Paper: arXiv 2312.16179

Authors: Albert Cohen, Jimmy Risk

Abstract

This paper presents a new framework for player valuation in European football, by fusing principles from financial mathematics and network theory. The valuation model leverages a “passing matrix” to encapsulate player interactions on the field, utilizing centrality measures to quantify individual influence. Unlike traditional approaches, such as regressing on past performance-salary data, this model focuses on in-game performance as a player’s contributions evolve over time. Consequently, our model provides a dynamic and individualized framework for ascertaining a player’s fair market value. The methodology is empirically validated through a case study in European football, employing real-world match and financial data. This cross-disciplinary mechanism for player valuation adapts the effect of connecting pay with performance, first seen in Scully (1974), to include in-game contributions as well as expected present valuation of stochastic variables.

Complexity vs Empirical Score

  • Math Complexity: 8.0/10
  • Empirical Rigor: 6.5/10
  • Quadrant: Holy Grail — high math complexity, high empirical rigor

Why this score: The paper employs advanced stochastic calculus and network theory (Markov chains, GBM, PDEs), fitting high math complexity, while it includes a detailed empirical case study with real-world EPL data, indicating strong implementation and backtest potential but less than full trading system rigor.

Research Flowchart

  flowchart TD
  A["Research Goal: Dynamic Player Valuation"] --> B["Data Inputs"]
  B --> C["Model Construction"]
  C --> D["Computational Process"]
  D --> E["Valuation Outcomes"]

  subgraph B ["Data Inputs"]
      B1["Match Event Data"]
      B2["Financial Data"]
  end

  subgraph C ["Model Construction"]
      C1["Passing Matrix"]
      C2["Network Centrality"]
  end

  subgraph D ["Computational Process"]
      D1["Stochastic Variable Integration"]
      D2["Financial Mathematics Fusion"]
  end

  subgraph E ["Key Findings"]
      E1["Dynamic Fair Market Value"]
      E2["Individualized Player Worth"]
  end

Evidence Card

What the authors claim, on what data, and who has checked it. Read from the paper's own text by a language model: every quote is verbatim, nothing was re-run. How to read this card

Proposes

The paper proposes a new player valuation framework that combines financial mathematics with network theory using a passing matrix and centrality measures.

Source sentence
This paper presents a new framework for player valuation in European football, by fusing principles from financial mathematics and network theory.

Assumes 6

  • The Opta Index of player performance follows a Geometric Brownian Motion. model
    Source sentence
    which is assumed to follow a Geometric Brownian Motion (GBM) [27], a common assumption in contingent claims modeling.
  • Passing matrices are row‑normalized to form Markov chains. model
    Source sentence
    By row-normalizing, these matrices become Markov chains, allowing relevant theory to be used.
  • The proportion of a player’s index relative to the sum of all players’ indices can be modeled directly. model
    Source sentence
    We directly model the proportion of a player’s index relative to the sum of all players’ indices.
  • The stochastic process πt is bounded on (0,1) and follows a Pearson diffusion with a stationary Beta distribution. model
    Source sentence
    πt is guaranteed to be bounded on (0, 1) and πt has stationary (long-term) distribution Beta with known parameters
Show 2 more
  • Future salary (payment) structure is known at the valuation time. execution
    Source sentence
    assume the payment structure is known in the future so that at+h ∈Ft.
  • Row‑normalization yields empirical probabilities of passes, treating them as transition probabilities. statistical
    Source sentence
    provides the proportion of passes from player i to player j, or equivalently the empirical probability of that transition.

Data

  • Opta Index (player performance) · Opta alternative commercial
    Source sentence
    They use the Opta Index [29] as an index of player performance, which is assumed to follow a Geometric Brownian Motion
  • Per‑game player share data alternative unknown
    Liverpool, Arsenal, Brighton players
    Source sentence
    Data is collected per game1 and player shares are calibrated using the estimation methods in Appendix D.1.
  • Team revenue and player index sum data macro unknown
    Source sentence
    Team revenue and the sum of player indices to determine revenue per index point are also modeled by GBM’s

Costs

Not applicable The paper reports no trading or execution result.

Finds 5

  • The valuation framework accurately captures contributions of players across different salary levels, aligning compensation with performance. performance
    Source sentence
    Our empirical analysis indicates that the valuation framework accurately captures the contributions of players across different salary levels, providing a balanced assessment that aligns compensation with performance.
  • Results sometimes show over‑ or under‑compensation, but many cases suggest projected underpayment aligns with a pay raise. performance
    Source sentence
    While our results sometimes indicate over- or under-compensation, we find many settings where our projected underpayment aligns with a pay raise.
  • Virgil van Dijk shows close alignment between his compensation and performance throughout the evaluated seasons. performance
    Source sentence
    Players like Virgil van Dijk exhibit a close alignment between compensation and performance.
  • The model improves upon the TCV approach by directly modeling the performance proportion rather than separate GBM numerators and denominators. theoretical
    Source sentence
    We directly model the proportion of a player’s index relative to the sum of all players’ indices.
  • Augmented passing matrices provide closed‑form probabilities of player involvement in scoring possessions. theoretical
    Source sentence
    This probability can be derived in closed form using traditional Markov chain techniques

Checked by others

No other paper in the archive has been found to support or contradict these findings (compared with 12 related papers). Coverage is partial, so this is neither replication nor refutation.

Code

No repository by the authors found in the paper.

Extracted 2026-11-14 · same card as JSON · A language model chose these items: check the source sentence before relying on one.