Improve your roster
Waivers
Who to claim this week, in your league's scoring: who to add and what to bid or whether to claim — and compare drop options. Nothing is submitted for you.
Protected players are never suggested as drops. Nothing is protected until you tick a player: each drop option shows its rest-of-season cost where it can be priced, and a drop whose cost cannot be priced gets no bid. Your choices are saved in this browser for this league and team only. IR/taxi players are never offered as drops. This is not an instruction to move anyone out of IR.
The desk loads from the league panel above: enter your Sleeper username, or choose a team to view. Refresh it there before placing a claim.
Modeled top option
Bids and claim order are heuristics, not predictions of success. Refresh from the league panel, and verify injuries, roles, eligibility and deadlines in Sleeper, before placing a claim.
Claim options
| Add | Drop options (rest of season) | This week | Bid or priority | Signals | Why |
|---|
Worth watching
Available players not in the claim options above. A research queue, not buy recommendations: same-team RBs are possible insurance, not confirmed backups or beneficiaries, and popularity is not projected opportunity. Position filter applies here too.
Preseason watchlist · ECR order (PPR reference ranking), not a buy ranking
Bye-week coverage
Bye counts cover dedicated positions only, assuming other active-roster players are healthy. FLEX coverage and unknown byes need separate review.
How this works · data and limits
- Availability means absent from every league roster, including IR/taxi. It does not prove a player is an immediate free agent rather than awaiting waivers.
- Current-week recommendations require fresh aligned projections and a complete modeled roster baseline. Otherwise recommendations are withheld; research lists are not bid advice.
- For FAAB, the budget reserve is your choice, not a measured optimum. For rolling waivers, no dollar amounts are invented: current priority is context and add/drop alternatives remain value-based.
- Weak signals: an alternative whose move value (this week's gain plus the priced rest-of-season change, averaged per week through NFL week 17; this week's gain alone where the rest of season is not priced) is under 1.0 projected point per week stays listed for research but gets no bid range and no priority-claim suggestion. The 1-point cutoff (like the 2- and 5-point FAAB bands) is a heuristic in your league's points, not a validated noise or confidence threshold, and projection quantiles are not claim-success probabilities.
- Drop cost is priced as the change in the roster's lineup projected points over the remaining weeks, from the model's remaining-season projections (NFL weeks through 17, regardless of your league's schedule): what the add contributes to the roster, and what the drop then forfeits with the add on it. A bench player who never starts forfeits little whatever his season total. Values assume participation; injuries and returns are not forecast. A measured accuracy is shown under "Data sources and timestamps" only when it was measured under your league's scoring and this model.
- Rest-of-season values are a sum of weekly medians through NFL week 17, not a season median. Simulation-based drop costs are off for every league in this version.
- Spend guidance reads this week's gain plus that rest-of-season change, averaged per remaining week, through the same uncalibrated bands as before. A swap whose drop forfeits more than the add returns shows both numbers and gets no bid or priority claim; it is not called wrong. When a player has no rest-of-season projection, or the projections are stale or misaligned, the drop cost is unassessed and no spend guidance is offered for that swap — unknown is never priced as zero.
- No inferred hidden bids, calibrated win probabilities, automated transactions, or K/DST/IDP projections. No keeper-value premium is priced yet.
This week's warnings
This week's options
Reading the claim table
League format and heuristics
Roster coverage and rest of season
Data sources and timestamps
Historical results and their scope are on the about page; they are not claims about your league's scoring.
Research signals and market
Usage flags describe observed workload—not a breakout probability. Growth requires at least two comparable prior same-team games. Routes and red-zone usage are not available in this layer. Snap share is not route participation.
Usage sources: nflverse player statistics and PFR snap counts via nflverse.
Trending data: Sleeper. Counts are platform-wide activity, not selected-league demand. “Not listed” is not zero. No net-add rate or claim-success probability is inferred.
The preseason watchlist entries in "Worth watching" are ordered by preseason ECR (PPR reference ranking), not a buy ranking. They are unrostered now and may be stash candidates even without a projected lineup gain; role changes and injury contingencies are not priced by this version.
Open selected league in Sleeper · Read-only API documentation