Information Cascade, Herding
The Information Cascade
The Information Cascade is an Information Cascade and Herding scenario. The core lesson: When everyone follows the crowd, the crowd can be confidently wrong. You're choosing between two restaurants. You have a private tip that Restaurant A is better. DecisionPlay maps the players, payoffs, and equilibrium dynamics that shape how this situation typically resolves.
The situation
You're choosing between two restaurants. You have a private tip that Restaurant A is better. But you see a long line outside Restaurant B and nobody at A. Do you follow your tip or follow the crowd?
Background
Information cascades (Banerjee 1992, Bikhchandani et al. 1992) occur when people rationally ignore their own private information and follow the observable choices of others. The logic: if many people chose B, they probably know something. But the cascade can start from just 2-3 early movers, and once started, everyone's private information gets suppressed. Berger, Feldhaus, and Ockenfels (2018) found that shared group identity makes cascades even stronger: when you see people 'like you' choosing B, you're more likely to follow, even when your own evidence points to A.
What this reveals
Cascades amplify noise into confidence
Information cascades make the crowd look knowledgeable when it's actually just imitating. The first few movers set the direction. Everyone after is following, not deciding independently. Shared identity makes this worse, not better, you're more likely to follow 'your people' even when their choices carry no more information than a stranger's.
How to counter it: When you see 'everyone' choosing something, ask how many independent data points actually support the choice versus how many are just following the same cascade. Trace it back to the origin.
A question to sit with
What's a belief or choice in your professional community that 'everyone' agrees on? Have you ever traced it back to see what the original evidence actually was?
Frequently asked questions
- What game theory concept does The Information Cascade illustrate?
- The Information Cascade illustrates Information Cascade, Herding. When everyone follows the crowd, the crowd can be confidently wrong.
- What is the situation in The Information Cascade?
- You're choosing between two restaurants. You have a private tip that Restaurant A is better. But you see a long line outside Restaurant B and nobody at A.
- What does The Information Cascade reveal about how people decide?
- Cascades amplify noise into confidence. Information cascades make the crowd look knowledgeable when it's actually just imitating. The first few movers set the direction. Everyone after is following, not deciding independently. Shared identity makes this worse, not better, you're more likely to follow 'your people' even when their choices carry no more information than a stranger's.
- How do you avoid the trap in The Information Cascade?
- When you see 'everyone' choosing something, ask how many independent data points actually support the choice versus how many are just following the same cascade. Trace it back to the origin.
- What is the research behind The Information Cascade?
- Information cascades (Banerjee 1992, Bikhchandani et al. 1992) occur when people rationally ignore their own private information and follow the observable choices of others. The logic: if many people chose B, they probably know something.
- How long does The Information Cascade take to play?
- About 9 min, at core difficulty, across 4 decision points. It runs in your browser with no account and no sign-in.
Keep exploring
More Classical Game Theory scenarios, or browse all scenarios. New to this? Start with how DecisionPlay works or the game theory glossary.
Topics: cascade, herding, conformity, identity