A huge amount of them can be summarized as "if your X fits these highly unlikely criteria, a highly specialized Y can give you 1% better results than a long-sta

A huge amount of them can be summarized as "if your X fits these highly unlikely criteria, a highly specialized Y can give you 1% better results than a long-standing general algorithm that already works and you can probably use within 5 minutes". No. Really? shutupandtakemymoney.gif