If you have error bars and showed a statistically significant difference. Only mean was used (so only one method to calculate average, that shows nothing about distribution), we could instead use median and even an ANOVA analysis for different categories.
For consistency, the distribution of individual data points for each category could be used (per season, or for the full set of games as separate data points). Wouldn't be possible for AA or B&Fs because the the former is just a voting system at the end of a full season, while the latter we would need every single game's B&F read out. It is accepted that midfielders with very consistent elite performance across a season do well in those categories. If you didn't win an AA selection or a B&F, it's fair to assume there was some variance between your best and worst games. E.g a large period of the season just "good", with a hot streak early or late.
I will once again emphasise that no system is perfect, but it's entertaining to see something more comprehensive mocked by a person (not you) who thought "career goal/disposal averages" was the most accurate and definitive analysis there was. 4 different key statistical metrics, and 4 different voting metrics, then a tallying and ranking system, is undoubtedly a stronger approach to validate 21st century midfielder consistency/longevity. For the thousandth time, I have never claimed it measured "dominance of peak" or "most brilliant footballer I've seen".
Last edited:
