How the numbers work
The League Times converts each platform into one normalized season model before any statistic runs. That keeps definitions consistent across Sleeper today and Yahoo next.
Official reality
Actual scores use the platform's reported totals, including commissioner corrections. Regular-season boundaries come from league settings.
Legal alternate history
Optimal lineups obey every configured starter slot, multi-position eligibility, flex rule, and negative score. Missing historical roster data is surfaced as a warning.
Explicit edge cases
Ties count as half a win in expected-win rates. Sleeper median games are included in actual and optimal records. Playoffs and consolation games are excluded from core season analytics; completed winners-bracket games are imported separately for rivalry records.
Definitions
pointsLeaderThe manager whose starting lineups scored the most total points during the regular season.
Format: points · Inputs: weekly lineups · Audit tolerance: 0.001lineupEfficiencyActual starter points divided by the maximum points from a legal lineup each week.
Format: percent · Inputs: weekly rosters, lineup slots · Audit tolerance: 0.001scheduleLuckActual wins minus expected wins based on weekly all-play performance.
Format: wins · Inputs: weekly matchups · Audit tolerance: 0.001biggestLineupMissThe single week with the largest difference between actual points and the highest-scoring legal lineup.
Format: points · Inputs: weekly rosters, lineup slots · Audit tolerance: 0.001actualStandingsTeams ranked by their official regular-season record, including platform median games when enabled, with points scored as the tiebreaker.
Format: record · Inputs: weekly matchups · Audit tolerance: 0optimalStandingsEvery weekly matchup is replayed after giving each manager their highest-scoring legal lineup. Opponents, schedule, and enabled median games stay the same.
Format: record · Inputs: weekly rosters, lineup slots · Audit tolerance: 0scheduleLabSeason superlatives derived from weekly opponents, scoring variance, and optimal-lineup standings.
Format: collection · Inputs: weekly matchups · Audit tolerance: 0.001scheduleStrengthThe highest average points scored by a manager's weekly head-to-head opponents.
Format: points per week · Inputs: weekly matchups · Audit tolerance: 0.001consistencyThe population standard deviation of actual weekly starter points. Lower variation means more consistent scoring.
Format: points · Inputs: weekly scores · Audit tolerance: 0.001heartbreakThe highest team score that still resulted in a head-to-head loss during the regular season.
Format: points · Inputs: weekly matchups · Audit tolerance: 0.001greatEscapeThe lowest team score that still resulted in a head-to-head win during the regular season.
Format: points · Inputs: weekly matchups · Audit tolerance: 0.001biggestRankJumpThe largest improvement from actual standings rank to optimal-lineup standings rank.
Format: rank · Inputs: actual and optimal records · Audit tolerance: 0allTimeLeaderboardRegular-season results combined across every linked completed season. Managers are matched by stable platform owner ID even if team names change.
Format: table · Inputs: linked seasons · Audit tolerance: 0.001seasonsPlayedNumber of completed linked seasons in which this owner managed a roster.
Format: count · Inputs: linked seasons · Audit tolerance: 0historicalRecordCombined official regular-season wins, losses, and ties across the available league history.
Format: record · Inputs: linked seasons · Audit tolerance: 0historicalOptimalRecordCombined wins, losses, and ties if every matchup used optimal legal lineups.
Format: record · Inputs: linked seasons · Audit tolerance: 0historicalPointsCombined actual regular-season starter points across all available seasons.
Format: points · Inputs: linked seasons · Audit tolerance: 0.001historicalEfficiencyCombined actual points divided by combined optimal lineup points across all available seasons.
Format: percent · Inputs: linked seasons, weekly rosters · Audit tolerance: 0.001managerScorecardA side-by-side view of results, scoring strength, schedule luck, and lineup decision-making for every manager.
Format: table · Inputs: weekly matchups, weekly rosters · Audit tolerance: 0.001actualRecordThe manager's official regular-season wins, losses, and ties.
Format: record · Inputs: weekly matchups · Audit tolerance: 0optimalRecordWins, losses, and ties after replaying each matchup with both teams' highest-scoring legal lineup.
Format: record · Inputs: weekly rosters · Audit tolerance: 0pointsForTotal points scored by the manager's actual starting lineups.
Format: points · Inputs: weekly scores · Audit tolerance: 0.001allPlayThe record produced by comparing this team's score with every other team each week, removing schedule effects.
Format: record · Inputs: weekly scores · Audit tolerance: 0luckActual wins minus expected wins. Positive values indicate favorable weekly matchups; negative values indicate an unfavorable schedule.
Format: wins · Inputs: weekly scores, actual record · Audit tolerance: 0.001efficiencyActual starter points divided by maximum legal lineup points. Higher means fewer usable points were left on the bench.
Format: percent · Inputs: weekly rosters, lineup slots · Audit tolerance: 0.001benchPointsTotal points this manager could have gained by starting their highest-scoring legal lineup every week.
Format: points · Inputs: weekly rosters, lineup slots · Audit tolerance: 0.001weeklyActualPointsPoints scored by the lineup the manager actually started that week, including platform score corrections.
Format: points · Inputs: weekly scores · Audit tolerance: 0.001weeklyOptimalPointsMaximum points possible from a legal lineup using players on that week's roster.
Format: points · Inputs: weekly rosters, lineup slots · Audit tolerance: 0.001weeklyPointsLeftOptimal lineup points minus actual starter points for this week, never below zero.
Format: points · Inputs: weekly rosters, lineup slots · Audit tolerance: 0.001lineupChangesPlayers marked Start appear in the week's optimal legal lineup but were benched. Players marked Sit were actual starters omitted from that optimal lineup.
Format: players · Inputs: weekly rosters, lineup slots · Audit tolerance: 0Retrospective positional value
Each non-keeper selection is compared with every other drafted player at the same position. Expected rank comes from draft order; actual rank comes from points scored while the player appeared on a league roster. A pick starts at 82, then gains or loses as many as 36 points based on how far it beat or missed its positional expectation. Grades are capped from 45 to 100. Keepers are shown on the board but excluded from pick and manager grades.
Repeatable snapshots
Completed reports are cached for 24 hours and carry their import timestamp and calculation-engine version. Active seasons refresh every five minutes. Changing the calculation version makes future cached reports independently identifiable and auditable.