Gotta say, using 3 just feels like giving up due to laziness, even in 1200BC.
Also it’s interesting how the Chinese entries basically stop between 1400 and 1949, whereas European names are far more present during this era. Some Japanese ones, too. I wonder how comprehensive this page is.
Rounding pi to 3 is just the engineering way. It’s close enough to get the job done and then I don’t have to worry about decimal places. However, using pi=3 typically undershoots your calculations, so personally I like to use pi=4
AFAIK the Chinese knew that the value between that of the encompassing shape that meets the circle at tangeants to the inscribed shape whose edges meet the same equidistant points gives us the approximation of pi. So did archimedes, and maybe even the babylonians.
So while a triangle yields about 3 and satisfies the theorem, you could also theoretically draw a 96 gon and 192 gon like Liu Hui for an accuracy of 9x10^5.
Personally I just memorize 22/7 or use the Leibniz infinite series if I have to.
Gotta say, using 3 just feels like giving up due to laziness, even in 1200BC.
Also it’s interesting how the Chinese entries basically stop between 1400 and 1949, whereas European names are far more present during this era. Some Japanese ones, too. I wonder how comprehensive this page is.
Rounding pi to 3 is just the engineering way. It’s close enough to get the job done and then I don’t have to worry about decimal places. However, using pi=3 typically undershoots your calculations, so personally I like to use pi=4
An error margin of less than 5% (even better, biased in a known direction) is more than good enough for plenty of use cases.
An error margin of more than 25% on the other hand, is seldom acceptable.
Nah, it’s fine. Trust me I use pi=4 in every calculation I do that uses pi and I haven’t ever run into any issues at all
(I’m not that type of engineer, I never do anything with pi)
It’s called safety factor
AFAIK the Chinese knew that the value between that of the encompassing shape that meets the circle at tangeants to the inscribed shape whose edges meet the same equidistant points gives us the approximation of pi. So did archimedes, and maybe even the babylonians.
So while a triangle yields about 3 and satisfies the theorem, you could also theoretically draw a 96 gon and 192 gon like Liu Hui for an accuracy of 9x10^5.
Personally I just memorize 22/7 or use the Leibniz infinite series if I have to.
Sometimes zero decimals is enough precision even in 2025…
…but also because of laziness…