The proposal
A deal one manager sent another on this site, with the numbers their page showed them when they sent it. It records an agreement to go and make a trade on ESPN, and nothing more: the site makes nothing happen there.
The sender's page judged the deal from their side, and its figure for your side was its estimate of your roster's needs. Opened here, the same deal is judged again as your team, from your own roster, on the day you open it: both are shown, and you decide.
Agree here says you will make it on ESPN. Then one of you proposes the trade on ESPN and the other accepts it there, where the league's own review applies.
Start from a player
Deals worth a look
Every even deal, one for one and two for two, that holds the men you tapped: the men you would give, the men you want, or both. With no team chosen and only your own men tapped, every other team is searched.
The change in expected categories won in an ordinary week, for your roster and for theirs, with the deal against without it. It is the season part of the judgement only: open a row to have it judged whole, with this week, the playoff weeks and the finish. Both figures count each category once.
Deals in which both rosters gain categories come first, the one with the larger smaller gain on top. Nothing is left out for being lopsided; each row says what it costs the other side. On a build the order and Both gain categories are read for your side as your build counts it; the figure printed is not changed by it.
Both gain categories is a statement about the categories, and no guide to what the other manager would say. Measured on one league's history, 3 of the 50 trades both managers accepted had both rosters gaining by these figures, and half the sides lost by them (the star value study, 2026-10-13).
Beside each deal, what its players cost at this season's auction: the league's own market. A deal can be even on the categories and far from even there (a star whose free throws cost him categories, for a role player), and the other manager is likelier to read it the second way.
The deal, judged
As one number
How much to trust the number
Send it
to the other managerNothing here happens on ESPN. The trade is proposed and made there.
Build a trade
Pick the other team, then tap the men on each roster to move them into the deal. Tap a man again to take him back out.
Our roster
what we would give
Their roster
what we would get
Who fills it
The deal in full
What it does to
The number
The risk model's read
Each side is read on two things, never added together and never added to the judgement above: the change in its chance of taking this week's matchup, in points, with the noise of the draws beside it (inside about one point either way is near-neutral); and where the two men stand over the next four weeks, as percentiles among the league's rostered men, an order and never a size.
What it rests on
Every line is the one knowable on the morning of the deal. A man with few games of his own is marked thin: the rate under him could be anything.
Box Out's line: his own games this season, leaning on ESPN's projection for him while they are few. That is ESPN's latest projection where one was saved by the morning of the deal, else its preseason one; a man ESPN projected nothing for rests on his own games alone.
| Player | To | A week | Games so far | Games left | Playoff games | Rests on |
|---|
Proposals
| When | With | The deal | State | Open |
|---|