Ask HN: Review a startup - Grafighters
We are building a game that aims to bring your drawings to life exactly the way you envisioned them. graFighters is an online fighting game for your hand drawn characters. By taking a picture of your drawing with you phone and going through a quick process we can bring your character to life to battle other drawings on the site. The interesting part is that you don't control the character, they take on a life/attitude/stats/abilities all on their own based on our systems analysis of how it was drawn. We are calling this algorithm "Cornelius".
I'm reaching out to you guys for feedback on the concept and the potential to spread the Kickstarter link to people who might be interested. This algorithm has been underway for a year and it is not cheap, so we are looking to raise 20,000. Thanks!
graFighters on Kickstarter - http://kck.st/cnTzcP
Main Site- http://www.graFighters.com
Footage from the Demo - http://bit.ly/9J1wdc
More videos - http://vimeo.com/user1881676/videos
Feedback, criticism, verbal abuse, and donations will all be highly valued.
36 comments
[ 4.5 ms ] story [ 73.9 ms ] threadMain Site- http://www.graFighters.com
Footage from the Demo - http://bit.ly/9J1wdc
More videos - http://vimeo.com/user1881676/videos
Providing a quality product means meeting people's expectations. You say that many people expect to control their character, so let them. You're in awe of your excellent algorithm, but people are not going to be so happy when they don't get your algorithm on the first try.
You could still map different moves to different keystrokes depending on the output from your algorithm and make the result depend on the characteristics .
Last but not least, don't look down on "every other fighting game out there". The only users that will be interested in your game will be people that are deeply in love with every other fighting game. They will expect to be able to use their honed fighting game skills in your game as well, but with the awesome addition of bringing their dream character to life.
Above all, they will not get back to the drawing board when their awesome character is shredded to pieces by a simulation over and over.
In reality you are able to do alot about it. Aside from the actual fight, everything you do controls what will happen. How your draw, who you choose to fight, what arena you fight in, what items or weapons you select etc etc. Simply because you are not pressing buttons at the time of action does not mean all hope is lost and everything will become a random simulation. A la the video, we are working on the balance of the game.
"The only users that will be interested in your game will be people that are deeply in love with every other fighting game"
Disagree, almost all of the people who have sent us sketches are not from the typical gamer crowd. They are drawn to the fact they don't have to control them, it takes the pressure off.
"You say that many people expect to control their character, so let them."
This is a really interesting argument that I go both ways on. History has shown both times where it pays to listen and times it pays to ignore your users requests and show them something new and better that they never even thought of. "Providing a quality product means meeting people's expectations" Or exceeding them :)
Thanks for your feedback though, you bring up some interesting points. We aren't ruling out controlling your characters forever, but for the initial launch we think it will ruin the purity of the game. We are exploring adding more decisive elements without adding live "control" if you know what I mean. Mini games etc.
Also, you won't have a character getting pummeled again and again - but there are some opponents you might just not have any luck beating, and for those fights you want to have a diverse base of characters.
"The only users that will be interested in your game will be people that are deeply in love with every other fighting game." - I don't think this is true at all. The feedback we've been getting range from avid gamers, casual gamers, to even people who don't game, but would love to see their hand-drawn character brought to life. The problem with setting this up as a traditional fighting game is that is then becomes immediately too technical for 95%+ of our audience, which isn't really ideal.
btw: I bet people think of your game that way because you present it as a fighting game. If I see a fighting game, of course my first idea is, that I should interact with my character to win a fight. If I am not interested in that kind of stuff, I will probably go to look for another game. So maybe you can also get something out of positioning your product in another way.
It's like when you give your kid an large expensive toy, and you help them unpack it and show them it, and they then spend the day playing spaceships and forts with the big carboard box it came in. You can't dictate fun, and the cool features you managed to script in the background are great and all, but if you get hung up on them and they block people having fun with it then you killed 100% of it because you got fixated on 30% of it.
I've worked on a lot of video games (indie and commercial), and this is a perfect example of something that is much cooler to you (the programmer) but not to the player.
You 100% need to have a player-controlled vs mode or the whole thing is a waste of time. You can ALSO have the autoplay strategy/design/creation (what does that even mean?) if you want. And maybe it will emerge as the winner.
But don't take my word for it. Spend two days and make a simple versus mode (two players on one keyboard, for example). Sit down five or ten random people from your target audience in front of each one and ask them which one they like better.
I understand your point, but as an avid fighting game enthusiast myself you've got to understand that the game has a broad audience of everyone who wants to upload their hand-drawn characters. From a technical aspect, it's got to be easier to grasp than even say, Super Smash Bros. Brawl, and in the eyes of the fighting game community that's about as non-technical as you can get. Developing a hard-core fighting game is a different task all in it's own, and since the idea of graFighters is to have a move pool of thousands with a character pool of infinity, extrapolating those figures with a traditional fighting game engine and we'd be in for trouble.
graFighters isn't just a fighting game, and that isn't meant to be a sly against fighting games - the truth is, graFighters does a bit of a bunch of different genres; There's fighting, but then there's also strategy, there's also social networking in a sense, and there's also drawing. It's not "a fighting game" or "a strategy game", but saying that it's a fighting game is the easiest way to explain it to the largest audience in the least amount of words.
EDIT: The $5000 pledge on your Kickstarter page says that you get to design a character that defies the visual rules other players are held to. But I didn't see any visual rules on the site, and the "Send us a sketch" button just opens an email. I'm afraid my character may violate some rules. Just let me know. /EDIT
It's a late night in the office, so I sent you a whiteboard sketch of my company's robot mascot, Jango Webbot: http://cdn2.alfajango.com/images/jango_web.jpg?1279741285
I would love to see him battle our other mascot, Alfa Automabot: http://cdn1.alfajango.com/images/alfa_web.jpg?1279741285
Funny story, my co-founder and I designed these guys to be sort of an odd couple, so this seems like the perfect outlet for them! I had actually forgotten all about these little mascot guys until just now.
A graFighter that "defies the rules" would be one with animation, with multiple forms, with unique behaviors - things you can't convey in a sketch.
> it is not cheap, so we are looking to raise 20,000
Why not just submit to YC then? You'll get $20K and much better publicity and VC connections.
Occasionally, fanatical TV viewers will try to get together and pledge enough money to produce another season of their favorite cancelled TV show. In return for that money, they get to see another season of their favorite cancelled TV show (well, they would if the networks were smart enough to work that way).
So if I donated $5 to this startup and got a life membership with like 3 fighting dudes, while they charge $1/mo post-launch with the same plan, I'd call that a pretty fair bargain. Of course, I don't quite know the legality of selling a chance to possibly in the future own a product that isn't developed yet and might not ever exist, but at a discounted rate.
If you don't have to give up equity then why opt-in to do it? If enough people believe in an idea/want it built then it's a fine way to make the money that you need. From the sounds of it they've made it this far in a year with almost no funding so imagine what $20k could do...
Also, I believe Diaspora was asking for $10k and simply had an idea, at least this team has executed.
The idea behind many huge movements today is to be a part of something. If you take a risk to donate $5, you could potentially lose it yes, but you could also say that you helped launch something big.
I think it would be more appealing if you let the people control their characters. The problem with this is it will require more CPU work on your part- expensive for a start up. You could have them scan the characters in and use a "wire-frame", like the use in art, to lay out the different body parts. You could even use some OCR to see it the character had legs or if it is on some kind of "tracks".
Awesome art, awesome idea, can't wait to see a working version.
P.S. RazaBlaza is too powerful
Edit: About the wire framing and OCR. Instead, to make your life easier, you could have your own custom character creator. Kind of defeats the purpose of bringing your drawings to life, but hell of a lot easier.
1. It celebrates 'creation' as opposed to performance. The lose-control-of-performance-during-fight is an aspect that puts the performance of the character squarely in the court of its 'designed' attributes, rather than 'performed' action.
2. It is very intuitive and even 'humanist' if one can dare... Imperfections in character created by user translate into performance (or lack), and tend to become logical extensions of its behavior. Thus, this product becomes a celebration of this very imperfection - in the same manner that flaws in human nature are what make us human.
3. This product celebrates the hand. And how. It seems to capture the nuances of hand-drawn art, and I hope it eventually discerns medium - the quality of paper, the weight of the finger, and the stroke of the pen. If one can dare to think of the ideological similarities between the arts of calligraphy and sword-play - and if this product can begin to map those aspects, these will be very interesting times to live in.
The negatives [or fears, actually - being the incorrigible cynic one is]:
1. I am curious about the level of human intervention from the product-developer end once a user submits a character. I can imagine figuring weapon-attributes, body-limitations, some vectorization and animation 'rules'. If Cornelius can do all this himself - over time, if not immediately - it would be great.
2. The nature of 'conflict' - from the demos I see so far - each character's primary objective is attack-and-kill. If human-centric intelligence becomes a gradual attribute of the character - knowing its strengths and limitations, knowing the same for its opponents, being environment and constraint conscious, etc - it would be a great build-up. I am talking experience here, with which a character grows
3. Where do we go from here? I already see this becoming a cult activity. It can go further from there into the leisure/gaming/media end. It can also become a basis for a better tactical combat simulator. This is where the creators would have to answer the question - is this a product that treads the path of creation, art and design - or goes in the direction of solving(?)/mapping(?)human combat and conflict
This whole "you can't control the player" is a bad idea, it doesn't drive the usual gamer and you're going to find that maybe it's exciting for people to watch their creature fight the first time but there's no repeat value in that (watching the same moves over and over and over again just isn't exciting).
The perfect blend of your idea and the wishes of the masses is to have your 'Cornelius' algorithm analyze the drawing then return a skill set. So it gives the player 6 movies + 10 combo moves based on their drawings attributes. You get a combination of the "my ninja vs. your ninja" as well as the ability to actually influence the outcome and promote replay value.
I've read all of the comments thus far and I have to say that I agree with many points on both sides. However, I would have to definitively say that if you want to control your player, this concept isn't for you. This is of course through the basic player moves that so many people are used to. Click A for jump, B for attack, different combo moves etc. Yes this appeals to many for great gaming, but I would argue that it is a gaming genre that may be at it's absolute peak. Even with today's most complicated games, there is a certain amount of creativity, street smarts and ingenuity in how you play the game.
What I love most about these arguments is that the creator (argued most notably for by dave_chenell) is telling you that you in fact DO have control over your character and the fight. It's just a higher level of control than one is used to. Basically, the 'player' gets to play 'god' in creating characters. By changing the slightest drawing, weapon accessory, and proportions (as is my understanding of this algorithm), one has complete control in how one should act in an all out brawl. If you want your gorilla to have great grappling moves, provide him with powerful arms, etc. I would argue that this feature gives complete control to the gamer.
This game type wouldn't be for anyone, that is certain. However, look at the massive amount of popularity games like Armadillo Run (http://www.armadillorun.com/) have received. You don't have any control over the armadillo in this either. One must use their knowledge of physics, materials, and common sense in order to build a device to complete the objective, like a puzzle. This graFighters idea seems very similar; spend some time playing to learn how the game works, and it'll be just as addicting.
I understand that this concept isn't for everyone and it'll be tough to spread word to gamers who are willing to give it a shot based just on concept alone. But let me ask these gamers something. How many times have you been bored maybe in class or talking on the phone, and you doodle some complex robot. Maybe you give him a lightsaber for good measure. Wouldn't it be nice to put him to some use instead of tossing him away with your class notes?