Team DAP cards Week 4 and 5 analysis

Week 4 and Week 5 decks are on disk. Here is the test against actual results.

Lean cuts are the locked ones: Pass under 40% rush share, Balanced 40–55%, Run over 55%. Untrusted Run leans (R/P at or above 2.4) are flagged but still binned as Run. A win is a team win; ties count as non-wins. Yard hits use the same tercile method as before (high/mid/low within the week). Pass yards use the starting QB’s yards as the proxy because full team net yards are not uniformly published.

Favorite and underdog win rate by lean bin

Bin

Favorites

Underdogs

Pass lean

9–14 (39%)

8–16 (50%)  

Balanced

3–4 (75%)

1–4 (25%)

Run lean

4–5 (80%)

0–2 (0%)

Run lean untrusted

3–4 (75%)

0–1 (0%)

Overall favorites won 16 of 30 decided games (53%). The sample is small, but the split is clear: Pass-lean favorites underperformed, and Pass-lean underdogs covered half their games. Run-lean sides, trusted or not, won when they were favored and lost when they were underdogs.

Yard forecast hits and misses

Pass-yard tercile hit rate was 11 of 30 sides in Week 4 and 9 of 30 in Week 5, about 33% overall, close to chance. Extreme misses (high predicted vs low actual, or the reverse) were 9 of 30 in Week 4 and 8 of 30 in Week 5.

Rush-yard tercile hit rate was lower, about 27% combined, with more extreme misses. The cards that printed extreme Run scripts (BAL 210 rush, CHI 260, MIN 237, CLE 278, NE 258, DEN 217) all finished in the low or mid actual rush tercile. The 180-pass floor was not applied on the cards; several of those Run-lean sides (BUF 69, CHI 88, MIN 53, NE 85, CLE 91, CIN 107) also finished well under their printed pass numbers.

Lean direction

Card lean matched the actual game lean (rush share of total yards) on 14 of 31 sides in Week 4 and 12 of 30 in Week 5, about 43% combined. 

Pass-lean cards were the most reliable; Run-lean cards flipped to Pass or Balanced more often than they held.