> I have a strong suspicion that if we could run the counterfactual, 29% would be right on the money. Conditioned on what? How would you model that? Go back an

> I have a strong suspicion that if we could run the counterfactual, 29% would be right on the money. Conditioned on what? How would you model that? Go back and change the weather? Redo polls?