Da Cinco Model — EPL Strategic Decision Framework

Problem

English Premier League clubs face an asymmetric risk environment: small performance gains deliver modest rewards, while relegation results in severe financial losses (£100M+). Decision-makers need a framework that quantifies how players, ticket pricing, and strategic interventions influence both on-field performance and off-field financial risk.

Approach

Player Performance & Availability: Built a position-specific metric (PP90) using z-scores across attacking, midfield, defensive, and goalkeeper stats. Extended with injury risk via logistic regression (age, minutes, recurrent injury) and Poisson modeling for expected games missed. Player value = performance × availability.

Ticket Pricing: Demand index aggregates opponent attractiveness, timing, market size. Log-linear demand with capacity constraint. Optimal prices solve a season-level revenue maximization with game-specific demand shifts.

Revenue Model: Matchday (utilization × yield), broadcast (league-position merit), commercial (AR process). OLS-estimated coefficients from financial and attendance data.

Dynamic Decisions: At each decision point, evaluate actions (transfers, manager change, injury insurance) by comparing marginal cost to expected relegation loss reduction: L_R × ΔP_rel ≥ Cost(action).

Data & Scale

• Crystal Palace 2024–25 squad (FBref, Understat)
• EPL financial and attendance data (Kinnaird, European Football Statistics)
• Applied to 2026–27 season planning

Results

• Identified squad strengths (defense/midfield > attack) and key injury risks (e.g. Chris Richards)
• Optimal ticket pricing: ~£45 base for representative demand (validated vs. actual £48)
• Transfer and managerial-change recommendations justified only when relegation probability exceeds threshold; injury insurance (loans) recommended for high-impact players

Tech Stack

Python, NumPy, statistical modeling (logistic regression, OLS), optimization.

Links

• View PDF