of professional football matches. Based on its Bi-P-Var-X model (Bivariate-Poisson-Vector Autoregressive-X[cov’s], for those who care), for each match, expected goals will now be published (the forecast mean number of goals), as well as probabilities (expressed as %s) of each result outcome, the most likely scoreline pick, as well as the most likely scoreline conditional on the most likely result prediction happening, and the probability forecasts for several other possible common scorelines.

This Premier League season kicks off with Liverpool taking on newly promoted Norwich. Liverpool came so close last season to their well-funded rivals from Manchester. It is hard to look past those two sides for the title this season. Norwich visit Anfield first up, and RED suggests Liverpool have a 71% chance of victory, with the most likely scoreline outcome being 2-0 (12.5% chance). Norwich have just 11% chance of starting their season with a win.

Chelsea may have a transfer ban, but they poached a new manager over the summer in the form of club legend Frank Lampard. He faces a historically tough trip to Old Trafford on Sunday. However, Lampard himself has a 100% record there as a manager, having been in charge of Derby County’s thrashing of the Red Devils in the Football League Cup last September (2-2 and 8-7 on penalties, but it felt like a thrashing at the time). RED says that these two teams will start the season closely matched, with a slight advantage to the visitors Chelsea (38% chance of a win), though still the most likely outcome is 1-1 (12.5% chance).

The table below gives the rest of the Model’s forecasts for Round 1 of the Premier League.

Premier League Forecasts R1 2019/20
  • Expected Goals: the forecast average number of goals the Model expects for Home and Away teams
  • Outcome Probs: the predicted % chance of either a Home or Away win, with 100 minus these two numbers being the % chance of a draw.
  • Score Picks: the Most likely forecast scoreline outcome, as well as the most likely conditional on the most likely result outcome happening.
  • Home wins / Draws / Away wins: the predicted % chance of various potential scoreline outcomes of the match.