Well, I think the chat is set in slow mode, so you can only send a command once every 30 seconds.
I tried to implement a voting algorithm on the Dark Souls stream I did today, but I didn't have enough people for it to work. The idea was to have up to 10 actions queued at any given time, and requiring each action to have been voted on N number of times before it is queued. Eventually I just removed the voting and let people control directly, since it was going to slow with the 50-100 or so people we had.
The system already ignores most of the commands due to how the game works. Essentially, it picks a move at random every frames or so. And if you look at the live stats of the voting, you can see that the people tend to vote in a single direction at a time already: https://sheltered-reef-4687.herokuapp.com/
I think that the long stream delay is a bigger problem then the randomness. People tend to vote according to what the stream is currently showing but the votes will actually be counted 20 to 40 seconds in the future.
This would change it completely - rather to use the spam chat, just aggregate the data from the last performed command, maybe half to 1 second delay, then poll the commands for 1s or something and perform the highest voted action.
It'd be nice to have a little green/red indicator when commands are being accepted.
I setup something similar to this today, but for Dark Souls instead. (Pokemon to Dark Souls -- that escalated quickly.)
It was actually pretty fun, but the Twitch delay made it difficult. We managed to kill a bunch of enemies and slowly make our way up various staircases, but the cliffs always got the best of us.
That reminds me I still have not finished Dark Souls, thanks!
That game is so hard sometimes, and I love it! It's more fun when you play with others, because if you are alone it's easy to get frustrated at not being able to do something or not knowing where to go.
There should be some vods on the channel, I think...but it's not that fun to watch, since it's about 5 hours with long stretches of nothing happening when people got bored. We had about 50-100 people in chat at most times.
I may do it again in the future, but I'm considering looking into a streaming service with a shorter stream delay.
I submitted a link to the twitch yesterdy. One of the comments links to a google doc with current progress recorded. There's also been a program made in C# that analyzes the stream to draw a map of Red's movements. All in all, it is a pretty cool community project right now!
Why is the assumption I didn't read the article?
What's even more curious is that not only do you start with the assumption that I did not read the article, you also go ahead and question yourself wether or not you fully understood me.
Which is funny, because you STARTED with questioning if I had read the article.
Regardless - you get all the starters in Yellow + Pikachu.
And assuming things aren't fully random, I wanted to see what the hive mind wanted to go with.
He made that assumption under the (incorrect) belief that you didn't fully understand that it was only a single game of pokemon being played.
He had this incorrect belief because he inferred from your statement, "it would be cool to see which starting Pokemon got used most", that you thought multiple games were being played.
He made this inferral because in the Pokemon games that he has played(recall he hasn't played yellow), you cannot get all of the starting Pokemon in a single game.
His own anecdotal experience resulted in a differing interpretation of your statement than was intended, and I would be remiss to not mention that fact that I 'mis'-interpreted your original comment in exactly the same manner as he did.
What a retarded and unfortunate happenstance this has been.
Not to mention the 50000 people watching the stream, an extremely active chat, and 20+seconds of latency; how exactly do you think that any qualitative findings in a post-mortem of this social experiment would be in any way indicative of the communities preferred starting Pokemon.
Furthermore, are they really "starting Pokemon" in yellow if you only really 'start' with Pikachu?
I actually was blackout drunk for the first time in awhile and woke up to a bunch of extremely in depth comments about trivial stuff I had made here and on Reddit. Riddled with the standard grammatical errors.
Wish I had done something more productive really, come on drunk me.
But that's not how it works, thousands of people just enter button presses so which pokemon gets chosen won't be any indication of what the majority of those players wanted to choose
