How Wise is the Crowd in Prediction Markets

5.June 2026

If you’ve ever scrolled through Polymarket or Kalshi wondering whether the “wisdom of crowds” is actually wisdom—or just organized noise—you’re not alone. A new paper, “How Wise is the Crowd? Bias and Edge in Prediction Markets,” tears into the microstructure of modern prediction markets to ask a practical question: Who’s actually making money, and who’s just paying for the privilege of being loud? By engineering a high-frequency data pipeline that ingests tick-level order flow, on-chain wallet histories, and social commentary across decentralized finance and regulated venues, the authors expose structural inefficiencies that most traders overlook. The verdict? Market efficiency in Web3 betting isn’t dead—but it’s wearing a very clever disguise.

First, let’s bury a zombie: the Favorite-Longshot Bias. Classic behavioral finance literature tells us that longshots are overpriced and favorites underpriced. But when the authors applied continuous, multivariate spline regressions—controlling for contract lifecycle timing and liquidity regimes—that neat S-curve evaporated in most markets. For example, in “Mention Markets” (e.g., “Will Trump say X on Fox News?”), what looked like a favorite-longshot distortion was actually a pervasive “Yes Bias”: traders systematically overpay for the affirmative outcome, driven by narrative conviction and temporal volatility spikes near resolution. This isn’t just an academic nuance; it’s a market microstructure signal. If you’re fading the late-stage “YES” frenzy in low-liquidity narrative contracts, you’re not fighting the crowd—you’re harvesting its optimism premium.

Second, forget the whale mythology. In traditional finance, size often signals sophistication. Not here. By reconstructing wallet-level P&L in real-time, the paper shows that whales—the most capitalized participants—systematically bleed expected value to small-order traders via adverse selection. These large actors aren’t sharp; they’re ideological, overpaying for narrative exposure while nimble retail traders pick off their stale limits. Even more telling: using a Natural Language Inference framework to score trader “vocality,” the authors find zero correlation between sentiment intensity and edge. The loudest voices in Polymarket chats aren’t alpha generators—they’re noise. For practitioners, the takeaway is clear: denoise your signals by adjusting for temporal lifecycle, liquidity regime, and the ideological bias of oversized positions. In the wild west of Kalshi and on-chain prediction markets, edge isn’t about being first or loudest—it’s about being structurally aware.

Authors: Deleep, Avaneesh and Lee, John and Bai, Jenny and Suresh, Dhruv and Dhawan, Harsh

Title: How Wise is the Crowd? Bias and Edge in Prediction Markets

Link: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=6322678

Abstract:

Prediction markets are increasingly relied upon as real-time probability oracles, yet their predictive signals remain polluted by structural inefficiencies. While prior literature documents anomalies like the favorite-longshot bias at an aggregate level, the microstructural origins of these distortions—specifically, who generates and exploits them—remain unstudied in modern ecosystems. To investigate this, we engineer a scalable, multi-threaded data architecture capable of synchronously ingesting and persisting tick-level order flow, decentralized wallet histories, and user commentary across Polymarket and Kalshi.

As ever, we present several interesting figures and tables:

Notable quotations from the academic research paper:

“Prediction markets are increasingly relied upon as real-time probability oracles, yet their predictive signals remain polluted by structural inefficiencies. While prior literature documents anomalies like the favorite-longshot bias at an aggregate level, the microstructural origins of these distortions—specifically, who generates and exploits them—remain unstudied in modern ecosystems.

The Apparent Illusion of the Favorite-Longshot Bias: Our preliminary univariate analysis suggested the presence of a classic favorite-longshot bias in Mention Markets. However, by deploying multivariate penalized splines to control for the confounding effects of contract lifecycle timing, we reveal this to be a statistical artifact. Instead, these markets exhibit a pervasive “Yes Bias,” whereby participants consistently overpay for the “Yes” outcome — that is, the contract that pays off if the reference event occurs — relative to the corresponding “No” outcome.

Furthermore, we find that “Whales”, or the most capitalized players, are not the most sophisticated. By dynamically reconstructing participant positions, we demonstrate that Whales, on average, systematically bleed expected value to small-order traders. Rather than acting as sharp informed players, these large actors likely trade on ideological conviction, structurally overpaying for specific narratives and suffering from adverse selection against smaller participants.

We find no statistically significant correlation between sentiment intensity and informational edge, indicating that the most prominent voices in these ecosystems function primarily as sources of communicative noise.

Ultimately, this research provides a reproducible methodology for actively denoising prediction market probabilities by adjusting for temporal lifecycles, liquidity environments, and the ideological biases held by heavily capitalized, unsophisticated traders.”

Are you looking for more strategies to read about? Sign up for our newsletter or visit our Blog or Screener.

Do you want to learn more about Quantpedia Premium service? Check how Quantpedia works, our mission and Premium pricing offer.

Do you want to learn more about Quantpedia Pro service? Check its description, watch videos, review reporting capabilities and visit our pricing offer.

Do you want algorithmic access to the full Quantpedia database via the API? Subscribe to Quantpedia Pro, ask for an API key, and explore the in/out-of-sample statistics, source academic papers, and code snippets — ideal for quantitative research, systematic trading workflows, and AI model training.

Are you looking for historical data or backtesting platforms? Check our list of Algo Trading Discounts.

Or follow us on:

Facebook Group, Facebook Page, Twitter, Linkedin, Medium or Youtube

Share on Refer to a friend

We’ve already analysed tens of thousands of financial research papers and identified more than 700 attractive trading systems together with hundreds of related academic papers.

Browse Strategies