Hover over a player to see their stats and rank.
Player Tier Results
In this analysis, all statistics are weighted equally when comparing players. This is a simplification, as not all stats contribute equally to winning or player value. A more sophisticated approach would reweight the statistics based on historical outcomes (e.g., correlation with wins or Pro Bowl selections) or based on individual beliefs about what makes a player valuable. The equal-weight approach shown here serves as a baseline that can be refined with domain-specific weighting schemes. Additionally, the statistics chosen for each position may not be the optimal set of variables for player evaluation. I welcome any feedback on both the stat selection and weighting methodology.
Season:
Players
-
Optimal Tiers
-
Cut Imbalance
-
Preference Matrix
Entry (i, j) = number of stats where row player i outperforms column player j
Stats Used for Comparison
Each position uses different statistics that capture performance relevant to that role. Players are compared pairwise across these stats to build a preference matrix, then clustered into tiers using cut imbalance optimization.
Quarterbacks (10 stats)
Passing Yards
Passing TDs
Completion %
Yards/Attempt
TD %
INT %
Interceptions
Passer Rating
Sack Rate
20+ Yd Pass Rate
Wide Receivers (10 stats)
Receptions
Receiving Yards
Receiving TDs
Targets
Catch Rate
Yards/Reception
Yards After Catch
First Downs
20+ Yd Receptions
Yards/Target
Cornerbacks (5 stats)
Pass Defenses
Interceptions
Solo Tackles
Tackles for Loss
Forced Fumbles
Red items indicate "lower is better" statistics.
Qualification thresholds: QBs need 200+ pass attempts, WRs need 50+ targets, and CBs need 20+ solo tackles to be included in the analysis. This ensures we're comparing players with meaningful sample sizes.