Value is in percentage points of weekly matchup win probability: 12 means about 12 points more than an average rostered player. Pick a punt and values are recalculated for a team that gives that category up. Everything is computed for the league size chosen at the top, and cell shading shows the z-score against that league's rosterable pool.
Z total is the sum of the nine category z-scores, leaving out the punted category. FG% and FT% are volume-weighted impact scores, so a low-volume shooter does not top the list. Turnovers are flipped: higher is better.
A Marcel-style projection: three seasons weighted toward the most recent, pulled toward the position-group average, with an experience adjustment and separately modeled minutes. The Now column is what each player has done so far this season, so the gap flags players running hot or cold against expectation.
Projected and current totals are sums of per-game z-scores. Early-season samples are small, so treat the gap as a prompt to look closer, not a verdict.
Add the players leaving and joining your roster. Each category shows the net change in z-score for your team, and the total value uses the punt you pick.
The comparison adds per-game z-scores and does not account for games played per week, injuries, or who fills an open roster spot.
How often each category decides a close matchup, from simulated snake drafts played out over real weekly game logs. A higher number means winning that category more often flips the result.
Scale: 1 means the category always decides the matchup, 0 means the other eight have already settled it. All nine sit in a narrow band for an average team. The gaps widen once a team commits to a build.