Awesome game Kappa

It's an interesting idea to develop a game specifically for streamers where viewing audiences can actively participate and influence the course of the game. I don't know any other game that did that before. As always, the most important aspect is probably the right balance. The game needs to be fun to play by itself, the people in chat needs to feel like they actually have an impact on what's happening and at the same time they should not be able to block progress completely and ruin the fun.

From a academic perspective, it is amazing how you attempt to harvest entropy, in effect create an random number generator, from twitch chat. Of course the quality of this entropy will vary for different streamers. Just think of forsen, I think it might be safe to assume, that his chat will have on of the highest if not the highest entropy on twitch. I think one of the hardest problemes you will face, might be how you can turn this source of chaos into something that can be percieved as a response to what the chat wrote. The chat is a player, too then. How to make it fun for the chat? Is it only to make the life of the streamer harder? All in all, I think you have a great idea for a fun experience for streamers and their chat!

What is your favorite aspect of the proposed game? Why?
The asynchronous nature in the sense that from the perspective of the streamer it plays like a normal game with typical orderly inputs (one decision, one outcome) while from the perspective of the chat it is a way to interact with the streamer by using the chat to input commands in a rather chaotic fashion (many decisions, still one outcome).

What is your least favorite aspect? Why?
The fact that twitch chat does not have a clearly defined goal and/or win condition. At this point it is unclear to me whether the game mode is "streamer vs. twitch chat" or "streamer + twitch chat vs. game (highscore)" or something completely else.

Which single change or addition would you suggest to most improve the game?
I would add clear game modes like the ones mentioned above so that it is clear what the chat is supposed to do and also how the streamer is going to interact and react to the poll results. Furthermore another incentive for the chat to actually participate in the game seems to be a good idea, like a win condition. The reward could be something like channel points (if thats possible with the twitch API), which fits the target audience well.

1. Including Twitch chat is awsome!
2. The amount of needed chat input might be difficult to balance. Imagine different viewer amounts (10 to 10 000). Maybe less is more. (Even Marbels on Stream is a great game with basicly no input.) Another challange could be individual stream delays so I think the gameplay has to be quite slow for every user to follow.
3. Maybe split the game in two phases, the first ist setup: done by twitch chat. The second is gameplay by the streamer. Those phases could also alternate during the game.

I dig the fact that you made this game twitch chat compatible

I think my least favorite aspect is that you require a twitch chat to play with.

Which leads me to my suggestion that you could have a random inputter for the twitch commands so you can play without any actual random user interaction

I like the idea of incorporating the twitch chat, as it is unique in this course. 

One problem might be to find enough people to play the game for it to be fun, as it heavily depends on viewers/chat input. Another problem might be the blurry goal. 

I feel like if you could somehow overcome those, it might be a fun game to play as a streamer. 

What is your favorite aspect of the proposed game? Why?
It's an interesting idea to play a game via twitch. I'm very curious how your game will be like in the end!

What is your least favorite aspect? Why?
I'm not sure how good the game can be balanced and how easy it will be to give a score for the streamers performance.

Which single change or addition would you suggest to most improve the game?
Maybe give the streamer the ability to lock single tiles for a limited time to create a little more order.

To me the idea of using Twitch chat to generate chaos is equal parts hilarious and genius. 

My problem would be that the player/streamer is at the complete mercy of their chat without much to anything about it.

Personally I could imagine mechanics that allow the player to deal with the fiesta better, such as being able to choose a specific ingredient or tile once. I could also see a mechanic in which the streamer can influence the voting procedure by for example freezing the votes for a few seconds while the clock is still counting down. These mechanics could for instance run on a cooldown.

Either way I am really looking forward to how your game will turn out!

What do I like the most?
I think that the idea of basing the gameplay on Twitch Chat more or less entirely is a pretty cool and very rarely seen concept.

What do I like the least?
Obviously the game will not be playable (at least not as intended) without a twitch chat, making it more or less impossible to play unless you are a somewhat known streamer.

What change would I suggest?
If there is enough time after reaching the planned targets, I'd suggest either integration of other live streaming platforms (e.g. YouTube) or something like a singleplayer mode with either procedural or hardcoded interferences.
This could also serve as a demo version of the final game.

I think this could be a fun game for streamers to play while they wait for their next match-up in a different game and help them keep the chat engaged.

It could be difficult to keep the game fair for the streamer. I would expect twitch chat to always pick the worst possible option for the streamer they can find, just because it is funny.

Like: Using the twitch chat as an input medium
Fear: Balancing the game. Especially with different chat sizes.
Suggest: Giving the chat 'goals' so they can quantifiably win instead of just achieving chaos and hindering the streamer.

I love the idea of making a game with a Twitch integration as a project and I'm excited to try it out.
Maybe try to make the visuals appealing enough for viewers to stay in the stream and participate over longer periods of time.
Also, from experience in watching streams of similar games, I think that not that many people actually want to see the streamer lose. So maybe don't expect every chatter to try to make the streamer suffer, as other people here in the comments expect them to. Of course I might be wrong, though.

1. As a regular Twitch user, I really love your game idea and am really excited how the final game will look like!
2. Due to Twitch chats reaching from very few to a lot of twitch viewers the balancing of the game might be quite hard.
3. Maybe you could use the current number of viewers in the chat to balance the game. This might also be important so the streamer also has a fair chance as Twitch viewers often want to see the streamer suffer.

Favourite aspects: The idea of leveraging Twitch is both great and hilarious, and is a smart decision for a game to keep up with current trends. The game idea feels like it has drawn inspiration from Overcooked as well as the Twitch Plays Pokemon experminet, both of which were great successes in their own right. I'm thus really looking forward to what happens when you combine both of these.

Least Favourite aspect: I understand that this game is being designed specifically for a Twitch audience, but that also means that its applicability is quite limited. The gameplay experience could vary greatly based on the number of simultaneous viewers. Furthermore, it would be hard to port this to any other platform.

Suggestion: Maybe incorporate an AI behind the scenes that can help simulate Twitch viewers in required cases? This could be programmed with different social characteristics (sadistic AI, cooperative AI etc.) such that it helps 'smooth the curve' for the player even at different viewer counts. This AI could then also be leveraged in the game even if Twitch was not connected. That way your game would be a lot more robust and portable.

Overall, I look forward to playing as a viewer and making you lose!

The Twitch chat integration is imo both the strength and the curse of this game. I understand that your vision is to have a twitch chat only game which is a super cool and new idea, however, realistically speaking, most of us will not have the viewers to really experience this game in all its potential. Imagine having no viewers and also not being able to play the game because you have no viewers. Feels bad man. And could turn out really frustrating. Consider that in you design process,

Favorite aspect: Twitch integration. Integrating your viewers into the game is always fun and can lead to more fun for everyone.

Least favorite aspect: Twitch integration. A game that revolves arround streaming it can only really be used by streamers, which greatly limits the possiblity to play it.

Single change: Give clear objective to viewers. Causing mayhem is fun, but having a goal leads to greater teamwork possiblities.

What is your favorite aspect of the proposed game? Why?
I like that you're trying something very different in terms of technical aspects by making a Twitch API game.

What is your least favorite aspect? Why?
I'm not a big fan of Twitch chat culture, so I honestly can't estimate how much fun this will be simply because I can't imagine myself having fun by spamming Pepega. Of course this doesn't have to mean anything for your game, but be aware that it's a small niche. You need a big streamer with a community, which has fun playing something like this. I'm curious how you'll conduct the playtesting.

Which single change or addition would you suggest to most improve the game?
Making singleplayer mode/normal multiplayer mode for a few people would make this game playable by a lot more people and also save you a lot of headaches regarding playtesting. But I also don't know how this would look with the current concept.

Looking forward to what you'll make out of it!

It's safe to say that your game is the most unconventional when it comes to the technical innovation and I mean that in a very good way. I think it's a great idea to exploit the Twitch API for your game and I also love how the whole concept of your game is based on a pun. Since Twitch chat is already quite chaotic in most cases, I am certain that your technical achievement will complement the theme so well. I would only suggest adding some constraints for the viewers so that it's actually possible to realize order during the gameplay. Although the idea of pure chaos is very tempting, you would probably want to make it possible for the player to eventually achieve order and win the game. I am sure that your game would be hella fun to play and I'm really excited to see your final demo!

Here's one anonymous comment:

Incredibly creative take on the theme, I absolutely love it and cant
wait to play it....as a viewer especially. Maybe you could also develop
a twitch chat simulator so this game will be playble alone aswell.
Something along simulated chat inputs would be needed for the debugging
anyway and is something to incorporate even more twitch chat energy