Julia Galef on the Sunk Costs Fallacy (Big Think)

youtube.com

3 points·shankarro··4 comments

This makes perfect sense. What doesn't make sense at all is when coworkers use the sunk cost fallacy as reasoning for the following situation: your team has been working on a rewrite of the codebase for months. It's looking pretty close to being done, when the boss says, "why are we still building this?" The PM who was pushing it has been fired, and the guy who started the thing has left the company. The team still would benefit from having this rewrite done, but because the last few deadlines have been missed (due to the firings/leaving), it is declared a "sunk cost" and to be abandoned.

This makes no logical sense, because a project you needed, that you still need, is 90% done. There is no sunk cost fallacy here. "Sunk cost" is overused and overrated.

The problem is the old saying "The first 90% takes 90% of the time. The last 10% also takes 90% of the time."

That is, it's 90% done, or seems to be, but how far is it from being actually done? Just 10%? History suggests that's far too optimistic. So if it's maybe-10%-but-probably-more-by-an-unknown-amount from being done, is it better to abandon it as a sunk cost? Depends on how far away it actually is - and nobody knows the actual truth of that.

I agree, estimations are hard. I just don't see how this is the same issue as that mentioned in the video at all. To me, "sunk cost" is a buzzword used to justify the cancellation of any project that takes a long time to develop.

The sunk cost fallacy applied to my latest project. It was perfect for me. But not for anyone else. Didn't stop me from finishing it though. It would have been better if I had given up on it.

Reply on news.ycombinator.com