Conditional Forecasts and Proper Scoring Rules for Reliable and Accurate Performative Predictions

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Boeken, Philip, Zoeter, Onno, Mooij, Joris M.
Formato: Preprint
Publicado: 2025
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866911229670326272
author Boeken, Philip
Zoeter, Onno
Mooij, Joris M.
author_facet Boeken, Philip
Zoeter, Onno
Mooij, Joris M.
contents Performative predictions are forecasts which influence the outcomes they aim to predict, undermining the existence of correct forecasts and standard methods of elicitation and estimation. We show that conditioning forecasts on covariates that separate them from the outcome renders the target distribution forecast-invariant, guaranteeing well-posedness of the forecasting problem. However, even under this condition, classical proper scoring rules fail to elicit correct forecasts. We prove a general impossibility result and identify two solutions: (i) in decision-theoretic settings, elicitation of correct and incentive-compatible forecasts is possible if forecasts are separating; (ii) scoring with unbiased estimates of the divergence between the forecast and the induced distribution of the target variable yields correct forecasts. Applying these insights to parameter estimation, conditional forecasts and proper scoring rules enable performatively stable estimation of performatively correct parameters, resolving the issues raised by Perdomo et al. (2020). Our results expose fundamental limits of classical forecast evaluation and offer new tools for reliable and accurate forecasting in performative settings.
format Preprint
id arxiv_https___arxiv_org_abs_2510_21335
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Conditional Forecasts and Proper Scoring Rules for Reliable and Accurate Performative Predictions
Boeken, Philip
Zoeter, Onno
Mooij, Joris M.
Statistics Theory
Machine Learning
Performative predictions are forecasts which influence the outcomes they aim to predict, undermining the existence of correct forecasts and standard methods of elicitation and estimation. We show that conditioning forecasts on covariates that separate them from the outcome renders the target distribution forecast-invariant, guaranteeing well-posedness of the forecasting problem. However, even under this condition, classical proper scoring rules fail to elicit correct forecasts. We prove a general impossibility result and identify two solutions: (i) in decision-theoretic settings, elicitation of correct and incentive-compatible forecasts is possible if forecasts are separating; (ii) scoring with unbiased estimates of the divergence between the forecast and the induced distribution of the target variable yields correct forecasts. Applying these insights to parameter estimation, conditional forecasts and proper scoring rules enable performatively stable estimation of performatively correct parameters, resolving the issues raised by Perdomo et al. (2020). Our results expose fundamental limits of classical forecast evaluation and offer new tools for reliable and accurate forecasting in performative settings.
title Conditional Forecasts and Proper Scoring Rules for Reliable and Accurate Performative Predictions
topic Statistics Theory
Machine Learning
url https://arxiv.org/abs/2510.21335