Herewith, I am sharing an insightful article (link given below) by Prof. Amartya Lahiri, discussing why the exit polls failed to gauge the nation's mood accurately in 2024.
The article emphasizes that relying on predictions based on the conclusions of competitors and peers can often lead to incorrect outcomes. The article suggests a strategy that somehow mirrors Stackelberg's game. A game in game theory is any interaction between multiple people in which each person’s payoff is affected by the decisions of others. In a Stackelberg game, there is one player (say “leader”) who moves first, and all other players (the “followers”) move after him. However, such strategies should not be adopted while doing forecasting.
The forecasts, whether they are related to business, economics, or any other field, should be based on proven and tested models that are independent of the forecasts of other forecasters. However, it can be challenging because customers/users/audience of a forecaster may be influenced by forecasts released by well-known market leaders, leading them to cast doubts about the accuracy of the forecast if it differs from the market leader's forecast. To address this issue, the forecaster needs to focus on gaining the confidence of the customer rather than trying to outplay other forecasters.