Trade Ideas builds every deal it can find from the rosters and draft picks in your league: swaps that fit both lineups, each partner’s players bought with your picks and bench, and your veterans or stars sold for younger players, starters and picks. It checks each one against both rosters and the rules below, keeps the best few, and then Claude reads the shortlist with each owner’s trading history and ranks the ones that fit where your team is heading and that the other owner has a reason to accept, with the reasoning and a pitch. Claude can’t add a trade of its own, so every idea you see was built and checked by the app; one that missed a check is marked “A stretch”. Every idea should move your lineup or your value by at least 3% of the value traded, and the other side can’t be left more than about a fifth behind with nothing for their lineup. When fewer than five pass every check, the closest ones fill the run, marked “A stretch” with what they missed; you see fewer than five only when the league has nothing more to offer, and then the page tells you before the run is counted (“Only 3 trades fit your league with these choices right now”) so you can run it anyway or change a choice. When more than five pass every check you get up to eight: the first five, then “Show more”. Cards are ranked by what the deal does for you, scaled by how likely the owner is to say yes (the “Ranked …/100” line under each card shows the reasons), and each says what the other team gets out of it, a claim the server checks against their roster.
What the model sees for each partner, beyond their roster: a “Wants now” line (their goal, record and playoff odds, where they are short or deep, FAAB left, how they price players and how they negotiate, from their profile), their weakest starter at each position, injuries and this week’s byes on both sides, and how many weeks are left before the trade deadline.
Trade Ideas is part of Pro (20 runs a month); on the Free plan the page keeps Best fits, the closest trades on value and need worked out from the rosters, and any ideas from earlier runs.
Your team’s goal decides who you trade with
Trade with: leave it on Any owner, or pick one owner for five ideas with them alone (an owner’s profile on the Owners page links here as “Get trade ideas with …”). Best fits follows the choice too.
Each idea card says whether the other owner is likely to say yes (Likely, Worth a try or Long shot) and why: how the deal reads for them, whether they usually come out ahead in trades, whether they’ve gone quiet, and their chance of a trade in the next 30 days. Under it, the first notes from their owner profile on how they trade.
When your team is Win now, no idea brings you a player who is Out, on IR, PUP, suspended or Doubtful: he can’t help the lineup you’re trading to improve (Questionable players are fine, and a rebuilding team may still buy an injured player low). Every team is labeled Win now, In between or Rebuilding, the same labels the Trade Block shows. A contender is never paired with another contender, and a rebuilding team is never paired with another rebuilding team: a contender won’t help a rival win this season, and two rebuilding teams want the same picks and young players. The page lists the teams that were left out. While you’re in the playoff race, no idea makes another team in the race better this season either: an In-between team with a real playoff chance can take your picks, young players or depth, never a starting-lineup upgrade.
Auto takes your label from your max PF, lineup value, picks and age. If you see your team differently, choose Win now or Rebuilding and the ideas follow your call. The choice is remembered for that league in this browser, and on your account when you’re signed in.
Looking for something in particular
- Pick a focus (a position, Draft picks or Youth) and the ideas where one of those is a real part of what you get (at least 30% of it) come first. When too few of them fit together to make five, the closest deals at other positions fill the run under “More deals worth a look”, so a short list never comes back for a focus. Youth means a player aged 25 or younger, or a pick.
- Draft picks are part of the deals: when picks are in play, at least two of a full run’s five ideas include one (one, in a shorter run), priced Early, Mid or Late like the Draft Picks page.
- Running it again gives you different ideas. A new run never repeats an idea from your last five runs, and never comes back to the same partner for the same main player.
- Each idea shows both sides with values, the reasoning, and a pitch you can copy into Sleeper chat.
- Each idea is measured on both rosters: what it does to your best starting lineup, your value less anyone you’d have to cut, and how it reads to the other owner (they gain, close to fair, or a stretch). Ideas that don’t help you for the way you’re heading, or that no one would accept, are thrown out before you see them.
- "Best fits right now", under the goal, focus and partner choices and the run button, is free and needs no AI: the three teams whose rosters line up best with yours and a starting trade with each, from the same numbers as an owner’s Deal starters.
- "Open in Trade Calculator" loads the exact trade, picks included, so you can check the numbers and adjust it.
- A player who wouldn’t start for the team getting him counts for less than his value: a contender gives little for depth, and a rebuilding team little for an older backup. So no idea pays for a starter with two bench players. In a superflex or 2QB league neither side gives up a quarterback aged 25 or under without one coming back.
- The line under a card’s title says what kind of deal it is: the partner’s direction, "2-for-1" for a consolidation trade, and how likely they are to say yes. The spot you free is noted on the card but not counted as value, so a consolidation has to win on value or on your starting lineup.
- A run takes about half a minute in the background, and earlier runs stay available.
- Trade Ideas is off while your league’s trades are closed: past the deadline, in the playoffs, or switched off by the commissioner. A run also waits for the league’s history to load; if it fails, the page tries again for a few minutes. A one-for-one swap stays within 20% of even on value.