I won’t include here the complete history of Twitch Plays Pokemon, which you can grab from the news. I am writing this at 5 Days, 20 Hours and 40 minutes into the “social experiment,” which many people say has “failed” with the introduction of Democracy Mode.
The experiment has not “failed.” The experiment has collected a trove of fascinating player experiences and made an adjustment in the face of a plateau to revitalize it and collect more fascinating player experiences of a slightly different sort. In artificial intelligence, plateauing refers to an AI getting stuck in a situation where it can no longer meaningfully improve itself. The players got stuck in the same small maze for more than a day straight due to an inability to further improve their co-ordination with the tools provided. At this point, Democracy Mode was introduced (and the original mode named Anarchy Mode) and a mechanism for toggling between them was installed some small time afterwards. The maze was then solved.
This has upset many fans of the game because it “ruined the point” or otherwise seemed to them to have fundamentally damaged the experience. The suggestion put forth by some of them was to simply let the game remain stuck indefinitely, until players quit by the tens of thousands so that a small number of hardcore true fans could co-ordinate their way through the maze. Ostensibly, the thousands of ejected players would then be expected to return.
I hope these people do not apply the same philosophy to everything. They are clinging to absolute ideological purity in the face of undeniable evidence it currently isn’t working. The experiment has not “failed” because a new mode of operating was established. The experiment has shown that Anarchy Mode worked well for a small number of players, was somewhat tenable for a few tens of thousands of players, and become deadlocked when there were many tens of thousands of players faced with an intricate task. When a vote-based system was introduced, it showed that Democracy Mode could sustain the game through the massive playerbase growth.
The game is now toggling between Democracy Mode and Anarchy Mode on a fairly regular basis as people tip the meter through mass voting and occasional use of the “start9” deadlock protest. Democracy Mode is more effective for intricate footwork and Anarchy Mode for battles. Both are equally ineffective at taking care of things like getting out of a dungeon to go heal and getting back, because the exact plan is not necessarily evident to all players, or agreed upon - whereas in puzzles and battles, the goal is completely unambiguous. In puzzles, in particular, the only “dissenting” option is that of the troll, attempting to destroy progress for the sake of ruining others’ fun.
Imagine you founded a town with a few hundred people, and agreed upon some laws with some novel features. All went very well as your town expanded to a few thousand people. At about twenty thousand, however, serious congestion and sanitation issues started becoming evident, and at fifty thousand, it had practically devolved into a war zone. Some relatively minor changes to the laws would clean up the situation substantially. Do you a) make minor changes to the laws or b) tell everyone that 90% of them need to leave the town they moved to because they saw something desirable about it?
Since implementing Democracy Mode, the max simultaneous views has increased by about fifteen thousand, so I don’t think the changes have “ruined the fun” for most people. It allowed the game to remain functionally a game without having to resort to kicking most of the players out. This is not a “failed experiment,” this is an absolutely awesome ongoing experiment which is allowing more and more people to participate.
Thousands of people controlling one Gameboy have co-ordinated sufficiently well, in the face of hordes of trolls, to get four of eight badges and counting. We’re going to Victory Road.