League-Wide Roster Analysis

After auditing my own team, I expanded the analysis to all 10 teams in the league to identify the “Optimal Strategy” for our specific league settings.

I tested evaluation windows from 3 to 90 days and correlated roster moves with final team success.

< Back to Team PJR Audit

1. The Optimal Lookback Window

One of the hardest questions in fantasy is: “How much recent performance history matters?”

I calculated the predictive power (correlation with future performance) for lookback windows ranging from 3 to 90 days across the entire league.

Winner: 42 Days While a 90-day window technically had the highest correlation, a 42-day window (6 weeks) captures 92% of the predictive power but allows you to react nearly 7 weeks faster.

xychart-beta
    title "Predictive Power vs Lookback Window"
    x-axis "Days" [3, 6, 9, 12, 15, 18, 21, 24, 27, 30, 33, 36, 39, 42, 45, 48, 51, 54, 57, 60, 63, 66, 69, 72, 75, 78, 81, 84, 87, 90]
    y-axis "Correlation" 0.127 --> 0.365
    line [0.1334, 0.1768, 0.2041, 0.2279, 0.2472, 0.2597, 0.2688, 0.2757, 0.2831, 0.2895, 0.2978, 0.3051, 0.3122, 0.3192, 0.3227, 0.3249, 0.3272, 0.3254, 0.3235, 0.3233, 0.3241, 0.3251, 0.3258, 0.3284, 0.3328, 0.3385, 0.3445, 0.3467, 0.3456, 0.3472]

Key Takeaway: Don’t overreact to a bad week (3-15 days is noise). But if a player has been bad for 6 weeks, they are likely to stay bad.

2. Roster Management Styles

I analyzed the “Manager DNA” of every team by looking at two metrics:

  1. Churn Rate: How many adds per week?
  2. Patience: How long do they hold a player before dropping? (Avg Hold Days)

The “Churn” Correlation

There is a significant positive correlation (+0.39) between Churn and Success. Teams that made more moves generally accrued more value. In contrast, “Patience” was negatively correlated (-0.47) with success.

quadrantChart
    title "Roster Management Style"
    x-axis "Patience (Avg Hold Time)" --> "Stubborness"
    y-axis "Low Value" --> "High Value"
    quadrant-1 "Diamond Hands (High Value)"
    quadrant-2 "Churn & Burn (High Value)"
    quadrant-3 "Panic Dropper (Low Value)"
    quadrant-4 "Sleeping at Wheel (Low Value)"
    PJR: [0.36, 0.95]
    HILL: [0.06, 0.82]
    AFFO: [0.15, 0.68]
    DO: [0.66, 0.44]
    CHER: [0.07, 0.40]
    ELLI: [0.31, 0.33]
    BP: [0.05, 0.31]
    YBSD: [0.95, 0.29]
    GIBB: [0.90, 0.29]
    $$$: [0.68, 0.05]

Quadrant Analysis:

  • Top Left (High Value, High Churn): This is the sweet spot. Teams like HILL and AFFO lived here.
  • Top Right (High Value, High Patience): PJR (Me) is the outlier here. High value despite stubbornly holding players. This indicates a dominant draft class.
  • Bottom Right (Low Value, High Patience): The “Sleeping at the Wheel” quadrant. Teams like DO, YBSD, and GIBB held players for 100+ days and finished at the bottom.

3. Member Breakdown & Optimal Benchmarks

Based on the Top 3 Teams, the “Optimal” strategy for our league is:

  • Target Churn: 2.2 Adds per Week.
  • Target Hold Limit: ~60 Days for underperformers.
Team Total Value Adds/Week vs Optimal Avg Hold
PJR (Me) 1385.6 1.2 -1.0 (Too Passive) 82.0d
HILL 1271.4 3.1 +0.9 (Aggressive) 45.4d
AFFO 1143.9 2.3 +0.1 (Optimal) 56.7d
DO 926.0 0.5 -1.8 117.1d
CHER 889.2 2.8 +0.6 47.2d
ELLI 826.5 1.4 -0.8 75.6d
BP 809.5 3.3 +1.0 44.6d
YBSD 788.8 0.1 -2.1 152.2d
GIBB 788.4 0.3 -1.9 146.4d
$$$ 574.3 0.5 -1.8 120.1d

Home

Last Updated: 2026-02-16

This site uses Just the Docs, a documentation theme for Jekyll.