We Forced 10 Draft Strategies Through Tens of Thousands of Simulated Leagues. One Kept Winning.

Draft strategy arguments never end because nobody can run the experiment. So we did. Our season simulator plays full leagues that draft, set lineups, work the waiver wire, suffer real injuries, and crown champions across five replayed NFL seasons (2021 to 2025, real box scores). We gave one seat a strategy, held everything else identical, same rankings, same opponents, same luck, and replayed it once per strategy. Tens of thousands of leagues later, one rule kept winning.

The contestants

Baseline is Best Value: every pick is the highest-value player your roster can still use (need-gated, at most one backup QB and one backup TE). Every challenger is that same drafter with one rule added. Robust RB: running backs with your first three picks. Robust WR: the same with receivers (a control). Zero RB: no back until round six. Hero RB: one back in rounds one to two, then none until six. Late QB: no quarterback before round eight. No Backup QB / No Backup TE: never draft a second starter, cover from the wire. Early Elite TE: force the best tight end in round two.

The test was kept honest: every seat drafts from the same board, opponents draft the market's order with real roster needs, and because the wire decides whether thin rosters live, we ran it across five waiver worlds (reverse-standings priority, rolling priority, diligent leagues, casual leagues, and you as the sharpest manager in the room). Every team could stream a replacement on a bye or injury, so the no-backup plans got a fair fight.

Robust RB won almost everywhere. The scoreboard:

Why it works: RB points are only sold early

The mechanism is in five years of real half-PPR scoring, and it is a straight number comparison. Early, the back outscores the receiver: RB6 16.4 points per game, WR6 14.2. Then the curves cross. Waiting from the 12th to the 30th player at the position costs 4.2 points per game at running back but only 2.8 at receiver, and by the 40s it has flipped: WR42 over 9 points a game, RB42 under 8. Late receivers are players; late backs are filler. Durability widens the gap: about 35 percent of top-24 backs fall out of the startable pool each year against 26 percent of receivers. The Draft Timing chart in the Lab shows this season's projected version of the same curves, a cliff at running back and a gentle slope at receiver.

Zero RB's defense is the wire, and the wire is not the problem: the hottest available back each week was startable about 62 percent of the time, nearly identical to receivers. But there is one of him, every team wants him, and he is patching a hole you chose to dig with picks spent on the worst assets on the board.

Two honest caveats

The best tight end of the era appeared in 60 to 80 percent of championship lineups, yet forcing one in round two lost. Both are true: owning a tier-of-one tight end wins, but the value board already knows when he is worth it, so take him when he is the value rather than reserving the slot. And skipping backups flipped sign across waiver worlds (slightly negative under rolling priority), so we will not sell it as a law. The one rule that never lost in any world: never delay your starting quarterback.

What we changed

The draft Recommend view now pushes the Robust RB start: in your first three rounds it surfaces the best back on the board, and headlines him when his value is close to the top. The rankings have not moved, deliberately. The tournament varied construction, not grades, so the lesson is about when to take positions: in the early rounds, when the values are close, take the back.