The Future is Now: New and Developing Gaming Technologies: Case Study: Twitch plays Pokemon

Twitch Plays Pokémon (TPP) was an exciting era in gaming, especially for game designers who were curious about how such a project could work. For those who are unfamiliar, Twitch Pays Pokémon is a social-experiment-slash-game played live on Twitch, where actions and decisions are crowdsourced via the chat function.
Dan Shefelman lived through that moment and remembers all the fascinating ways the game evolved around its clear obstacles.
“I thought we’d seen it all. We’ve seen multiplayer games, we did platform games, we had FPS, we’ve just seen every genre possible. And then, Twitch Plays Pokémon comes around.”
The most glaring obstacle was how thousands of commands given by the chat could be distilled into one cohesive decision so the game could progress.
“[I asked if] it needs to take one input and the character moves. And they said no, all these people are putting in that input and they’re – somehow – getting averaged. Well, you know, at one point, then a voting system came out.”
The technology also had to answer the question of lag. If a user gave an input from South Korea, and another user gave an input from Brazil, which would arrive first? Would the first one count before the second? These questions were slowly rising, and communities developed out of the solutions people found. Eventually, they developed a voting system, until another problem arose.
“The voting system went away and it was just this incredible phenomenon of how we deal with, basically, overpopulation . . . Maybe it’s almost like a democracy deciding on a policy. And they did.”
This happened outside of the game, and players returned ready to navigate it as a team, like a crowd-sourced game.
“We as a team, a world team, we can beat that by coming together as communities to plan and strategize, to play one game,” says Shefelman. “Like a million people played one game together with one character. And that was basically a new genre.”