Footnote
GuideStart Free
Guide
As of May 2026Footnote Metrics2 min read2 references cited

Fair Market Value Logic — Statistical + Comparative + Club Calibration

Squad ROI's Fair Market Value is Footnote's Phase M core logic. FMV computation has three layers: (1) statistical model (positionBase × ageCurve × pvsMultiplier × minutesPlayed × clubScarcity), (2) comparative market (median of similar verified transfers), (3) club calibration (G9 EMA learning from outcome data). This article publishes coefficients, thresholds, and integration weights, ensuring reinforcement staff can trust FMV as a basis for decisions.

3-Layer Construction

FMV is not a single formula but 3 layers. Statistical sets the baseline, comparative-market grounds it in real transactions, club-calibration optimizes per-club.

  • Layer 1: Statistical model (statistical-model.ts) — fixed coefficients (Wave M3 baseline, J-League 10-year average)
  • Layer 2: Comparative market (comparative-market.ts) — transfer_records median, ≥3 samples for reliability
  • Layer 3: Club calibration (calibration.ts, G9) — EMA from roi_decisions outcome, applied with ≥3 samples, clipped to [0.5, 2.0]

Final FMV = (Layer1 × 0.6 + Layer2 × 0.4) × correction_factor. Pre-data-accumulation, Layer 1 is 100%.

Statistical Model Coefficients

Position base values: GK ¥15M, CB ¥25M, CMF ¥30M, CF ¥35M. Age curves peak at different ages by position. PVS multiplier non-linear (PVS 80 → 2.3, PVS 90 → 3.0).

Position base values

  • GK: ¥15M (low scarcity, suppressed salaries)
  • CB: ¥25M (standard mid-tier)
  • SB: ¥20M (high attrition, hard to extend)
  • DM: ¥25M (rare specialists)
  • CM: ¥30M (versatile, in-demand)
  • AM: ¥35M (scoring + creativity, commercial appeal)
  • WG: ¥32M (individual-skill dependent)
  • CF: ¥35M (direct goal contribution, top tier)

PVS multiplier

Non-linear scaling: PVS 50 = 1.0 baseline; PVS 60 → 1.3; PVS 70 → 1.7; PVS 80 → 2.3; PVS 90+ → 3.0. Top prospects get exponentially higher valuations.

Minutes Played × Club Scarcity

Minutes factor reflects 'actually deployed?' Club scarcity reflects 'few competitors at this position?' Combined to reflect practical evaluation.

  • Minutes 80%+: 1.2 multiplier
  • 60-80%: 1.0
  • 40-60%: 0.85
  • 20-40%: 0.6
  • Under 20%: 0.4 (bench-warming reality)

Club scarcity: 1 player at position = 1.3 (irreplaceable), 2 = 1.1, 3 = 1.0, 4+ = 0.8 (oversupply).

G9 EMA Calibration — Per-Club Learning

G9 updates correction_factor via EMA: factor_new = factor_old × 0.8 + (actual/predicted) × 0.2. Clipped [0.5, 2.0]. Applied with sample ≥3.

Example: factor_old=1.0, new sample ratio=1.5 → factor_new = 1.0×0.8 + 1.5×0.2 = 1.1. Repeated 20 times with same ratio → converges to 1.5. The [0.5, 2.0] clip prevents single-outlier runaway.

References

  1. [1] Frick B. (2007). “The football players' labor market: Empirical evidence from the major European leagues” Scottish Journal of Political Economy.
  2. [2] He M., Cachucho R., Knobbe A. (2015). “Football player's performance and market value” Machine Learning and Data Mining for Sports Analytics.

Related Articles

Footnote Metrics

Squad ROI Explained — Football Manager-style Decision Support for J Club Reinforcement

3 min read

Footnote Metrics

The Business Value of roi_decisions Archive — Reinforcement Decisions as Organizational Memory

3 min read

Footnote Metrics

PVS (Player Value Score) — Complete Guide to Footnote's Composite Player Metric

7 min read

Footnote Metrics

PVS vs Other Metrics — Transfermarkt, Wyscout, WhoScored Compared

3 min read

Footnote Metrics

fit_score Explained — Club Philosophy × Player Evaluation Vector Inner Product

2 min read

Track Your Growth with Footnote

Just record your matches — AI analyzes every 5 games. Visualize growth with PVS Score. All features free during beta.

30-second signup · No credit card required

← Back to Guide

Last updated: 2026-05-18 ・ Footnote Editorial