Prospect Pipeline

No single target: a ranking of current performance (Perf+) and age-for-level, labeled by readiness, with an expected next-season Perf+.

• MLB Stats API minor-league lines 2025–26, by league and club
• Birth dates (MLB Stats API people)
• FanGraphs prospect board (FV, ETA) and its trade-value table
• Savant AAA and FSL pitch tracking for MiLB Stuff
• The organization today, Rule 5 status and options (keystone_models.org)
• The Winter Plan’s Rule 5, trade and 2027 roster calls (decisions feed)

• Perf+: OPS+ (hitters) or K−BB% (pitchers) against his own league first (the PCL and Cal League inflate offense), PA-weighted across leagues, then standardized among qualified players at the level (100 = average, 10 = 1 SD)
• Park (hitters): each affiliate’s park read from its team lines both ways (its hitters’ OPS+ and the OPS+ its pitchers allowed, averaged), 2025–26, regressed by the reading’s year-to-year repeat rate. Tested before use on hitters with full seasons in both years: a change in his park index (each club’s index without his own line) showed up in raw Perf+ at only about a third of its size (± 0.18), so Perf+ takes out that share, never more than the test’s upper 90% bound
• Whole season: each level’s Perf+ weighted by its sample (Murf Gray: Single-A, then High-A, then Double-A, each level’s Perf+ shown); the tag reads the season, the level path is shown
• Expected next season: 100 + (Perf+ − 100) × n / (n + k), k fit on every club’s minor leaguers 2025 → 2026 (about 675 PA for hitters, 550 batters faced for pitchers); Trend marked only beyond the ordinary year-to-year change (about ±10)
• Age vs. level: his decimal age on July 1 (birth date) minus the PA- or BF-weighted average decimal age at the level
• Readiness: push (Perf+ ≥ 110, no more than half a year older than the level’s average age; both as printed), on track, watch, stalling (< 92 and half a year or more older than the level’s average), small sample; 26+ and ungraded = org depth (no development tag), the same age line that defines a development prospect; a player already promoted past the highest level he reached is labeled with where he is now
• Grade vs. performance (a screening read): expected Perf+ against the median for his FanGraphs grade across every organization, flagged only outside the 80% band of his own estimate
• Other organizations’ lines: a season played mostly with another club (complex and DSL clubs matched by name) is labelled with it and left out of the development counts
• Rule 5 eligibility from keystone_models.org (CBA rule: draft or signing date and age on June 5); the protect/expose call, the 40-man count and pick risk are the Winter Plan’s
• Trade value: FanGraphs’ FV value table, the same pool the trade-package rule draws from (top 3 kept out by default; a tie at that line is shown); a 2026 draftee carries the trade pool’s provisional value: Keystone’s draft curve at his pick scaled to FanGraphs’ table (the table-to-curve ratio on the graded 2025 class), labelled with its range at that scale

• Box scores only below Triple-A and the FSL: no tracking at Double-A or High-A.
• No home/road splits for the minors in our data: the park index mixes in team and organization talent, which is why only the tested share of it is taken out.
• Perf+ is one season of results (about a third of a median season’s distance from 100 carries into the next, about 43% of a full 500-PA season’s); scouting grades (FV) carry the tools it can’t see.
• MiLB Stuff is pitch quality, not a forecast: AAA Stuff barely predicted MLB K−BB% on the forward test.

• Share of “push” players who hold Perf+ ≥ 100 the next season (2025 → 2026: about 70%, against about 45% of everyone else)
• Rule 5 losses among exposed players