The 2023 Draft

• Picks
  180
  15 rounds · 12 seats
• First off the board
  Justin Jefferson
  JD
  1.01
• Keepers
  21
  prices paid before the night
• Picks that changed hands
  9
• Graded through
  Week 14
  Larry leads at +3.0 a pick
  of 14

The board

Rounds down, seats across in the order the room drafted. Each cell carries its position; the tint is the same fact, said twice. A keeper is marked in wax, and a pick standing in another seat’s column says who used it. The signed number is the pick’s surplus through week 14 — see the grade.

When the board came for each position

For each position, the pick at which the first, second, third of them left the board. A steep line is a run; a flat one is a position the room let sit.

Twelve shapes

What every seat walked out with, by position, in draft order. It reads as strategy: who went heavy at a position, who left one bare.

The keepers

21 slots were consumed before the night. The cost is the round: two rounds dearer than where he sat the year before, and two more every year he is held. The ladder is read off the earlier boards.

Through the full fantasy regular season, the price against what it bought. A keeper is not graded as a pick — his slot is a formula’s output, not the room’s judgment — but he cost a ninth; what does a ninth return? is answerable. Stands is where he finished among the season’s drafted players at his position, against the ones taken ahead of his slot; Return is that as a surplus; The round returns is what a selection made on the clock in that round has returned across 2021–2025, every board on file; the last column is the difference — the part of the return that was the price being right.

The picks that changed hands

9 picks were used by a seat other than the one the snake gave them to. Both ends: whose slot it was, who used it, and what it was spent on.

• JD sat out round 3; picked twice in round 11.
• Ryan sat out rounds 13 and 15; picked twice in round 3 and twice in round 5.
• Evan sat out round 11; picked twice in round 7.
• AJ sat out rounds 5 and 11; picked twice in round 13 and twice in round 15.
• Justin sat out round 7; picked twice in round 11.

The grade, through the full fantasy regular season

Every selection scored by where it finished among that season’s drafted players at its position, against where it was taken: the eighth back off the board who gave the second-best back season is +6. Within a position, within a season, against the eleven other people in the room — never against a projection nobody here ever saw. A keeper is a price, not a call, and is not graded. A player with no stat line is counted, never zeroed.

Three columns, always together. Surplus is the record. Adjusted is a model, and says so: the same selections with injury luck neutralised — a man is credited only the weeks he missed to injury beyond what he was priced to miss, at his own points per game, and the whole pool is re-ranked on the result. Nothing is ever debited: a man who missed exactly his price keeps his record grade, unforgiven. This season every man is priced at his position’s own rate — QB 1.4, RB 2.0, WR 1.8, TE 2.2 weeks, measured on this league’s own boards 2021–2024 — and a rookie at the rookie rate, because the availability model did not beat that rate out of sample and the plan said in advance what happens then. Risk taken is a fact about the draft, fixed from Week 1: every slot’s price weighted by where it was taken, the first pick in full and the last for one part in 180. With every price at a positional rate it moves only with how much early capital went to backs and tight ends, and this sentence is here so nobody reads more into it than that. The widest gap on the board is Joe’s: −13 on the field, +28 on health.

The adjusted board is a full re-rank, so pardoning one seat’s injured back can move another seat’s healthy back down a rank and cost him a point he did not lose on the field. That is the question the column answers — who drafted best in a world where injury luck was neutral — and in that world other people’s players are better too.

The pardons

What the adjusted column gave back, the 12 largest of 51: priced at so many weeks, missed so many to injury, and the difference credited at his healthy rate. † priced at the rookie rate; ‡ under 3 games played, the rate is the position’s median among drafted players.

The surplus is the record, not a model: nothing in that column is adjusted for anything. Re-read every morning in season; a rank moves when a stat correction lands, and every movement is a commit.