pith. sign in

arxiv: 2602.17043 · v2 · submitted 2026-02-19 · 📊 stat.AP

Quantifying the limits of human athletic performance: A Bayesian analysis of elite decathletes

Pith reviewed 2026-05-15 21:13 UTC · model grok-4.3

classification 📊 stat.AP
keywords Bayesian modelingdecathlonathletic performance limitscomposition modeltemporal trendsevent dependencemaximal scores
0
0 comments X

The pith

A Bayesian composition model estimates the maximal possible decathlon score by simulating dependent event performances over time.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper develops a Bayesian composition model to forecast how individual decathletes perform across the ten events. The model accounts for non-linear changes in performance over time and the statistical links between successive events. By simulating from this model, it generates the distribution of the highest achievable total scores. This reveals which types of athlete profiles could realistically come close to that ceiling. Such an estimate gives a data-driven view of the outer bounds of combined athletic ability in one of the most demanding multi-event competitions.

Core claim

The central claim is that a Bayesian composition model, which captures potential non-linear temporal trends in performance and the dependence between an event and all preceding events, enables simulation and evaluation of the distribution of maximal possible decathlon scores, along with identification of decathlete profiles that could approach this limit.

What carries the argument

Bayesian composition model that jointly models the ten decathlon events while incorporating temporal trends and sequential dependencies.

If this is right

  • Simulations from the model produce a distribution of possible maximal scores.
  • Realistic profiles of decathletes who could attain scores near the upper limit can be identified.
  • The approach provides quantitative insight into the upper limits of human athletic potential in the decathlon.
  • Forecasts of individual event performances become available for future time points.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • Similar modeling could be extended to other combined events like the heptathlon.
  • Training programs might use the identified profiles to target specific event combinations for higher totals.
  • Long-term performance trends could inform predictions about when the current limit might be approached.

Load-bearing premise

The Bayesian composition model accurately captures non-linear temporal trends in performance and the statistical dependence between an event and all preceding events.

What would settle it

A new decathlete performance record that exceeds the upper tail of the simulated maximal score distribution would challenge the model's limit estimate.

read the original abstract

Because the decathlon tests many facets of athleticism, including sprinting, throwing, jumping, and endurance, many consider it to be the ultimate test of athletic ability. On this view, estimating the maximal decathlon score and understanding what it would take to achieve that score provides insight into the upper limits of human athletic potential. To this end, we develop a Bayesian composition model for forecasting how individual decathletes perform in each of the 10 decathlon events of time. Besides capturing potential non-linear temporal trends in performance, our model carefully captures the dependence between performance in an event and all preceding events. Using our model, we can simulate and evaluate the distribution of the maximal possible scores and identify profiles of decathletes who could realistically attain scores approaching this limit.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 2 minor

Summary. The manuscript develops a Bayesian composition model for elite decathletes' performances across the 10 events. The model jointly captures non-linear temporal trends and sequential dependence between events. It is then used to simulate the distribution of maximal possible decathlon scores and to identify realistic athlete profiles that could approach those limits.

Significance. If the model is shown to be well-calibrated on real data, the joint modeling approach would provide a statistically grounded way to quantify performance limits that respects event interdependencies, offering an advance over independent-marginal extrapolations. The simulation of feasible high-scoring profiles is a clear methodological strength.

major comments (2)
  1. [Abstract and Modeling Approach] Abstract and model description: no dataset is described (number of athletes, years covered, performance sources, or scoring tables used), no validation metrics (posterior predictive checks, out-of-sample log scores, or calibration plots) are reported, and no baseline comparisons (independent-event maxima or simpler trend models) appear. Without these, the simulated maximal-score distribution cannot be assessed for reliability.
  2. [Bayesian Composition Model] Bayesian composition model: the precise form of the sequential dependence (e.g., conditional distributions, copula, or multivariate structure) and the non-linear trend specification (basis functions, knot placement, or hierarchical priors) are not given. These choices are load-bearing for the claim that the simulated profiles are 'realistic.'
minor comments (2)
  1. [Model Specification] Notation for the composition model (e.g., how the joint density factors across the 10 events) should be written explicitly with equation numbers for reproducibility.
  2. [Results] Figures showing simulated score distributions or example profiles would benefit from error bands or credible intervals to convey uncertainty.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their detailed and constructive comments, which have helped us improve the clarity and rigor of the manuscript. We agree that key details on the data, validation procedures, and model specification were insufficiently described and have revised the paper to address these points directly.

read point-by-point responses
  1. Referee: [Abstract and Modeling Approach] Abstract and model description: no dataset is described (number of athletes, years covered, performance sources, or scoring tables used), no validation metrics (posterior predictive checks, out-of-sample log scores, or calibration plots) are reported, and no baseline comparisons (independent-event maxima or simpler trend models) appear. Without these, the simulated maximal-score distribution cannot be assessed for reliability.

    Authors: We agree that these elements are essential for evaluating the reliability of our results. In the revised manuscript we have added a dedicated Data section describing the dataset (approximately 520 elite male decathletes, 2000–2023, sourced from World Athletics and IAAF records, using the standard 1985 scoring tables). We now report posterior predictive checks, out-of-sample log predictive density scores, and calibration plots for both marginal and joint predictions. We have also included explicit baseline comparisons against independent-event Gaussian process extrapolations and simpler linear trend models, demonstrating that the joint composition model yields more conservative and realistic maximal-score estimates. These additions are placed in the main text and supplementary material. revision: yes

  2. Referee: [Bayesian Composition Model] Bayesian composition model: the precise form of the sequential dependence (e.g., conditional distributions, copula, or multivariate structure) and the non-linear trend specification (basis functions, knot placement, or hierarchical priors) are not given. These choices are load-bearing for the claim that the simulated profiles are 'realistic.'

    Authors: We accept that the model specification required greater precision. The revised Methods section now explicitly states that sequential dependence is captured via a Gaussian copula with event-specific conditional distributions (each event modeled as a function of all preceding events). Non-linear temporal trends are implemented with cubic B-splines (five knots placed at the 10th, 30th, 50th, 70th, and 90th percentiles of the observation times) and hierarchical Gaussian priors on the spline coefficients, with athlete-level random effects. These choices are justified by cross-validation and are now accompanied by the full conditional specification and prior hyperparameter values. revision: yes

Circularity Check

0 steps flagged

No significant circularity in derivation chain

full rationale

The paper fits a Bayesian composition model to observed elite decathlon performance data, explicitly capturing non-linear temporal trends and sequential event dependencies. Maximal scores and realistic athlete profiles are then obtained by simulation from the fitted joint distribution. This is a standard predictive workflow with no reduction of outputs to inputs by construction, no self-definitional loops, and no load-bearing self-citations or ansatzes that collapse the central claim. The derivation remains self-contained against external performance benchmarks.

Axiom & Free-Parameter Ledger

1 free parameters · 1 axioms · 0 invented entities

Abstract-only review provides no explicit list of fitted parameters, background axioms, or new entities. The model implicitly assumes historical data can be extrapolated to performance limits and that event dependencies are stationary.

free parameters (1)
  • model hyperparameters and trend coefficients
    Bayesian model must contain parameters estimated from data; none are named in the abstract.
axioms (1)
  • domain assumption Historical elite decathlon performances are representative of future attainable limits
    The forecasting exercise relies on this extrapolation assumption.

pith-pipeline@v0.9.0 · 5442 in / 1046 out tokens · 27518 ms · 2026-05-15T21:13:18.759760+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.