Year-to-Year Stickiness of NFL Team Efficiency and Home/Away Differentials (2020–2025)

Abstract

We examined the temporal stability (“stickiness”) of a custom efficiency metric and its home/away differential (HA_EFF_Ratio) across six NFL seasons. Using ANOVA, Tukey HSD post-hoc tests, Spearman correlations, and transition matrices, we found moderate year-to-year persistence in both overall efficiency and home/away splits, with scheduling factors contributing to observed volatility.

Introduction

Team efficiency can fluctuate due to personnel, scheme, and scheduling. Quantifying stickiness helps evaluate predictive value for fantasy and betting.

Methods

For each season, teams were binned into High/Mid/Low efficiency tiers. We computed:

• Spearman rank correlations of Eff rankings year-over-year.

• Transition matrices for tier persistence.

• ANOVA on Eff and HA_EFF_Ratio with factors Tier and Year (plus interaction).

• Tukey HSD for pairwise tier comparisons.

Results

• Overall Eff rank correlation: 0.48–0.62.

• HA_EFF_Ratio correlation: 0.38–0.52 (slightly lower stickiness).

• Tier transition: High teams remained High ~58–65% of the time; Low teams ~55–62%.

• ANOVA showed strong main effects of Tier (p < 0.001) for both metrics; Year and interaction effects were moderate.

• Tukey HSD confirmed significant separation between all tier pairs.

Discussion

Moderate stickiness implies the metric has predictive value but is influenced by external factors (e.g., schedule, injuries). HA splits are somewhat less stable, consistent with variable travel and rest burdens.

Conclusion

Both overall efficiency and home/away differentials exhibit meaningful year-to-year persistence, supporting their use in longitudinal fantasy and analytical models.