/sci/ - Science & Math

[Anonymous No.2575824]

Hey /sci/. I don't usually come here, but I found this trollscience pic, and I don't understand how this DOESN'T work. I know it doesn't because it's a trollscience pic, but it really seems like it should work.

[Anonymous No.2575841]

bumping with cool fish

[Scientist No.2575843]

>>2575824
Because perimeter isn't defined that way. You can talk about taxicab metrics and so on, but the short version is that's simply not its definition.

You need the approximating shape to be tangent everywhere, or at least approaching tangent everywhere, or something like that. In the OP pic, you get arbitrarily close in distance, but still just as far from tangent as when you start. That's how you define perimeter.

[Anonymous No.2575840]

lines are not curves

[Anonymous No.2575853]

Because if it did work, then since an independent calculation puts <span class="math">\pi < \frac{7}{2}[/spoiler], that identity would cause a horrible breakdown in other axioms we like.

For an actual understanding, it's just that the limit of the lengths of a sequence of curves is not the length of the "limit" of those curves, because the "limit" of those spikey block curves is not going to be nice and smooth like the circle. In a grand waving of the hand. Swish, swish.

[Anonymous No.2575905]

Get measuring tape, wrap it around a square object with a parameter of four inches. Now take that same measuring tape and wrap it around a cylindrical object with a diameter of 1 inch.

Does it go 4 inches around? No, and you're a shit head.

All troll sciences take advantage of something that looks logical but isn't. As it approaches 'cutting the corners' you lose what is initially nonexistent amount of surface area, but as you get closer to the circle, it has to bend around it. This gradually increasing amount of error will take you from 4 to ~3.14

[Anonymous No.2575929]

lol thats great!

[Anonymous No.2575943]

The line in a Koch snowflake has an infinite length... just sayin. it might help you see.

[Anonymous No.2575949]

I think its because by removing same length corners, you would get a octahedral shape, not a circle shape.
Also, I trolled my math teacher with this, whole class discussion about it.

[Anonymous No.2575959]

The length of the curves in a series does not necessarily equal the length of the limit curve of that series.

[Anonymous No.2575962]

>>2575943
This. Each iteration of the Koch snowflake has a different perimeter, and the limit (the actual Koch snowflake) has infinite perimeter.

[Anonymous No.2575972]

The length of the limit curve does not in general equal the limit of the curve length.

[Anonymous No.2575983]

>>2575972
>>2575959
These

[Anonymous No.2576010]

ummmm...because circles don't have straight edges?

[Anonymous No.2576023]

I post this every time I see that picture.
http://www.askamathematician.com/?p=4937

[Anonymous No.2576029]

>>2576010
Circle=infinite sided polygon

[Anonymous No.2577227]

This reminds me of surfaces with lots of nucleation sites, like mentos. Things that appear superficially "smooth" from a distance, but if you look at them with a microscope (or if you show 'em some diet coke) it becomes apparent how much surface area there really is.

That's what's going on here, OP. It looks like a circle, but it's really all pixels, which has a higher perimeter than a circle. How much higher?

Why, 4/pi, of course.

[Anonymous No.2578758]

Glad to burst your bubble