Skip to content
DraconiticResearch
Search Sources API
194,226 studies31 sourcesUpdated 59 min ago
PreprintCC BY 4.0arXiv

Proper Calibeating

  • Dean P. Foster
  • Sergiu Hart
Open at sourcePDFDOICite

Abstract

The classic concept of "calibrated forecasts" and its more recent refinement, "calibeating," are defined with respect to the standard quadratic scoring rule. We extend these notions to the class of proper scoring rules (for which the true distribution is an optimal forecast) and define \textit{proper calibration} and \textit{proper calibeating} by requiring the corresponding guarantees to hold uniformly over all bounded proper scoring rules. We first establish that calibration always implies proper calibration, whereas calibeating need not imply proper calibeating. Second, we show how to guarantee proper calibeating and proper multicalibeating; in particular, \textit{complete calibeating}---a strong form of calibeating that calibeats the joint binning---is always proper. Finally, we consider \textit{decision-making under uncertainty}, where one best replies to the forecasts. We establish that (proper) calibration is equivalent to universal no regret, and (proper) complete calibeating is equivalent to universal no regret together with universal subsuming of the reference sequence, where \textit{universal} refers to all bounded utility functions.

Related

  1. Refined proper descent for left completed sheaves on locally compact Hausdorff spaces
    Li He
    arXivPreprintCC BY 4.0Full textAlgebra & number theoryGeometry & topologyarXiv
  2. Proper scoring rules for survival times
    Philip Dawid
    arXivPreprintCC BY-NC-SA 4.0Full textStatisticsarXiv
  3. Finiteness of open substacks with proper moduli spaces
    Xucheng Zhang
    arXivPreprintCC BY 4.0Full textAlgebra & number theoryarXiv
  4. A proper arithmetic surface with infinite étale fundamental group
    Pablo Chenal
    arXivPreprintarXiv licenceFull textAlgebra & number theoryarXiv
  5. Fundamental groups of proper algebraic stacks are finitely presented
    Chirantan Chowdhury, Ludvig Modin
    arXivPreprintCC BY 4.0Full textAlgebra & number theoryarXiv

Cite

BibTeXRIS
BibTeX
@misc{foster2026proper,
  title = {{Proper Calibeating}},
  author = {Dean P. Foster and Sergiu Hart},
  howpublished = {arXiv},
  year = {2026},
  month = sep,
  doi = {10.48550/arxiv.2605.26703},
  eprint = {2605.26703},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2605.26703},
  note = {CC BY 4.0},
}
RIS
TY  - UNPB
TI  - Proper Calibeating
AU  - Dean P. Foster
AU  - Sergiu Hart
PY  - 2026
DA  - 2026/09/12
PB  - arXiv
DO  - 10.48550/arxiv.2605.26703
UR  - https://arxiv.org/abs/2605.26703
AB  - The classic concept of "calibrated forecasts" and its more recent refinement, "calibeating," are defined with respect to the standard quadratic scoring rule. We extend these notions to the class of proper scoring rules (for which the true distribution is an optimal forecast) and define \textit{proper calibration} and \textit{proper calibeating} by requiring the corresponding guarantees to hold uniformly over all bounded proper scoring rules. We first establish that calibration always implies proper calibration, whereas calibeating need not imply proper calibeating. Second, we show how to guarantee proper calibeating and proper multicalibeating; in particular, \textit{complete calibeating}---a strong form of calibeating that calibeats the joint binning---is always proper. Finally, we consider \textit{decision-making under uncertainty}, where one best replies to the forecasts. We establish that (proper) calibration is equivalent to universal no regret, and (proper) complete calibeating is equivalent to universal no regret together with universal subsuming of the reference sequence, where \textit{universal} refers to all bounded utility functions.
ER  -