Basketball Trade Analyzer

Why z-scores, and why the percentages are weighted

Points, rebounds and steals are counted in different units and spread across the league very differently. A gap of three rebounds a game between two forwards is ordinary; a gap of three steals a game is the whole league. To compare a player's contribution in one column with his contribution in another, each column has to be put on the same scale, and the scale the category sites have used for years is the z-score: how far a player's projected line sits from the average of the pool, measured in standard deviations of that pool. A z of plus one in assists means he is one standard deviation above the pool in assists; a z of minus half in threes means he is a bit below. The pool here is the players our projections price for a league of your size and roster rule, the same pool that builds the category board, so a player's z on this page is the number beside him there.

Field-goal and free-throw percentage need one more step, and it is the step most quick calculators skip. A percentage is a rate, and a rate only matters in proportion to the attempts behind it. A backup center who shoots 62 percent on four attempts a game barely moves a team's field-goal percentage; a star who shoots 52 percent on twenty attempts moves it a great deal in the other direction. So the z for each percentage is weighted by attempts: the pool's average and spread are attempt-weighted, and each player's z is scaled by his share of the attempts, so a low-volume shooter cannot swing a column the way a high-volume one can. The ledger shows that z, and beside it the plain arithmetic a manager can check: each side's made-over-attempted rate and its attempts per game. Turnovers are the one column where fewer is better, and the API turns them before they reach the page, so a good ball-handler always reads as a positive figure and the ledger never needs a footnote about direction.

What the net means, and when it is too close

The net is the sum of the differences down the ledger: Team A's total z minus Team B's, over the categories still in play. A net of plus one says that, spread across the categories, Team A comes out about one standard deviation of one category ahead, which over a nine-column ledger is a real but modest edge. The stamp calls the trade too close when the net is inside a quarter of a z either way, because our projections are one model's opinion and a quarter of a z is inside what that opinion could be wrong by. It is not a coin flip in that case; it is a trade that the columns, not the total, should decide. Look down the ledger and ask which side gains the categories you have been losing on Sunday nights, and remember that a category won by a hair counts the same as one won by a mile: in head-to-head the ledger's edges are the score, not the sum.

The punt case

Punting is the core category strategy in basketball, far more than in any other sport. A roster built around a dominant big man concedes free-throw percentage on purpose and wins rebounds, blocks and field-goal percentage every week; a roster built around guards concedes blocks and wins threes, assists and free throws. For a punting manager a trade that loses the punted column costs nothing, and a trade analyzer that counts it is scoring the wrong game. The punt toggle in the picker keeps a punted category on the ledger, greyed out and struck through so you can still see what changes hands, and drops it from the net and the verdict. The same toggles are written into the link, so a trade you send to a league mate opens on your punts, and if he runs a different build he can turn them off and see his own answer without retyping the players. A free-throw punter trading for a poor free-throw shooter should see the stamp move in his favour when he flips the toggle, and that is the whole point.

Two for one, and the seat that opens

The most common trade in a category league is two players for one, and the most common mistake in judging it is to add up two values against one. The side that gives two players gets a roster seat back and fills it from the waiver wire, so the honest comparison is one player plus the best free agent who can fill the seat that opened, against two players. The page does that. Our pool publishes a replacement level for every position, the value of the next player available there in a league of your size, and the ledger adds it to the shorter side as its own line, named as an open seat and never hidden inside a player's number. It uses the least valuable player that side gave away, since that is the seat that actually opens, and because that player may be eligible at two positions it takes the better of the two waiver options, which is what a manager would do.

Multi-position seats and the two-center rule

Positional scarcity is weaker in basketball than in football, because most players carry two positions and every roster has guard, forward and utility seats that take anyone in the group. The roster-shape line above the ledger therefore counts players by their whole eligibility, a PG/SG as one PG/SG and not as a point guard, so you can see whether a trade leaves you with four combo guards and nobody who can fill a forward seat. The one real scarcity is the center seat, and it depends on the roster rule: the default shape starts two centers and two utility players, and in that shape a true center is worth more than his stat line because there are fewer players who can sit there. Set the roster rule to your league's before you read a verdict, because the pool is re-priced for it and the replacement level at center is the figure that moves most.

Games this week, and a Tuesday trade

Lineups in basketball are daily, and a fantasy week is anything from two to five games per club, so the same two players can be worth very different amounts in a given week. The projections behind the ledger are season totals and know nothing about the calendar, which is right for a trade you expect to hold until April and wrong for the matchup you are in right now. The roster-shape line therefore adds the games each side gets this week, read from the schedule, so a manager making a trade on a Tuesday can see that he is taking on three games and sending out seven before the weekend decides the matchup. For the week-to-week picture the schedule planner lays out every club's games and back-to-backs, and the streamers page ranks the pickups the open seat could hold.

Points and best-game formats

Under the points format the ledger collapses to arithmetic, which is the honest answer there: each side's projected points under the table your league scores, the MustDraft value beside it, the open seat at replacement, and the net. Two tables are offered because the platforms disagree: one scores makes and misses, with field-goal and free-throw attempts as minuses, and rewards efficient volume; the other scores six counting stats and does not care how a point was scored. The newer best-game formats, where only each starter's single best game of the week counts, are not a points total at all, and this page does not pretend to model them; a player's ceiling matters more there than his average, and the ledger has no column for it.

What this page does not know

Every number here is ours. There is no crowd-sourced trade market for basketball the way there is for football, so this page cannot tell you what other managers would pay, only what our projections say the players are worth in your categories. It does not know your roster or your opponent's, so it cannot tell you whether the categories you gain are the ones you needed, and it does not know your league's trade deadline, its roster limits, its weekly add cap or a commissioner's veto. Position eligibility is read from our directory and can lag a platform's own updates by a few days, so a forward the platform has just made center-eligible may still show as a forward on the roster-shape line. Injuries change the answer daily and the projection changes slowly; a player listed out for the season is still priced on what he would do if he played. The players, the format, the categories and the punts are all written into the address, so the link opens the same ledger for anyone you send it to.

The pool is priced by our own valuation from our projections, refreshed daily. If it is unavailable the page says so rather than showing a stale ledger.