Hacker News new | ask | show | jobs
Show HN: How far do I have to go to run into 100k people? (imjasonh.github.io)
63 points by ImJasonH 5 days ago
28 comments

That seems like the distance to go from having 100k people in front of you to having 100k people behind you. The distance you'd have to go to run into them properly would depend critically on the width of the vehicle you're in. With about 30k people running the Boston marathon and 60k people running the NY marathon, if you can source a vehicle wide enough to cover the course, it seems like you could pull it off in just a bit over 52.4 miles... assuming that the bodies didn't produce enough drag to slow you down so much that runners were managing to stay ahead of you on the course. (And assuming that marathon runners prioritize completion of the course over running sideways away from obvious danger, but they're marathon runners, that seems like a safe assumption.)
> That seems like the distance to go from having 100k people in front of you to having 100k people behind you

0 metres. Just turn round.

To pass within say 5m of someone you just need to set up shop at a busy train platform and wait. With 1000 people passing every 2 minutes it won’t take long to pass 100k (well presumably about 3 hours)

Ah interesting, I seen the title "run into" as "bump into" or in this context, get to where 100k people are. I don't know any US addresses far from population centers so I pumped in Area 51!

Clarity needed on the title or description!

It says right at the top of the website:

> From the pin, a filled 5° slice in every direction. Petal length = distance to 100k people.

What does “distance to 100k people mean”?
You could replace with bump into and the comment would still apply though, no?
I must be slightly confused about what claims this visualization is making. Are these 5 degree rays supposed to be indicating non-overlapping triangles where each triangle contains 100k population? If so, a full circle (with no triangles missing) would be a population of 7.2 million. And yet, clicking around Los Angeles County you can easily find multiple completely non-overlapping full circles, despite the county population being under 10 million.
Yeah, this can't possibly be right. Los Angeles, the city, has less than 4 million people, and there are complete "circles" of presumably 7.2 million people entirely within its boundaries.

If you set the petal population threshold to a low value, like 10k or 20k, and drag it to a densely populated area like Koreatown, you'll get a degenerate "circle" with 0 feet in every direction.

Is this resident population or ambient?

I've been building a UK-wide 24-hour population dataset lately, and it's really opened my eyes to how much population data is best-guess-estimate - including census figures - and quite how difficult it is to say with any certainly how many people "touched" a location on any given day.

Cities on the water make fun shapes for obvious reasons. Atlantic City, Seattle, Chicago, Galveston, New Orleans. The slices can go very far or even be omitted due to not hitting the contiguous US. Chicago's is probably the most interesting, they cross Lake Michigan before getting sufficient density to stop.
It's rough not having Canada's population in here. There's way more than 100k people in all directions north from Buffalo and it doesn't feel accurate.
Neat! I've got a slice that's under 10 miles long going south into town, but I've also got 14 slices to the north (the ones that don't intersect with a couple up-state cities) that are over 3000 miles long.

Vancouver is similarly interesting, a lot of great-circle 2000+ mile segments that reach over towards New England, but everything west or north is invalid.

Would be neat if instead of a single yellow petal to 100k, each petal was a color gradient to 1k, 2k, 10k, 20k, 100k, 200k, or 1M people!

I live in the one of the densest parts of lower Manhattan, and often explain that when I take a walk around the neighborhood, I pass 100,000+ people.

This map pretty much proves it!

Even if you are colliding with 20 people a minute thats still 3.5 straight days of running into people. Even if gentle thats probably heavy bruising, and any conflict or discourse is gonna slow down that 20-people rate.

I imagine theres probably only a few cities where you could pull this off.

Anyways I know the article isnt this literal, this was just my first thought from the title and I had a chuckle lol.

heh, from my address in downtown Chicago, I get 2.34 miles in the shortest direction and > 3000 miles (aka don't reach 100k within the US) in the longest direction (slices of lake michigan passing through unpopulated stretches of the UP. Canada is not included, but not sure it would help much...
I got 1419 miles by putting the pin just east of Presidio, TX. The Gulf of Mexico doesn't have many Americans in it on the way out to the Florida Keys.

1,570 miles by pinning the lower Keweenaw Peninsula which null-counts a great circle route through Canada.

Having been to Presidio on a trip to Big Bend I'm not surprised
> Pin stays in the contiguous US — Alaska, Hawaii, and abroad aren’t covered.

It's kinda funny to hear an LLM use such americentric language, I suppose it didn't "know" it was working on something that would be public.

Just a note: using the noir plugin for safari i had to disable it as the map is dark and invisible. Maybe its a layer thing.

This is really neat. Could this data be linked to quality of life indexes and see if there’s any correlation?

Pretty cool, took me a few minutes to figure it out, there was a nice stretch of Florida that had no rays, clear cut wilderness
Doesn't work outside the US sadly.
It does say that it only works in contiguous US (so no Alaska nor Hawaii nor "abroad" :D ), but seems it also works in Bahamas, for some reason, which as far as I know, isn't part of the contiguous US (yet?).
Indeed, it doesn't seem to think that the fairly large cities of Vancouver, Calgary, Edmonton, Regina, Winnipeg, and Thunder Bay have a population of approximately zero. Anywhere in northern Montana has the "beam" shoot all the way over to Maine before reaching 100K, which isn't quite right since it passes through a couple medium Canadian cities and dozens of prairie towns.

I think this tool would be much more interesting in the western US if they had incorporated Canadian census data as well.

This is very cool - I really want to see it in other places outside of the US - I imagine the centre of Australia only has like 4 potential lines. Do you plan on expanding it beyond the US?
I don't fully understand what this is supposed to visualize, but at least for my location it tells me I'd have to run considerably far into golden gate park, which I imagine is negligible in terms of people.
Is this supposed to mean that 100k people live in each slice of the pie? Counting up slices and comparing to my metro statistical area population, that about works out.
Havre, MT has curved rays going east. Is that intentional?
That is a cool visualization.

Punched in my grandparents' town and a bunch of the slices stretched 1500km over three states away. A lot of the US is big and empty.

I was hoping this would estimate the mean free path from population density. Still fun though.
Fun! If you click in the center of Nebraska, US - there are some angles for which you will never find 100k people.
This is a killer app for Vancouver drivers.
Really cool. Makes me wonder where the most remote point in the contiguous U.S. is by this metric.
I'm having a hard time finding any place that produces a ray longer than about 900 miles. I found a spot in NW South Dakota whose shortest rays are nearly 300 miles.
The longest single ray I've found is 1089 miles around El Paso, headed northwest. There ain't much out there as long as you dodge Flagstaff and southwestern Utah.
Try a little NE of Seattle.
Would be cool to be able to change the width/angle of the petal
The SF Bay Area is suddenly not so dense when you realize how much is water.

From Union Square, you'd need to scan the city plus the entire East Bay.

This is cool and useful. I would love to see the drive time version.
Waiting for rest of the world
Global population density estimates are available at the ~100 m^2 level from the LandScan project at Oak Ridge National Labs.

https://landscan.ornl.gov/about

nicely done.