#error is a combination of the effects of your instrument’s accuracy precision and variance

[paydirt49]

At this point you’d probably start tracking your precision as an error term. So you’d have something like 1ft±0.01ft or whatever your precision is. Then when you do the conversion you also convert the error term. I’m this case 0.3048m±0.003048m.

Although now I’m realizing you said outside of highly technical contexts, which is what this sort of error tracking is for. I just got excited about engineering.

outside of highly technical contexts, when converting between units there's a subtle risk of introducing "false precision". like, there's this old tumblr post saying like 500 miles sounds "more poetic" or something than 804.672 kilometers.

and it is true that 500 miles precisely equals 804.672 kilometers, but that conversion has introduced a level of precision that wasn't there in the original measurement! it would be less precise but more helpful to say 800 kilometers instead.

anyway I've been thinking a lot about the general problem of how exactly to decide what level of precision is appropriate for a conversion (assuming you're not in a context where you definitely always want the exact conversion). in particular, given one measurement given in feet, what metric unit should be used? I reckon that decimeters would work best (smaller than a foot by a large enough factor to give an accurate-enough conversion, but not so much smaller that they introduce too much false precision), but the problem is that decimeters aren't nearly "normal enough" of a unit to have the same vibes to metric users as things measured in feet do for imperial/us customary users.

so like, do you use centimeters rounded to the nearest ten, or meters with one digit after the decimal point? when exactly is one of these options better than the other? much to consider! I'd be interested in hearing other people's thoughts on this :)