Our category board next to where players are really being drafted. The useful part is the gap: the players our board ranks above where the room takes them at their position, and the ones going earlier than our board would take them. Punt a category and our side re-ranks while the room stays put, which is exactly the disagreement a punt build drafts into.

COURT GUIDE

What the gap tells a category manager

The gap compares two orders inside one position: where our board ranks a player among, say, the power forwards, and where the room drafts him among them. Our side is the same category value the rankings sort on, scored in the format you pick and with your punts taken out; the room’s side is the average pick in real drafts. A +9 on a center means our board has him nine places higher among centers than the room takes him. We compare inside the position because the seats are uneven: a league with two center seats makes centers scarce, and an overall comparison would light up every big man at once for a reason that has nothing to do with the player.

A gap is a disagreement, not a prediction, and one side of it is ours. The room may already know about a minutes battle, a camp injury or a new role that our projections have not absorbed, so read the news before you spend a pick on a number.

Read the rest

Punting moves our side and never the room’s

Nobody publishes a draft market for managers who punt free-throw percentage. Every average pick on this page comes from drafts full of managers building every sort of team, so when you punt a category the room’s order stays exactly where it was and only our side moves. That is the point of the toggle. Punt free throws and the big men who shoot poorly from the line climb our board, because the category that was dragging their value down is gone from the sum, while the room still prices them as though free throws counted. The gap that opens is the discount your build is entitled to. Punt assists and the scoring forwards and shot-blocking centers rise; punt turnovers and the high-usage guards who handle the ball on every trip stop paying a penalty. Try two punts at once and the board becomes a very particular shopping list, which is the right way to read it: a list for one kind of team, not a ranking of who is good.

A punt is decided on your own roster, not on this page. If the players the gap points you toward are not the players your first three rounds set you up for, the gap is telling you about a different team.

Two platforms, two rooms, one blended position

The draft position here is built from two platforms’ average picks. Where both list a player we use the median of the two, which for two numbers is the point halfway between them, and where only one lists him that platform’s own pick stands alone. Nothing is estimated for a player a platform does not list. The Platform spread column shows the earliest and latest average pick across the platforms that list him and, underneath, how many picks separate the two. A dash means only one platform lists him, so there is no spread to show.

The same player can go rounds apart. Each platform shows its drafters its own pre-draft rankings, which anchor every pick made in its rooms, and the platforms differ in their default league size and default scoring, so a player whose value depends on the scoring can be a fourth-round pick in one room and a seventh-round pick in the other. The widest splits inside the first ten rounds are mostly centers. If your league drafts where he goes late, you can wait on him; if it drafts where he goes early, the blended number can make you wait a round or more too long. Your own draft room shows its average pick beside every name, and the spread tells you which names are worth that second look. A spread of a full round or more in a twelve-team draft is set in stronger type. The band above says how many players each platform prices and how many both do; the smaller room prices far fewer players than the larger one, so past the first ten rounds dashes become common.

If your league drafts on Yahoo

The My league drafts on switch decides which room the board reads. All platforms, the default, is the blended position described above. Pick Yahoo and the ADP column becomes Yahoo’s own average pick, Room rank becomes the player’s rank at his primary position inside Yahoo’s drafts, and the gap, the tiles, the draft order and the disagreement sort are all worked out again against that room, so a punt build sees its discount in the room it will really draft in. The platform spread, its sort and its tile step aside, because they compare two rooms and you have chosen one. Yahoo prices roughly the top 190 players, so deeper than that a row shows a dash and falls to the end of the draft order rather than carrying an estimate, and a line over the board says how many of its players Yahoo prices. The choice rides in the link the copy button makes. Yahoo is the only platform offered on its own because we show a single platform’s room only where that platform’s data may be shown by itself, and its terms allow it with the credit printed under the board.

The format decides our side

Nine-category leagues count points, rebounds, assists, steals, blocks, threes, both shooting percentages and turnovers. Eight-category leagues drop the turnovers, which quietly lifts the ball-dominant guards, because the one category that punished their usage is gone. Both shooting percentages are weighed by attempts: a center who makes seven of ten at the rim moves your field-goal percentage far more than a guard who makes two of three, and the board counts him that way. Points leagues have no categories to punt, so the punt row goes away and a table chooser takes its place: Makes and misses scores every made shot and charges every miss, and Six stats scores points, rebounds, assists, steals, blocks and turnovers with no shooting line at all. Roto uses the same nine categories as the head-to-head nine-category board, and our side is the same, because a category is worth the same in a season-long standings table as it is in a weekly matchup. What changes in roto is the punt: with no weekly matchup to concede, a punted category costs you its whole column of standings points for the season, so most roto managers punt less and later.

Four ways to sort the same board

Draft order is the room’s view, the board as the market is taking it, and the sort to keep open during a draft to see who is likely gone before your next turn. Biggest disagreement puts the widest positive gaps first, so the players our board likes most against the room lead the list, and the players going earlier than our board would take them sit at the bottom. Our value is our board alone: the sum of the player’s category scores in a category format, his projected fantasy points in a points format. Biggest platform split, once the platform spread is on the page, lists the players the two rooms disagree about most, which is where knowing your own platform pays. Narrow the position and every sort reads inside it, which is where a gap is worth the most.

Where the two numbers come from

Our side is the value pool behind the basketball rankings. For each statistic it takes the median of what the independent forecasters published, one vote per forecaster, and our own model fills in only for a player nobody outside projects; it never votes beside them. The Sources column counts the forecasters under each line, or reads Ours when our model’s line is all there is, and a gap on an Ours row is printed but never coloured, because our model reads last season’s box scores and sees little of a rookie. Our rank is that line ranked at the player’s primary position, and Room rank is the blended draft position ranked at the same position. The gap is the room’s rank minus ours, a subtraction you can check on the row. A player listed at two positions is filtered under both and ranked at the first.

Reading the late rounds honestly

An average pick deep in the draft is thin evidence. A player taken in a handful of rooms around pick 140 carries the same average as a player taken in every room at pick 140, and the page cannot tell the two apart. That is why the tiles at the top only look at the first ten rounds of the league size on the board, and why a big gap on a player the room takes after the tenth round deserves a second source before it deserves a pick.

Where to go next

The consensus table shows every forecaster behind a projection and how far apart they are. The mock draft room lets you practise a punt build against this market, the auction values price the same board in dollars, and the live companion is this board with players leaving it as they are drafted. The arithmetic behind the value is in the methodology.

Where our board disagrees with the room