Conditional distribution analyses of probabilistic forecasts

J. Frank Yates, Shawn P. Curley

Research output: Contribution to journalArticle

41 Citations (Scopus)

Abstract

Probabilistic forecasts have good ‘external correspondence’ if events that are assigned probabilities close to 1 tend to occur frequently, whereas those assigned probabilities near 0 tend to occur rarely. This paper describes simple procedures for analysing external correspondence into meaningful components that might guide efforts to understand and improve forecasting performance. The procedures focus on differences between the judgements made by the forecaster when the target event occurs, as compared to when it does not. The illustrations involve a professional oddsmaker's predictions of baseball game outcomes, meteorologists' precipitation forecasts and physicians' diagnoses of pneumonia. The illustrations demonstrate the ability of the procedures to highlight important forecasting tendencies that are sometimes more difficult to discern by other means.

Original languageEnglish (US)
Pages (from-to)61-73
Number of pages13
JournalJournal of Forecasting
Volume4
Issue number1
DOIs
StatePublished - Jan 1 1985

Fingerprint

Conditional Distribution
Forecast
Forecasting
Correspondence
Tend
Game
Target
Prediction
Demonstrate
Conditional distribution

Keywords

  • External correspondence
  • Probabilistic forecasts
  • Scoring rule
  • Subjective probability

Cite this

Conditional distribution analyses of probabilistic forecasts. / Yates, J. Frank; Curley, Shawn P.

In: Journal of Forecasting, Vol. 4, No. 1, 01.01.1985, p. 61-73.

Research output: Contribution to journalArticle

@article{638e5734d0584857bbdf21680c000ef9,
title = "Conditional distribution analyses of probabilistic forecasts",
abstract = "Probabilistic forecasts have good ‘external correspondence’ if events that are assigned probabilities close to 1 tend to occur frequently, whereas those assigned probabilities near 0 tend to occur rarely. This paper describes simple procedures for analysing external correspondence into meaningful components that might guide efforts to understand and improve forecasting performance. The procedures focus on differences between the judgements made by the forecaster when the target event occurs, as compared to when it does not. The illustrations involve a professional oddsmaker's predictions of baseball game outcomes, meteorologists' precipitation forecasts and physicians' diagnoses of pneumonia. The illustrations demonstrate the ability of the procedures to highlight important forecasting tendencies that are sometimes more difficult to discern by other means.",
keywords = "External correspondence, Probabilistic forecasts, Scoring rule, Subjective probability",
author = "Yates, {J. Frank} and Curley, {Shawn P.}",
year = "1985",
month = "1",
day = "1",
doi = "10.1002/for.3980040110",
language = "English (US)",
volume = "4",
pages = "61--73",
journal = "Journal of Forecasting",
issn = "0277-6693",
publisher = "John Wiley and Sons Ltd",
number = "1",

}

TY - JOUR

T1 - Conditional distribution analyses of probabilistic forecasts

AU - Yates, J. Frank

AU - Curley, Shawn P.

PY - 1985/1/1

Y1 - 1985/1/1

N2 - Probabilistic forecasts have good ‘external correspondence’ if events that are assigned probabilities close to 1 tend to occur frequently, whereas those assigned probabilities near 0 tend to occur rarely. This paper describes simple procedures for analysing external correspondence into meaningful components that might guide efforts to understand and improve forecasting performance. The procedures focus on differences between the judgements made by the forecaster when the target event occurs, as compared to when it does not. The illustrations involve a professional oddsmaker's predictions of baseball game outcomes, meteorologists' precipitation forecasts and physicians' diagnoses of pneumonia. The illustrations demonstrate the ability of the procedures to highlight important forecasting tendencies that are sometimes more difficult to discern by other means.

AB - Probabilistic forecasts have good ‘external correspondence’ if events that are assigned probabilities close to 1 tend to occur frequently, whereas those assigned probabilities near 0 tend to occur rarely. This paper describes simple procedures for analysing external correspondence into meaningful components that might guide efforts to understand and improve forecasting performance. The procedures focus on differences between the judgements made by the forecaster when the target event occurs, as compared to when it does not. The illustrations involve a professional oddsmaker's predictions of baseball game outcomes, meteorologists' precipitation forecasts and physicians' diagnoses of pneumonia. The illustrations demonstrate the ability of the procedures to highlight important forecasting tendencies that are sometimes more difficult to discern by other means.

KW - External correspondence

KW - Probabilistic forecasts

KW - Scoring rule

KW - Subjective probability

UR - http://www.scopus.com/inward/record.url?scp=84984488684&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84984488684&partnerID=8YFLogxK

U2 - 10.1002/for.3980040110

DO - 10.1002/for.3980040110

M3 - Article

VL - 4

SP - 61

EP - 73

JO - Journal of Forecasting

JF - Journal of Forecasting

SN - 0277-6693

IS - 1

ER -