What to track
Keep four live counts at every turn:
- Tier supply: how many players remain in each tier, not their absolute rank.
- Manager needs: which of the other nine teams still need QB, TE, RB2, and WR depth.
- Run pressure: whether a position is being drafted because a tier is truly ending or because managers are copying each other.
- Your replacement: compare a candidate to RB24, WR26, TE10, or QB10—not to zero.
The starting structure is 1 QB, 2 RB, 2 WR, 1 TE, 1 FLEX, K, and D/ST per team. Thus ten managers create fixed weekly demand for ten QBs and TEs, twenty RBs and WRs, plus ten FLEX spots; the current replacement model allocates that pressure to RB24 and WR26.
Turn discipline from 1.01
The sequence is 1, 20, 21, 40, 41, 60, 61, 80, 81, 100, 101, 120, 121, 140, 141, 160. There are eighteen opponent picks between 1 and 20. At each double turn, make two linked decisions:
- Take the player from the tier most likely to disappear before your next pick.
- Use the paired pick on the best remaining value, unless it is the same fragile tier and your roster already has that need covered.
Do not draft to repair a run that has already passed. If eight managers have QB and two do not, QB supply is tightening—but only buy it if the candidate is still meaningfully above QB10. The same logic applies to TE10 and to RB/WR tiers above their deeper replacement baselines.
The current data gate
The August 2 board, tier sheet, ADP, and projections are the current 10-team release. Any July 12-team player name, rank, or turn-specific availability claim is archival context, not an instruction. The draft date is late August but not fixed, so refresh the data, injuries, depth charts, and role news within 48 hours of the confirmed time.
Endgame
Ten teams leave more viable weekly options on waivers, so spend the bench on contingent role changes rather than safe substitutes. Keep a slot flexible for an early waiver claim. Take kicker and D/ST at the finish and stream them; for D/ST, exact sack/interception/fumble-recovery scoring is still undocumented, so use the internal model as an ordering only, never as a precise forecast.