5 ms·
Has anyone checked out the actual math behind this? The hypothesis sounds reasonable enough (though so do a lot of others), but something like >But they’ve suc
by ImprobableTruth 6y ago
Has anyone checked out the actual math behind this? The hypothesis sounds reasonable enough (though so do a lot of others), but something like
>But they’ve succeeded in impressing writers for nontechnical publications
isn't enough for me to give it any more credence than any other "this is why society is breaking apart" explanation.
- dogma1138 6y agoThe math might be right but it doesn’t make the theory or assumptions correct. GIGO is always an issue, especially in field where it’s very hard to make unbiased measurements of anything. We have the same thing in economics our models that explain things in hindsight are actually quite good but their predictive power is often at the level of reading tea leaves and chicken bones.
- arbitrage 6y agoThere are a few hints in the article that might suggest Turchin is pre-cultivating his data. It smells like p-hacking, honestly.
- titanomachy 6y agoGIGO == garbage in, garbage out
- rossdavidh 6y agoThe best summary would be in "Ages of Discord", where he goes into more detail on which datasets he is using, and what the equations are that he is using that data to fit coefficients for. The fact that it has the occasional equation in it, probably limits its sales, but I found it an interesting read.