Belleville News-Democrat: Why are there fewer Pokemon locations in black neighborhoods?
Within days of the launch of “Pokémon Go,” the new augmented reality game that uses phone cameras and GPS to show the classic ’90s ‘pocket monsters’ in real environments, a number of sites and museums started complaining that they were not appropriate places to catch ’em all. Formerly public sites turned private were also reported as being marked as “pokéstops” or “gyms.”
Some areas had the opposite problem: no pokéstops at all. Some social media users observed that their small town and rural neighborhoods lacked any pokéstops, though urban areas had plenty. {…}
In some areas where most residents are minorities, Twitter users noted pokéstops are hard to find. Kendra James, a writer at Racialicious, noted that the mostly-black city of Irvington, New Jersey has “no stops on all of its main roads, despite… several monuments/locations of note,” compared to Maplewood, a nearby majority-white town which she perceived to have many more.
Read more here: http://www.bnd.com/news/nation-world/national/article89562297.html#storylink=cpy
From: Why are there fewer Pokemon locations in black neighborhoods? | Belleville News-Democrat
#BlackPokemonMatterReport
#BlackPokemonLivesMatterReport
I couldn’t get to the article (I’m assuming it’s hammered) so forgive me if I’m saying something made explicit in the article.
I’m guessing that part of the template for Pokemon Go is the original Niantic game of Ingress. I’m guessing that the places that, in Ingress, were the biggest places for people to engage in combat or what have you were used as places for Pokemon Go.
And a very particular group of people have been playing Ingress for a while and Pokemon Go appeals to freakin’ everybody and not merely the people who enjoyed Ingress.Report
@jaybird — That’s my understanding. It’s based on the Ingress data, and thus Ingress-rich areas are well stocked, and Ingress-poor areas not so much.
That said, they sucks and they should fix it. The problem is, this fucker exploded fast, so I doubt their devs are ready for the kind of rapid change they need to properly “capture this moment of the zeitgeist.”
(They released an update last night that seems to have fixed the “I caught a Pokemon and now the game just kinda froze so I have to restart but then I lose my Pokemon” bug, which was needless to say irritating as fuck.)
I hope they manage. It’s a fun game. Obviously we want black folks to be able to share in the fun equally.
My neighborhood is okay. It is a kinda mixed-race, mixed-class, half-gentrified cuz the gays moved in place. Anyhow, I always get off at the subway stop a mile from my apartment. (There is a closer one, but I wanna be skinny.) Anyhow, I now walk two further blocks over so I can head down Dot Ave, the “big drag” in Dorchester. Along the way I encounter maybe five Pokestops in a mile. So, that’s not so bad I guess. There are two right by my apartment, plus a “gym” in the little breakfast joint by me. (It’s currently team red. I haven’t picked a team yet, but I don’t think I’ll pick red.)
I don’t know what it’s like over by Blue Hill and stuff, the parts of Dorchester that are black-black-black. I never go there. That said, they should have the same number of Pokestops I got. After all, they’re just bits on a server.Report
This is not an easy problem. Previous discussion here is quite applicable:
https://ordinary-times.com/2016/05/27/how-that-machine-algorithm-ended-up-being-racist/
Given that it appears the Pokestop and gym locations seem to be taken directly from Ingress, I would expect that the locations would be more white/male/20,30 something focused. It is hard for me to think of a data set that exists that would even begin to directly address this.Report
It is hard for me to think of a data set that exists that would even begin to directly address this.
An RNG based on where whomever is carrying a phone happens to be, with a miniscule multiplier for every additional person within 20 yards or so?Report
With a countdown, even, so that you’re guarandamnteed a pokemon every 2 hours, 45 minutes or something.
Whatever the psych department says is the perfect amount of time to maximize endorphin production.Report
Most pokestops are some kind of “landmark” or whatever. They usually have a photo. Humans have to make these.
It’ll take crowd sourcing.Report
Well, they’re gonna have some kind of access to the Google maps data they need, being a former Google sub. Likewise, getting pop. density data cannot be too hard. That stuff exists. (Which, from time to time I talk to the various geo-data folks and, there is a lot of stuff out there, which can be had.) So assume they have that.
(Note: I have no inside info on their access to Google Maps data. It’s just, that seems like something the various “decision makers” would be cool about. Everyone likes this thing, so they’ll want the “public glow.” Unless I’m wrong.)
So what locations?
Seems easy enough: create a website where folks can suggest and vote. Take the top votes over {period}, and then filter by population density — this can be a pretty straight-up heuristic. Google Maps already knows the borders of municipalities and (often) neighborhoods. You can run the algorithm on them one by one, so keep the dimensionality manageable.
Neighborhoods that already have a suffucient-according-to-population number of pokestops can be left alone. Existing pokestops would be maintained, but added to the algorithm as “fixed points.”
The diff-evo algorithm is pretty obvious. That’s what I’d use. (Diff-evo is continuous, but it’s pretty easy to do “discrete” stuff with it.) You score according to votes + evenness of distribution. Google maps also knows what are major-versus-minor streets, etc. That can apply also. It’s good when more pokestops are on the “main drag.”
Anyhow, you can play with the weights. You can pick a few sample neighborhoods, tweak them, and see if they look good.
It doesn’t need to be perfect. This is a problem where “good enough” really is good enough.
The hard part is, doing this fast. People wanna play right now. It’s “hot,” but won’t stay “hot” forever. Likewise, the media is yelling now. It sucks.
Once you have the data, you have the data. Just as this game could ramp up fast on the Ingress foundation, their next “thing” can ramp up fast on this data. Likewise, other companies wanting to make an “augmented reality” game might be willing to buy the data, along with the tools to tweak and customize the data.
It’s a new thing. People like it. They’ll want more. We’ll get good at it, very good, soon enough. (I wonder if they’re hiring. They could use me.)Report
Also, for Ingress they were generally asking people to suggest places with very specific characteristics: historical significance, unique art or architecture, local tourist spots, public libraries, and public places of worship. They disqualified candidates that were schools, hospitals, first responder facilities (police stations, firehouses), or that lacked “safe pedestrian access”. In the Denver area, that list of candidates preferentially avoids certain kinds of places. Not just poor neighborhoods; by at least one of the available maps, the entire SW quarter of the metro area is portal-free.Report
Obvs because Nintendo doesn’t want to watch the black people.
They’re poorer in the main, after all.Report
A problem that will no doubt get worse, since it appears that Niantic is about to roll out the feature that allows retail locations to pay to “attract” Pokémon characters to those locations in order to draw in the players.Report
This is truly the best indication available that we have not reached a post-racial society.Report
Oh man, it’s my home town in the news again! For being racist even! Although the article doesn’t quite cover things correctly – Maplewood is substantially whiter than Irvington, but the white population is not a “majority” in an appreciable way. They are dominant, for the usual reasons, but not a numerical majority. Or if they are, it’s under 60%. Not that that excuses anything, of course, the stops are probably all in the white parts of town (which there totally are).Report
I just looked up the census data: 55% white, 35% black, about 5% Latino, 5% everyone else. About what I expected to see.Report
Well today I encountered a new level of overwrought Pokémon paranoia: http://www.rawstory.com/2016/07/end-times-pastor-warns-isis-could-use-pokemon-go-to-target-christians-with-cyber-demons/
Sign me up for the demon wars!Report
Living life seems to be a daily terror with these sort of mindset.Report
Natch.
Pokemon is racist, the developers are racist. The internet is racist. Cell phones are racist.
(I have black friends)Report