All together, my friends are an accident

[desh]

Because I'm bored and a nerd, and an Excel master, I decided to calculate the geographic average of everyone on my Friends list right now. (Well, minus the people who I never talk to anymore or whose location I don't know.)

The Hawaii and England/Israel people approximately canceled each other out, as did the British Columbia and Florida folks.

And I end up with an average practically on top of a town called Accident in Western Maryland. Which puts me barely closest to bloomable, who just beats out the Pittsburgh folks, with the DC-area folks probably edging out Baltimore, Richmond, and Hurricane WV for third place.

Comments

[eboniorchid.livejournal.com]

You always seem to find the most intriguing ways to spend your bored-out-of-my-mind time. Very interesting.

[dagoski.livejournal.com]

Man, that seems really appropriate.

[lord-emo.livejournal.com]

Wait how do you calculate averages on the surface of a sphere? Shouldn't the average be somewhere in the interior? Or do you find that point and then use the point directly above it? I think too much.

[burr86.livejournal.com]

I assume Josh just took a weighted average of the latitude/longitude?

[desh]

Yeah, sorry to disappoint. It was just latitude/longitude, with adjustments in sign conventions to make sure that I was never wrapping around "the long way".

Everyone is within about 30 points of latitude anyway, so the rectangular projection of latitude/longitude is about as good as any other projection. But yes, I did want some sort of surface projection.

[gutwoman.livejournal.com]

yeah Accident, maryland!!
I've driven through before, I think. apparently it was plotted wrong when they first plotted it on the map or something, so they called it Accident.
and uh.... how did you do this?

[gutwoman.livejournal.com]

ah never mind... lat/long I get it

[pkzimmer.livejournal.com]

The Pittsburgh folks RULE. Way cooler than any dingy town in Western MD.

[bumonyou.livejournal.com]

now, let me ask you something, and it may just be that this question can be answered by my lack of mather-ma-tatical savvy, but here goes: given the placement of the dateline, wouldn't it be impossible to have an average time that rests near the dateline, even when it should (barring having a "true-er" averaging system than simply taking GMT+'s or GMT-'s and averaging them out)?

Say, for example, that your two sample times are Honolulu (GMT-10) and Auckland (GMT+12). Now, geographically speaking, your average time zone should rest between the short distance between the two, that is, GMT+11 or +12. BUT, just averaging the GMT's would place you...I dont know, somewhere in Europe, whatever.

And, incidentally, Honolulu (GMT-10) and Israel (GMT+2) are exactly 12 hours apart, no matter which way you cut it. So the real question is clearly: have you got a good system to decide which way to average if (God forbid) you only had friends in HI and IS?

[bumonyou.livejournal.com]

oh TWAT! looks like you already answered this question. Sort of?

[desh]

Yeah, I knew that my answer should be right around Philly, within maybe 500 miles or so. So I made sure that everything would "wrap around" in the right direction so they wouldn't go the long way toward Philly. If I had something right on the other site of the date line, I'd call its longitute something like 185 deg W (or -185, as is the convention), rather than 175 deg E (or +175).