[ejlangev]

> Some analysis is better than no analysis.

I don't agree with that. Doing some analysis on a limited and possibly skewed data set can lull you into a false sense of understanding. It makes your ideas seems objective when they can in fact be completely baseless.

[alexanderdmitri]

How is no analysis just as good?

It's hard to believe that the percentage of successes with no analysis is GTE the percentage of successes with some analysis.

[ACow_Adonis]

I presume because faulty analysis or bad data can actually get you more off track than when its just a "back of the napkin" based guess or hunch.

And additionally, you now have the pride/confidence thing in your even worse results because you did "analysis"...

[jacobush]

Often the pride is constant.

[jamesrcole]

As always, you need to take the limitations of the available information into account. That ought to be part of the analysis, though of course it often, or perhaps usually, isn't.

[08-15]

Not at all. Analysis of limited data gives you wide credible intervals, the exact thing that guards against unwarranted confidence.

[bpchaps]

Are you saying that zero middle ground exists?

[Dylan16807]

Obviously a middle ground exists, but it's possible that in part of the scale you get worse results as you head toward the middle.

[ejlangev]

I didn't downvote you :(

[bpchaps]

Sorry - edited. Still didn't answer the question, though :p