[quote name='freakchild' timestamp='1338839107' post='4946202']
A simple example of this would be the definition of 'grip points' on one wrestler - basically positions the other wrestler is trying to grab. The wrestler doing the grabbing will more than likely be animating via several blended animations and supplemented with dynamic animation via inverse kinematics in order to do that - no simple 'replay' going on there.

Nice man! I am in a position where I get to see the animations for one of the best wrestling series getting shaped and tuned everyday. You are spot on with that information there. And yes, the amount of custom tools "just" to have the animations looks great and work in all situations (eg. tall char with a short one or vice versa, fat and thin chars too and a lot of other situations) is astounding.

So yes, wrestling and most close contact sports are DEFINITELY not projects to pick up at the start. Pick something simpler like bowling, Slateboard.
[/quote]

Oh, this most definitely isn't an attempt at starting with pro wrestling. I'm merely working to obtain knowledge for the future.

It's something I've decided long ago. I'd either join a development team that was working on a game (seems THQ is the only game in town), or put forth the effort of making my own.

I'm currently in the same position as you, my goal is to make a pro wrestling game. I study Computer Aided Design and I've been independently learning to be a artist and animator(well mostly artist) from the age of 12. I remember a few years ago, I made this. http://i55.tinypic.com/23jgwh.png
http://i54.tinypic.com/s185xf.png
http://i51.tinypic.com/357hf6h.png

WOW I've seen a lot of complex talk about how wrestling games work and I'm soo confused.

Well in the future, I want to develop a ROH/TNA game but I have to start at a very very very easy project.

I'm currently in the same position as you, my goal is to make a pro wrestling game. I study Computer Aided Design and I've been independently learning to be a artist and animator(well mostly artist) from the age of 12. I remember a few years ago, I made this. http://i55.tinypic.com/23jgwh.png
http://i54.tinypic.com/s185xf.png
http://i51.tinypic.com/357hf6h.png

WOW I've seen a lot of complex talk about how wrestling games work and I'm soo confused.

Well in the future, I want to develop a ROH/TNA game but I have to start at a very very very easy project.

Those images look very nice.

As for it being confusing, I see it as being a work in progress. It's confusing because I haven't tried to learn it yet.

I've made even more stuff but some of them are not related to wrestling.

Let's say I wanted to make a powerbomb animation. How would I make it using one human model? I guess I need to learn more about animation also programming(I have beginner knowledge).

For any general wrestling move, the technique of authoring the entire move is going to be one of considering how to start the move (how to bring the characters together in order to do the move), how to do the move and how to end the move (which is really going to be how the characters separate OR how to transition to the next move).

Any of those stages is going to be a combination of authoring direct playback animations and then adding other blends, dynamic controllers (inverse kinematics and/or physics), meta data and potentially additional tween animations in order to fit that animation to the wider range of circumstances. Generally you just start with the simple direct playback and then add all the improvements in as necessary to cater for the required variety and to smooth things out. It’s not simple, but also not as hard as it sounds it’s more effort driven than anything and the success of it will largely be based on the variety of tools you have to help you meet the end goal. You see, all these controllers and other animations really are just tools to selectively apply in helping you to get the best result. Best to have many tools and use the best tool for the job then.


I should add that all that is going to be solved by the work of at least an animator AND a programmer (who is likely going to develop a very flexible animation engine but also supplement some of what the engine is doing with special case logic).

I don’t know names of specific moves, so I had to look powerbomb up. That move sounds like from a certain point the move could be a direct playback of two sequenced animations on both characters with some minor dynamic adjustment based on the sizing of the characters. Probably the biggest point of complexity in authoring that move is one of getting the characters to that point in the first place, which is largely a general problem for all moves as indicated above.

Assuming the move is actually possible given the location of the two wrestlers, it’s likely the AI will select a given animation (likely there is one or more that best suits that location) and then either use some level of AI to add additional blends and controllers to better fit that animation or sequence to the situation. This can be logic that is hard coded, which takes a lot of work but generally it’s best if the logic is driven by metadata that comes along with the animation – hints to the logic about what it should be doing if you like.

I should add that the most basic level (and fallback) of intro to move sequence is going to be walking to a specific position. That will work, but it will look crap. So again generally you start with basic intro moves and then add more specific ones in to make them look better.

When that animation is complete, the two players would in theory be in place and can playback a more synchronized animation specific to the technique. If your wrestlers are all the same size then there’s likely little else to do as you won’t have to cater for differing sized players, but if not you may have some blending and controls to take into account. You may also want to do that anyway just to add variation so if you replay the move over and over again it doesn’t look the same.

After the move is complete you’ll process animations to either separate the characters or start animations that blend or tween the wrestlers to link up with the next intro phase of the follow up move.

One thing to take into account is that depending on how you are implementing this, there may be parts of that animation that are driven solely by animation, inverse kinematics and other meta data…or there could be a physics controller to take into account too. It’s not uncommon for an animation to be largely driven by the animation data at various points in the entire cycle, but at other times acknowledge physics processing and take that as an input and control. A simple example of this would be picking someone up and throwing them – which could be fully animated until the point the throwing character ‘let’s go’.

Hope that helps. It might give you a better idea of what authoring and processing animation for any specific move might be.


You both want to make better wrestling games it seems - nothing wrong with that it's very honorable and well intended I might add. One of the reasons you might not be satisfied with the games you are seeing is because this stuff is very hard to get right. It’s very difficult for a game based on needing to animate to cover all corner and edge cases without applying constraints that affect the gameplay and even with dynamic input blending to animation there’s a limit to how much influence that can have/how many corner cases it can eliminate.

To understand this more…probably a good thing to look at is older wrestling or combat games (even 2D ones because they show the constraints quite well). Many of those games relied on playback only – no blending or dynamic control whatsoever. Hence why in many cases, you could only perform certain moves if you were in a specific position and you couldn’t always follow up one move with any other move. Combos or follow ups were quite often restricted to moves that were either authored specifically for or were just simply compatible with the current ‘state’.

That was just the state of the art at the time, but it also explains why those games had flaws that today wouldn’t be acceptable. Some of those games weren’t bad with those flaws or constraints of course and there’s probably some inspiration or a lesson to be drawn from that too.

Wow thanks for educating me. I now understand how animations for wrestling moves work. It seems like you have had history in making wrestling games. Hopefully I'll make a good animation one day.

Also I have another question. How would I make an object break and it will look realistic? For example a character falling into a table and the character responds realistic and the table breaks realistic(how it should in wrestling). I've never seen it PROPERLY done in a wrestling game.

Thanks

Thank you for the information freakchild.

It seems like you have had history in making wrestling games.

Nearly. I was once scheduled to be involved, but I just managed to just miss out on that while waiting for a work visa.

I did a lot of animation programming prior to that though and did one of the earlier examples that deployed a lot of real-time blending - which was a new technique back then. I kind of specialized in animation systems and programming for a few years as a result. I've followed the topic on and off ever since.

Also I have another question. How would I make an object break and it will look realistic? For example a character falling into a table and the character responds realistic and the table breaks realistic(how it should in wrestling). I've never seen it PROPERLY done in a wrestling game.

Yeah, the wrestling game I was supposed to work on had that problem I think.

To do it properly you need a real physics implementation. Usually you switch out the solid model of the table for one already pre-broken up into pieces and then apply physics correctly (transferring the energy from the falling body correctly) to the pieces.

That's typically how it would be done even today, although in the last few years there have been games (not wrestling) that have explored breaking objects up dynamically instead of switching to ones that are manually broken apart. It's had mixed results, but it is probably going to be 'state of the art' moving forward.

That's just for the table though. For the character you'd usually blend a ragdoll type of physics controller in there for the character - yet another type of dynamic controller.

I wonder how all of this would play into a move create mode.

I wonder how all of this would play into a move create mode.

I was thinking this aswell. If the moves are made complex. How hard would it be to assign moves to people?

Thanks for the reply. Freakchild

Also are you aloud to tell us what game you was going to make?