Andrew Gelman has a good post summarizing a lot of his research on the basically predictable nature of US Presidential election outcomes. In a lot of respects, this is the greatest story never told about American politics. After all, who could tell you but the mass media? And presidential elections are big business for the political press, giving them a strong incentive not to believe — and not to talk about — the strong evidence that these things are pretty predictable. I think I heard more cable news coverage of the idiotic Redskins rule than I did of the totally serious Hibbs forecasting model.