Author does informative critique of 538. I'm not sure at all that it has anything to do with Silver vs Taleb twitter war though. It's more like premise for expr

Author does informative critique of 538. I'm not sure at all that it has anything to do with Silver vs Taleb twitter war though. It's more like premise for expressing author's own thoughts. However I totally dig what 538 does and appreciate their analysis. I understand that there is a complex underlying model with many parameters taken subjectively. The results of Monte Carlo runs of this model are very interesting to me. The fact that they describe distribution of outcomes and not just a single number is important property of a Monte Carlo model and I would have an issue if that wouldn't provide that info. If I would like to analyze complex event that has lots of moving parts and uncertainties I would also build a model to see what kind of distribution I would get. The fact that somebody published results of their model is quite useful. The analogy would be European and American weather models - no one say that their results are exact and everybody understand that uncertainty in result comes from inherent uncertainty of initial state as well as shortcuts and approximations each of the model takes. No one says that the results of these models are useless because of that. And everybody finds it valuable if weather forecast predicts rain tomorrow with 40% chance. So I don't see how political prediction is only valuable (by the words of the author) if it comes without probabilities attached to it.