Historical Pace
This table shows the target number of points a championship team would take from each match, based on historical data. To read, find the row for the position of the opponent, then note whether the game was home or away and go to that column. The value in that row and column is how many points a typical championship team takes from that match.
For example, the hardest match of the year for a championship team is away to the team that finishes 2nd; looking at the row for Position 2 and the column for Away, we see that championship teams typically take 1.30 points from that fixture. On the other hand, the easiest match is home to the team finishing last; looking at the row for position 18 and the column for Home, we see that championship teams typically take 3.00 points from that fixture.
Estimated Standings
This table shows the projected standings for the current year. This accounts for the unpredicatability of the table during the early part of the season by adding in results from the previous season to make up the difference. More concretely:
- If a team has played at least half of its schedule, then we just use the current season's results, and project that out to a full season.
- If a team has not yet played half of its schedule, then we double their current season results (to make sure we're weighting current season performance more than previous season performance), then then we scale down the previous season's results to make up the difference to a full season.
- If a team is newly promoted, then for the previous season's performance, we substitute the performance of the worst non-relegated team from the previous season instead.
This tends to be fairly conservative on teams that get out to a hot start; for example, even after 11 matchdays in 2024, it still projected Nottingham Forest (5th in the table at the time) to finish 11th. However, this matches pretty closely with the betting markets; Forest had only the 10th best odds to make the Champions League at the time. So that suggests this approximation is reasonable.