The impression I got was that they told the major carriers what they were planning and the carriers all said, "oh, we can handle that traffic". The carriers have all done music festivals and such, where there are huge crowds, but the difference is, those don't involve every single person there hammering on the network every single second (presumably they're mostly texting and sending occasional selfies). I suspect the carriers had some combination of not thinking through the amount of traffic, and not caring too much (setting up temporary towers costs a them lot of money, and if people can't get the app to work, they'll blame Niantic, not the carriers).
Niantic should have a) discussed the event with the carriers, ahead of time, more forcefully, including negotiating to get multiple temporary towers from each major carrier at the event, even if it cost them money (in retrospect, it would have been a trade of up-front money for good PR vs bad PR for the game); and/or, b) they should have hired a highly skilled company, months in advance, to flood the park with really good, high-capacity WiFi for the event. Again, would have cost them money up-front, but would have been worth every penny.
They've essentially pulled in their most dedicated fans and supporters from around the world, to a highly visible and much anticipated event, and then given those fans a poor experience. Which means, when they go home and talk to friends or post on their YouTube channels, they'll be less enthusiastic about it than they would have been had things gone better. "Hey, I've got an idea, let's make all of our very best word-of-mouth advertisers unhappy and frustrated with our product".
I've been to an all-day Niantic event before, an Ingress Anomaly, and they did quite a good job of it. There were "merely" several thousand players there (vs the 15K at the PoGo fest), and because of the nature of the game, it was spread over many square miles of ground, in several parts of the city. Network/cell access wasn't really a problem even when there were several hundred players gathered in a tight crowd with the two teams furiously (electronically) battling for control of vital Portals (all PokeStops/Gyms are a subset of Ingress Portals).
In this arrangement (15K players in an enclosed park) the network problems were easily foreseeable (they knew the area, they knew how many tickets they sold, and they could talk to the carriers about coverage). When your entire highly-visible event depends on high network capacity and server capacity, these are details you really need to sweat ahead of time - just as, if you put on a music festival, you make
very sure the bands show up. I wish they had handled things better (from the coverage I saw, some people were upset, some had fun, most were a mix of the two, but if they had made the event an astounding success, the positive coverage might have gotten the game a bit of media spotlight - the game is still far from perfect, but considerably improved over last summer when everyone went nuts for a few months).
[doublepost=1500930703][/doublepost]
One year later and the game is still a bugfest. And I doubt it will ever be in a stable state given that Pokémon Go's popularity has died down significantly, to the point where Niantic will be wasting resources perfecting it.
"Died down significantly" in the sense that there's "only" tens of millions of people playing, instead of hundreds of millions (and many of those remaining are more dedicated players more willing to put money into the game). Some non-trivial percentage of the earth's population downloaded the game last summer - dropping off from that is still huge. A lot of game companies would kill to get that kind of traffic. I think your statement pretty severely minimizes the actual situation.
