Indeed. The overconfidence that certain people have re: their approaches is silly. The best one can do is review the data and various methods, tinker for themselves, carefully monitor the results and make adjustments accordingly. Crazy that so few people do the n=1 experiment on the most relevant test subject - themselves!