Most people use data from Fbref (Statsbomb) or Understat to assist their FPL decisions. However, an underlooked source of data is the FiveThirtyEight website.
FiveThirtyEight use data to analyse sports, politics and science. With regards to football, they are well-known for their predictions on every tournament imaginable, from the Champions league to League two. It’s the data behind these predictions that’s useful from an FPL perspective.
Their website has a full explanation on how these probabilities are calculated, but they are based on each team’s offensive and defensive strength. These are estimates of how many goals each team is expected to score or concede ‘against an average team’ i.e. the predictions are based on xG. This means we have another source of xG data!
The way these team strengths are calculated means it has advantages over xG data from Fbref ,particularly at the start of the season. As these ratings are calculated (and comparable) for teams in the Premier League and Championship, they make it easier to compare the quality of newly promoted teams to existing Premier League teams. Furthermore, team ratings are updated at the start of each season for new signings, which means they’re a more accurate reflection of a team’s quality than their previous season xG stats.
For each game, one can also download their predictions for the expected number of goals each team is expected to score and concede for all of their upcoming games (in the spi_matches_latest file). In terms of FPL, this is would be useful to help decide which team’s defenders are worth transferring in.
In conclusion, FiveThirtyEight’s team ratings are useful for FPL decision-making at the start of a new season as
- They allow you to assess the quality of newly-promoted teams
- They take into account new signings
Furthermore, they also provide predictions of how may goals a team will concede in each game throughout the season, which can be used to evaluate FPL defenders.
Statsfindings 