Team Intelligence Β· 2026 preseason

Team Intelligence 2026: All 32 Offenses, Priced and Measured

Every offense in the league, scored for how many fantasy starters it can support and whether the market is paying the right price for it.

ADY
Auto Draft Years
17 August 2026 Β· 6 min read
32
offenses
320
historical team-seasons
2.6
league-average starters
0–6
observed range

The short version

  • Every team page answers one question: how many fantasy starters can this offense support, and is the market paying the right price?
  • Capacity is read off the 320 team-seasons in the dataset, not asserted.
  • The phrase β€œthe market hasn't priced it in” is gated behind a measurable threshold, so most teams never get it.
  • Player calls are made on position-rank disagreement after price drift is removed β€” and often there is no call to make.

How to read this series

Thirty-two offenses, one framework, and a rule that stops the framework from saying the same thing thirty-two times. Every number below is computed from the same dataset at build time.

The three rules this series runs on

One. β€œThe market hasn't priced this in” is only written when the team's roster, priced at market ranks and at model ranks, differs by at least 8 league places in the same direction as the roster's own value against the league. That bar is set from the observed spread, not picked round: the median team disagrees by 5 places, so a five-place bar would hand the claim to half the league. That leaves 6 underpriced (Buffalo, Kansas City, New Orleans, New York Jets, San Francisco and Washington) and 6 expensive (Baltimore, Detroit, Green Bay, Jacksonville, Philadelphia and Tampa Bay). The other 20 teams do not get the sentence.

Two. Fantasy starter capacity is measured, not asserted. Each team's forecast is matched to its nearest historical offenses across 320 team-seasons (2016–2025), and the capacity is what those offenses actually produced. League average is 2.6; the observed range is 0 to 6.

Three. A player call has to clear a threshold set from the spread of the model's own disagreements at his position, after the drift affecting everyone at his price is removed. On 20 of the 32 rosters, nothing clears it β€” and those pages say so instead of manufacturing a target.

Every offense, ranked by capacity

TeamStartersProj PPGRankRelative strengthPriceGap
Kansas City3.726.36thWR 2ndUnderpriced+8
Buffalo3.628.31stRB 2ndUnderpriced+14
Tampa Bay3.626.45thQB 1stExpensive-18
Dallas3.527.33rdQB 4thFair+2
Detroit3.527.72ndQB 2ndExpensive-14
Baltimore3.227.24thRB 1stExpensive-17
Los Angeles Rams3.124.711thRB 7thFair+2
Green Bay3.025.19thTE 6thExpensive-21
Philadelphia3.025.88thRB 3rdExpensive-10
Cincinnati2.923.513thQB 7thFair+3
Minnesota2.923.115thTE 14thFair-4
New Orleans2.924.512thRB 9thUnderpriced+21
Seattle2.925.010thRB 13thFair-6
San Francisco2.925.97thRB 4thUnderpriced+11
Arizona2.722.818thRB 11thFair+3
Atlanta2.723.314thWR 7thFair0
Indianapolis2.722.023rdRB 10thFair+4
Los Angeles Chargers2.722.221stQB 6thFair-6
Miami2.723.016thQB 18thFair+2
New England2.722.817thRB 12thFair-5
Houston2.622.419thWR 15thFair-3
Washington2.622.220thRB 16thUnderpriced+12
Chicago2.421.424thRB 17thFair0
Pittsburgh2.422.022ndWR 14thFair+5
Las Vegas2.321.325thWR 17thFair-1
Carolina2.220.329thRB 19thFair-1
Cleveland2.220.528thRB 18thFair+3
Denver2.120.927thQB 22ndFair+6
Jacksonville2.121.026thWR 20thExpensive-12
Tennessee1.720.130thRB 22ndFair+5
New York Giants1.518.731stQB 29thFair+5
New York Jets1.317.832ndQB 31stUnderpriced+12

Starters is the expected number of QB1/RB2/WR3/TE1 finishes from comparable historical offenses. Gap is the team's market rank minus its model rank, in league places; positive means the model likes it more than the price does.

All 32, by division

Was this useful?

Sources: Auto Draft Years FF Model workbook; Auto Draft Years 2026 consensus board; Player outcomes 2016–2025. Generated from the dataset at build time.