Every team. One place.

Leave a bound blank for no limit. All active bounds apply together. Fantasy avg/game filters the per-game average; other stats use period totals.

Limits apply to displayed values. Production limits follow your per-minute or per-36 selection. Fantasy points / 36 and Fantasy points / min always use their labeled units. Leave a bound blank for no limit.

True shooting % = points ÷ [2 × (field-goal attempts + 0.44 × free-throw attempts)]. Effective FG % = (field goals made + 0.5 × threes made) ÷ field-goal attempts. Both are shown as percentages.

AST/TO divides total assists by turnovers. A dash means no turnovers or insufficient data. Production divides each stat by minutes played, then multiplies by 36 when selected. Rates use combined totals, not an average of game percentages.

Usage is a box-score estimate of the share of team plays a player uses while on court. It uses shots, free throws, turnovers and team minutes; across games, it weights the estimate by the player’s estimated opportunities. It is not ESPN’s published usage metric or an exact possession-tracking measure. Rounded box-score minutes can affect this estimate. Small samples and live games are provisional.

If this trade opens a roster spot, choose the player you would add. If it needs an extra spot, choose the player you would drop. Their production is included in the verdict.

Select up to the number of roster spots affected. Availability and legal starts still matter.

Uses actual ESPN regular-season totals and your scoring weights, including missed shots, offensive rebounds and double/triple-double bonuses. The date range on other tabs does not apply here.

The estimated fantasy points per game blends the latest season with up to three prior seasons. The latest season receives 70% weight after 30 games, proportionally less in smaller samples. Prior seasons with at least 10 games are weighted by recency. With no qualifying history, the estimate is the latest season average. These are Stew’s Fantasy Corner statistical estimates, not ESPN projections.

Rookies with 0–5 NBA games receive a provisional, semi-conservative estimate using available college stats, conference, NBA draft position and current roster competition. At 6 NBA games, actual NBA stats take over. Missing college data uses a labelled position baseline. These rookie assumptions have not been validated as predictive forecasts.

Trade tiers use the absolute difference in combined estimated fantasy points per game, rounded to two decimals: 0–3 is Even, above 3 through 10 is Conditional lean, and above 10 is Deny. The Deny tier flags an unbalanced package in either direction; the explanation identifies which side loses value. Roster size differences, injuries and incomplete history add context without changing the numerical tier. Explanations are tailored statistical analysis, not AI-generated text.

Historical range shows season averages, not a guaranteed floor or ceiling. The middle half of individual games describes scoring variation when game data is available. Injuries flag uncertainty; the calculator does not guess return dates or pretend to know future minutes, remaining schedules, your roster or waiver pool. Uneven deals show the extra roster spot’s break-even production.

Season data refreshes when you open or refresh this tab, with up to 15 minutes of caching. Advice updates when you analyze or save scoring settings. Offseason comparisons use the latest completed season and are labelled accordingly.

Manual mode: choose active slots. Daily estimates select the highest projected legal lineup each day; these are possible starts, not saved ESPN lineups.

For fixed lineups, select only your active starters in each roster. Future additions and drops are not assumed.

Team offensive and defensive efficiency use points per 100 estimated possessions. Possessions are estimated from field-goal attempts + 0.44 × free-throw attempts − offensive rebounds + turnovers, averaged across both teams. Pace uses 48 minutes and accounts for overtime. These are box-score estimates, not official NBA tracking metrics.

Player production blends full-season and recent per-minute production and minutes. True shooting is shown as an efficiency measure already reflected in that production. Position defense compares actual fantasy scoring with the production expected from the players faced, normalized for minutes. It uses ESPN’s primary positions, not tracked defensive assignments. Small samples are pulled toward neutral; correlated defensive indicators are blended.

When a rotation player is listed OUT for today, the model checks historical whole-game defensive results with and without that player. It requires at least five games in both groups, adjusts for opponent offense and limits the adjustment. This is an association, not proof of a defender’s individual impact. Missing evidence means no numerical boost. Future injury availability remains conditional; OUT today does not automatically mean OUT all week.

Validation predicts the last 56 available game dates using only earlier games, with no present-day injury information. Mean absolute error measures average error per player-game; smaller is better. If the matchup model fails to improve on the season-average baseline, totals use that baseline and the experimental estimate stays visible. This is an initial check, not a calibrated win probability.

Only complete game records are used. No results are invented for unavailable games. Refresh to retrieve current roster status and schedules; source caching can delay updates by up to 15 minutes. Projections pause during live games to avoid double-counting points.