Hey everyone, I want to make a game like Street Fighters for IOS, i have some queries to ask. Should i go for 3D models or 2D sprites what would be less easier and less time consumer?
Both can be solved. Modern fighter games usually use 3d, and I think it is much easier to animate.
Depends on your which skills you are stronger with. Both have their own challenges.
I would note that if intend to have a lot of assets, 2d fighting can get pretty expensive (in size and resources) on a mobile device. (see game in my sig)
You can have the best of both worlds. Watch this:
I’ve tried it. It works great and it is free! It’s time to dig up Illustrator again…
I’ve had some gruelling experience working with both sprite-based and 3D beat-em-ups. Both approaches have major pros and cons and it can have a huge effect on development if you pick the wrong one. The short and sweet is: if you think you’re going to have many thousands of frames of animation, go with 3D. .
If you’re just thinking of a small casual game as per zombiegorilla’s, or something with a retro vibe, then sprites will probably be easier and give you more control over the finished art quality/style.
Casual it might be, small it is not. That is why I brought up size. Just speaking in terms of characters, we have about 350 characters, with 7-20 animations each resulting in 900-1600 frames of animation per character and 50-200+ sprites. (some characters have upwards of 19k keys). All done by hand (art and animation, no bone rigs).
But the challenge is that a character, even heavily optimized, can have an atlas that is 4096x4096. FX (depending on character) may be fairly large as well. By the time all the elements queued up for a battle (audio, backgrounds ui elements, ect) and 6 characters with their fx, you have pretty much blown available memory. Size is an ongoing battle for us, as the many of the newer characters being added are larger and more complex. (and have more atlas busting elements like cape, wings and other flowy things).
3D definitely would have been more efficient and less problems with performance (and an easier pipeline), but it would have been a compromise on art direction. If the art style were lower fidelity (like retro/pixel), then that would also have been less of an issue for memory.
Yeah I hear you. We always had the opposite problem: a couple dozen characters with hundreds (up to ~500) frames of animation each, including FX overlays. It’s a hellish battle, especially when you have to start optimising and the only option is to write tools to cut things into sub-sprites…and then the fun really starts.
3D versions where more efficient memory wise for animation, but you swap one set of difficulties for another - as you say, nailing art direction is harder as you lose flexibility…can’t just touch up ugly frames - and you have to tackle issues with broken skins, bind poses, bones screwing up because someone scaled a bone somewhere by accident (or did they - FBX issues can be a bitch to debug), kinks in animation blending and so on.
There’s no real perfect solution. Now I’m older and wiser I avoid fighting games altogether 
Indeed, on the surface, a fighting game would appear to be relatively simple, in execution it has its challenges. Like you said regarding 3d, some stuff would be ok for an fps or 3rd person other types of games, but with fighting, the characters are in your face all the time, flaws are much more noticeable.
We are at the point and sheer scale of content in the game, that optimization is focused in characters. Even though there are things we could now do overall, it would mean manually editing nearly a thousand animation source files. Which is bad, so very bad. And so many times aspects where handled “one off” that automation isn’t effective.
No kidding, this is my second one here, and I am trying to avoid doing a third. (even though the IP for that one is the best ever) I am so OVER fighting games.
I feel you 
I think your thread should be required reading for anyone starting out and wanting build a game.
It is a very fascinating read and illustrates that perseverance is the key ingredient in building a game. Let’s be honest (and I hope you agree), in the beginning and early posts seem very much like any typical overly ambitious noob destined to give up in a few weeks. Or worse, finish with a lame half-ass “beta” in a few months, thinking was “teh awesome!” Instead, you hammered away at it for 4 years, improving every aspect and polishing every element of the game (not just the core elements or elements that you were most interested in). It is the pretty much the same game you described 4 years ago, just now it appears the be the product of a much more advanced developer. (presumably the skills honed from building this game). It’s a very positive example not only the full progression of building a game, but also great example of one done by someone doing the art, design/ engineering.
So, based on when you started, and where the game is now, how different is your opinion about how “easy” or “time consuming” building a fighting game would be now vs when you started? For example, did you expect it to take 4 years?
Usually in fighting games the sprites are fairly large, taking up more of the screen than in a typical platform game, so I think you will have a lot more to draw especially if it’s higher resolution. If you have the skills and time, 2D is cool… otherwise there can be some savings there by going 2.5D/3D particularly if you use bone animation.
Wow thanks ! I’m honored, especially since you are behind the juggernaut that is Avengers Alliance (congratulations on the release btw !)
Haha yes indeed, the first posts were totally misinformed about the quantity of work that needed to be done. I’ve came upon a post I wrote on Facebook to my friends back in 2009, stating “starting a fighting game, here we go for 6-12 monthes of work !”. LOL
I’m still working on it right now (the nightmare called synced multiplayer), but I agree that completing most of it did make me learn a lot of things I wouldn’t have even suspected handling 4 years ago. The simplicity of Unity helped a lot, along with the awesome community here.
Having a presentable, finished product brought me a lot of unbelievable opportunities too (jobs, business, travels, etc), and is still continuing. So yeah, I could never emphasize enough on the importance to never give up on something that we really want to create. Whatever the time it takes, it always pays in the end
I think we both agree that fighting games might be one of the most painful genre to develop
You nailed it earlier : « Indeed, on the surface, a fighting game would appear to be relatively simple, in execution it has its challenges. ».
I guess it differs from fighting game design to design, but personally what took the most time were the animations (which seems to concurr with your experience), AI and multiplayer. With a special mention to multiplayer … Actually anytime I’ve been asked about how difficult multiplayer is for a fighting game, I couldn’t hold a « run away from multiplayer !!1!zomg »