Stick, Perf, Carry, Edge, NTV, and TTV

The public performance ratings use a league-relative scale where 100 is average. Missing performance data displays as an em dash rather than an assumed average score.

Stick

Stick asks whether the player's performance is supported by repeatable underlying ability. PVL calculates it from MLB pitch-by-pitch and batted-ball records. Hitters are evaluated on contact ability, plate discipline, and contact quality. Pitchers are evaluated on missing bats, control, contact allowed, and pitch characteristics. An approximate fielding component has a smaller influence. High Stick means the underlying evidence supports the player's ability.

Inputs are compared with league peers, grouped to limit duplicate credit, and adjusted for sample size. Up to three seasons contribute, with older evidence receiving less weight. A score of 100 is average within the hitter or pitcher reference group; 120 is one standard deviation above it. This is a skill estimate, not a guarantee of future results or career length.

Missing measurements reduce confidence rather than becoming bad scores. Bat speed, running speed, throwing-arm speed, and precise catch probabilities are not included. Fielding difficulty uses comparable batted balls and remains an approximation. Hitter component weights are calibrated against a later season, then checked on a subsequent season. Pitcher weights use a fixed blend checked against later results. Neither is claimed to be an optimal forecasting formula.

Perf

Perf measures accumulated wins above replacement, anchored to actual team results. First, batting, baserunning, fielding, and pitching contributions allocate every team's actual wins among everyone who played for that team. Then we subtract the contribution expected from replacement players over the same workload. Traded players retain credit for each team stint; two-way players receive both batting and pitching credit.

Hitters, starters, and relievers share one rating scale. A 100 is the average accumulated contribution in the reference group; 120 is one standard deviation above it, and 140 is two above it. These are relative scores, not percentages. Playing time matters: this is season output, not a per-inning or per-appearance talent rating.

The reference group includes players with at least half a season-to-date role workload, using 600 plate appearances, 180 starter innings, or 65 relief innings as a full-season workload. Workloads combine for players with multiple roles. The current replacement rating is 85.9, below average. Replacement rates use players identified through preseason availability and remain an empirical estimate, not an official statistic.

Team results affect the final credit. This is an allocation of wins that happened, not a prediction of how many games a hypothetical roster would win. Negative contributions are retained. Only allocated Player Wins add up to actual team wins; wins above replacement and normalized ratings have different totals.

Check the team win totals

Carry

Carry starts from Perf, compares the player with the measured support around him, and adds a bounded adjustment for offensive, pitching, and defensive clutch value. Run value includes runs scored and the change in base-and-out expectancy. Close games and later innings increase the pressure bonus; large leads and deficits are treated symmetrically. Pitchers receive the expected value of contact allowed, while fielders receive the difference between that expectation and the actual outcome. Routine outs mostly credit the pitcher; difficult catches earn more defensive credit. Low-pressure plays receive no extra clutch bonus.

Great production receives more Carry credit when the rest of the team is weak and less when the player is surrounded by strong performers. A performance gate limits the roster adjustment for players with low production. The clutch adjustment uses factual play-by-play and our own models. Estimated fielding difficulty is based on comparable balls, not precise starting positions. Unidentified defensive credit remains unassigned. When clutch data is unavailable, that adjustment is omitted.

Edge

Edge is the flagship complete-player rating. It combines the legitimacy of the underlying skill profile with actual individual production and team-relative carrying power.

Weighted Edge = 40% Stick + 30% Perf + 30% Carry

Offset average = ((Stick − 100) + (Perf − 100) + (Carry − 100)) ÷ 3

Base Edge = Weighted Edge + Offset average

Final Edge applies a light positional multiplier to Base Edge. The adjustment ranges from 0.95 to 1.04, rewards scarce premium positions modestly, and uses the highest listed weight for multi-position players. Two-way players receive 1.04.

Longevity Index

Longevity Index estimates how well the player's skill set should age. Discipline, contact, command, and versatility generally age better than profiles dependent on speed, defense, max-effort velocity, or one-dimensional power.

Contract Efficiency

Contract Efficiency compares projected production with salary and below-market control. Scores above 100 represent favorable commitments; scores below 100 represent contracts that cost more than the projected performance supports.

NTV: Neutral Trade Value

NTV is the roster-independent asset value. Its performance inputs are the same Perf, Stick, and Carry ratings shown on player profiles, with Stick projected over the remaining control period. It also considers contract efficiency, control, age, longevity risk, position, scarcity, confidence, prospect pedigree, and downside severity. Zero is neutral, so harmful contracts and declining assets can become meaningfully negative.

TTV: Team Trade Value

TTV starts with NTV, then adds organization-specific context such as roster fit, positional need, overlap, contention window, and the other players moving in the trade.

2026 team win totals

Includes all contributors during the season, including players who have since left the roster. Values shown to two decimals.