Grade My Draft

The four components, and their weights

Market value is the average gap, in picks, between where each of your players was taken and where the draft market had him: a guard taken at pick 40 whose draft position is 28 counts twelve in your favour, and one taken twelve picks early counts twelve against. That average sits on a scale where zero, the structural middle of a snake draft, is a plain C. Starting lineup is the projected value of the best legal lineup the roster can start, with every player seated by his full eligibility list, ranked against every other team in the league. Depth is the same ranking for the bench. Balance starts at 100 and loses 25 for every required starting seat left empty at the end of the draft and 10 for every position where you drafted more than three players past the number of seats you could start them in. The weights are 35, 35, 15 and 15, and the letter is the weighted sum.

Multi-position eligibility changes what balance means

A basketball roster is not five silos. A player listed at guard fills the point guard seat, the shooting guard seat, the guard seat or any utility seat, so the balance component counts the seats a player could occupy, not the one his position name suggests. Under the two-center rule that leaves one real scarcity, and the strip under the report says so: a roster with eleven guard-eligible players and one center has every guard seat covered four times over and its second center seat empty, and the balance score reads it exactly that way. Under the three-utility rule the same roster grades better, because there is only one center seat to fill and three seats anyone can take. Choose the roster rule your league actually uses before you paste; the grade is a different number under each, honestly.

The category strip

Under a category format the report card carries a strip with one cell per category in play. Each cell is the roster's z-score in that category summed over every pick, so a positive figure means the roster is above the pool's average there, and the two thinnest kept categories are marked in the results colour, because a category draft is won five categories to four and a roster that is thin in two of them has two weekly coin flips it does not need. A punted category stays on the strip, struck through, and is left out of every value on the page. Field-goal and free-throw percentage are weighted by attempts before the z is taken, so a high-volume 48 percent shooter and a low-volume one do not read the same, and turnovers are flipped so that fewer counts as better. A points league has no strip, because a points league has one number.

Steals and reaches against the market

Under the league table every pick is listed with its market rank, the player's place in draft-position order across the draftable pool. A player taken twelve or more picks after his market rank is marked as value, one taken twelve or more picks before it is marked as a reach, and everything inside that band reads as on price. Twelve is deliberately a full round: draft position moves by a few picks from room to room, and calling a pick a reach because it landed four spots early would be reading noise. A player the market has no number for is marked as having no price, and the page counts how many of a team's picks the market priced, because that count is what the overall grade rests on; a team with a pick the market cannot price has its overall letter withheld rather than guessed, with the lineup, depth and balance tiles still shown.

Why draft position is the yardstick

The market here is average draft position from real drafts on the platform that publishes it, the same order the mock draft room seats its computer teams by. It is the yardstick rather than our projections for one reason: a grade should measure the decision against what was knowable in the room, and the room knew the market. A center taken in round two when the market had him in round five is a reach on this page even if he averages a double-double, because the grade is scoring the draft, not January. The projections come in through the lineup and depth components, where what a roster can actually produce is the question, and every one of those values is MustDraft's own; there is no crowd trade market for basketball to borrow from, and the page says so beside the results.

What the grade cannot see

It cannot see games per week, so a bench built for the four-game weeks grades the same as one that was not. It cannot see your league's injured-list seats, which your platform fills with players the grade counts as bench, and it cannot see keepers, which distort the market for everyone. It cannot grade a best-game format such as the ones the hosts now hand new leagues, where a single game a week counts; grade those drafts under the six-stat points format, which is the closest scoring to what a best game rewards, and read the letter loosely. And it cannot see the draft that happened in the room next door: the market is an average, and yours was not the average draft.

After the grade

The grade ends on draft night. From there the schedule planner shows how many games each of your clubs plays this week and where the back-to-backs land, the streamers page ranks the pickups on the thin nights, and the injury report tells you who is out before the first lineup locks. Share the card with the league; the manager with the lowest grade will want to argue about the punt, and the argument is the point.