Strictly Proper Mechanisms with Cooperating Players
This reveals a critical flaw in strictly proper scoring rules for real-world applications where cooperation is possible, such as through web portals, making it an incremental but important finding.
The paper tackles the problem of prediction markets failing to incentivize truthful forecasts when players can cooperate, showing that coalitions can exploit arbitrage opportunities to guarantee higher payouts than truthful reporting.
Prediction markets provide an efficient means to assess uncertain quantities from forecasters. Traditional and competitive strictly proper scoring rules have been shown to incentivize players to provide truthful probabilistic forecasts. However, we show that when those players can cooperate, these mechanisms can instead discourage them from reporting what they really believe. When players with different beliefs are able to cooperate and form a coalition, these mechanisms admit arbitrage and there is a report that will always pay coalition members more than their truthful forecasts. If the coalition were created by an intermediary, such as a web portal, the intermediary would be guaranteed a profit.