Name for things that are always underestimated, even if you allow for underestimation

[Princhester]

@control-z

I’ve always heard it as “Take your best estimate, and double it.” For some projects it’s more like triple.

This may work, but not if Hofstadter’s Law applies (as I find all too often it does). What the Law says is that when you do what you just said, your estimate will still not be enough.

For example I find when estimating legal costs I work out my best estimate. Then I double it because I know my estimate is never enough. Then (because I know that my estimates are never enough, even when I apply this method) I add a huge fudge factor. But they are still not enough. And so on. I give up. Next time I’m just going to say “the legal costs of this action will be infinite”.

Beat that, Hofstadter! [Shakes fist]

[Chronos]

Which infinity?

[schindler]

@Princhester

This may work, but not if Hofstadter’s Law applies (as I find all too often it does). What the Law says is that when you do what you just said, your estimate will still not be enough.

For example I find when estimating legal costs I work out my best estimate. Then I double it because I know my estimate is never enough. Then (because I know that my estimates are never enough, even when I apply this method) I add a huge fudge factor. But they are still not enough. And so on. I give up. Next time I’m just going to say “the legal costs of this action will be infinite”.

Beat that, Hofstadter! [Shakes fist]

he beats it in a dialogue called birthday cantatatatatata, in the same book. There are different infinities…

[iwakura43]

@Chronos

I can understand how you might confuse [Hofstadter] with Gardner.

Hofstadter actually had a column in Scientific American which was the successor of Martin Gardner’s column in the same publication. The Gardner column was called Mathematical Games; Hofstadter’s was an anagram: Metamagical Themas.