Skip to content
Longterm Wiki
Back

Credibility Rating

4/5
High(4)

High quality. Established institution or organization with editorial oversight and accountability.

Rating inherited from publication venue: ScienceDirect

Empirical study demonstrating prediction market accuracy compared to traditional polling methods, relevant to AI safety for evaluating forecasting approaches and epistemic techniques for existential risk assessment.

Paper Details

Citations
0

Metadata

journal articleprimary source

Summary

A study comparing prediction markets to polls across five U.S. Presidential elections found that market predictions were closer to the eventual outcome 74% of the time, particularly when forecasting over 100 days in advance.

Key Points

  • Prediction markets were closer to the actual election outcome 74% of the time
  • Markets significantly outperformed polls when forecasting more than 100 days in advance
  • Traders' financial stake creates strong incentives for accurate predictions

Review

This research examines the effectiveness of prediction markets, specifically the Iowa Electronic Markets (IEM), in forecasting election outcomes compared to traditional polling methods. The study analyzed 964 polls across five Presidential elections from 1988 to 2004, demonstrating that prediction markets provide more accurate forecasts, especially at longer time horizons. The methodology's strength lies in its direct comparison of market predictions to poll results, without complex statistical adjustments. The authors argue that prediction markets are superior due to several key factors: traders must invest real money, which incentivizes accurate predictions; the market aggregates diverse information dynamically; and participants are motivated to gather and process information effectively. The research significantly contributes to understanding alternative forecasting methods, suggesting that market-based predictive approaches can be more reliable than conventional polling techniques, particularly when trying to forecast election outcomes months in advance.

Cached Content Preview

HTTP 200Fetched Apr 9, 20265 KB
Prediction market accuracy in the long run - ScienceDirect

 

 
 
 
 

 Sep
 OCT
 Nov
 

 
 

 
 08
 
 

 
 

 2018
 2019
 2020
 

 
 
 

 

 

 
 
success

 
fail

 
 
 
 
 
 
 
 
 
 
 

 

 
 
 
 
 
 
 
 
 

 

 About this capture
 

 

 

 

 

 

 
COLLECTED BY

 

 

 
 Organization: Internet Archive
 

 

 These crawls are part of an effort to archive pages as they are created and archive the pages that they refer to. That way, as the pages that are referenced are changed or taken from the web, a link to the version that was live when the page was written will be preserved. 


Then the Internet Archive hopes that references to these archived pages will be put in place of a link that would be otherwise be broken, or a companion link to allow people to see what was originally intended by a page's authors.


The goal is to fix all broken links on the web. 

Crawls of supported "No More 404" sites.
 

 

 

 
 
Collection: GDELT

 

 

 A daily crawl of more than 200,000 home pages of news sites, including the pages linked from those home pages. Site list provided by The GDELT Project
 

 

 

 

 

 
TIMESTAMPS

 

 

 

 

 

 

The Wayback Machine - http://web.archive.org/web/20191008073512/https://www.sciencedirect.com/science/article/pii/S0169207008000320

 
Skip to main content
Skip to article

 

 
 
Please note that Internet Explorer version 8.x is not supported as of January 1, 2016.
 Please refer to this support page for more information.

 ×
 

 

ScienceDirect

Journals & Books

Create accountSign in

 

Sign inCreate account

Journals & Books

Help

 

Get AccessGet Access

Share

Export

Advanced

JavaScript is disabled on your browser. Please enable JavaScript to use all the features on this page.

International Journal of Forecasting

Volume 24, Issue 2, April–June 2008, Pages 285-300

Prediction market accuracy in the long run

Author links open overlay panelJoyce E.Berga

Forrest D.NelsonbThomas A.Rietzc

Show more

https://doi.org/10.1016/j.ijforecast.2008.03.007Get rights and content

Abstract

“Prediction markets” are designed specifically to forecast events such as elections. Though election prediction markets have been being conducted for almost twenty years, to date nearly all of the evidence on efficiency compares election eve forecasts with final pre-election polls and actual outcomes. Here, we present evidence that prediction markets outperform polls for longer horizons. We gather national polls for the 1988 through 2004 U.S. Presidential elections and ask whether either the poll or a contemporaneous Iowa Electronic Markets vote-share market prediction is closer to the eventual outcome for the two-major-party vote split. We compare market predictions to 964 polls over the five Presidential elections since 1988. The market is closer to the eventual outcome 74% of the time. Further, the market significantly outperforms the polls in every election when forecasting more than 100 days in advance.

Previous art

... (truncated, 5 KB total)
Resource ID: a039c6ec78c7a344 | Stable ID: sid_bir78JYBSf