The number
Picks priced in points, not stars
A five-star sticker and a printed magazine board flatten every player into a rank. Surplus prices the same decision in fantasy points, under the scoring you actually play.
Local guest state No trackers Published methodology Works offline
Value over baseline
VOB is points above the last startable player at that position. It is the ranking number on this board — not raw projected points, which make a quarterback look like a bargain against a running back because the scales are different.
Two receivers in the same tier on a sticker sheet can be eight points apart once your PPR and your roster depth are applied. The sheet cannot see that. The board recomputes when you change scoring.
What a printed board cannot do
A sticker kit or a magazine pull-out is fixed the week it is printed. It does not know whether you play half-PPR. It does not know you kept a back in the third. It does not move when the room reaches for tight ends. Those are not flaws in the paper — they are what paper is.
Surplus is a live board: projections in your scoring, ADP next to the model rank so the gap is visible, and the clock on the room you are actually in. The mark on the app is that split — the grey base is what the market pays, the green cap is the surplus above it.
A shorter page that states the same fact for search: the draft board that prices picks in points.
NBA, MLB, and NHL
Those boards use the same VOB idea, but season value is rate × eGP and the football missed-game haircut is not applied. Points columns print per game (Proj/g, VOB/g); ranking still uses season VOB. The house-model figures on this page and on the backtest are fantasy football only.
A number we can show our work on
The house projection keeps vendor volume and replaces vendor efficiency with
shrunk prior-season rates. Cross-validated over seven season pairs — fitted
on six, scored on the seventh, rotated — against extrapolating last season’s
efficiency, the assumption most projections ship with: mean absolute error
dropped 21.8% (18.04 → 14.09). The script is
tools/backtest.mjs in the repo. That is a comparison to a
baseline, not a claim that any other product is wrong.
That number hands both models the player’s actual future volume, which nobody has in August. Give them both last season’s volume instead and the gain falls to 0.9%. Production sits between the two, because the vendor projects volume rather than copying it. Both arms, every fold.
See the numbers on a real board. Free, no account.
Open the draft board