Show HN: Are You a Good Estimator? | Hacker News Reader

Show HN: Are You a Good Estimator?

estimator.dylancastillo.co

15 pointsdcastm9 comments

Took it (didn't go too well score-wise).

Would be nice to have the units for each question on top. Or in the input boxes themselves. I was on mobile and had to scroll down to see it. Mixing raw numbers and units of billions was confusing as well. There were definitely questions like the one about volume where I had no idea what a cubic mile really was conceptually (as opposed to a liter or something).

Regardless, reading about the study and comparing my own results was interesting.

I wonder if the inputs could be done using a slider given you are always selecting a range, that would be more intuitive.

Makes sense! I like the slider idea, but not sure if it’d introduce some bias to the results.

One quick improvement would be to style the number boxes to include decimal separators. For a few answers I was squinting to check if I had typed six or seven zeroes, for example.

I definitely messed up some units, particularly a million/billion. Two were really close to my estimates but still outside. One I could have gotten it right if I thinked a bit more.

4/10 overall, good reminder that we always have to double check. So easy to be wrong even when giving outrageously wide estimates

It would be nice to how wrong I was, as in, off by 1 order of magnitude or 5 orders of magnitude. Because I was super close on some of them.

6/10 by doing 1-1000000

Fun test. 4/10

I never realized how absolutely massive a blue whale can be!

Total volume of the Great Lakes

Your estimate:

10,000 to 25,000 cubic miles

Correct answer:

22,684 cubic miles

Uhhhh... pretty sure I was right and it marked it wrong.

---

Also, I think that marking answers right or wrong is the incorrect way to measure this. By deeming it right or wrong, you are making it trivia rather than estimation.

For instance, I estimated that Alexander the Great was born between 300 BCE and 100 BCE, it was marked wrong but the right answer is 356 BCE. That is a fairly accurate estimation still.

You should get rid of the Buzzfeed-esque "number people got right" and move towards measuring mean percent off from the correct answer so you could say like "Most people are within 30% of the right answer" or something like that. You should also make it where you can't game it by doing something like 0 to 100,000,000,000,000,000 and being told you are right: https://i.imgur.com/Ysnz1gr.png ;)

If I say no and I’m wrong does that mean I’m right?

Reply on news.ycombinator.com