GameDevBlog

Migrating to Unity 4 Mecanim

Mecanim does some things I really like:

• Makes it scary easy to retarget animations to different models.
  With Unity 3, I did a lot of my work with the stock character Unity came with, and then when I wanted to use a new character, I had to use all new animations. With Mecanim that isn't a problem, you can simply retarget them. Even though Mixamo claimed their animations didn't work with Mecanim, I didn't have a problem - simply opened them up, selected Rig, targeted humanoid, Apply, done. (EDIT: the Mixamo animations I downloaded from the Asset Store as part of the Action Hero Motion Pack worked, but the ones I get from the Mixamo Store don't. :(  Good to know...and you can still get working ones from their website.)
• Lets you chop up your animations, set their loop and pivot points from inside the Unity editor.
  In Unity 3, I had to take my jump animation, note at what times in the animation the transitions I cared about happened (crouch to leap; going up to falling down; hitting the ground...), and hardcode that into the script. Plus now you can just click a button and moves the pivot to their feet - which is where I almost always want it.
  With Mecanim I just sliced it up into separate clips.
• And the demo has some nice smooth run and idle clips that I like better than the ones I was using. (There were some missing frames in my run. Did anybody notice?) I simply retargeted to my character and they're mine now.

What I don't like so much is the animation state machine I had to create.

Apparently, this will be fixed in a later update - you can't really use Any State as it stands right now, so I've got this mess of transitions of, well, just about any state to any other state, because it's easy to go from jumping to flying - to running (on a wall) - to crouch-running - but then you might stop and just be crouching - and so on.

AFAICT, you can't use mecanim clips in the legacy animation system. So I couldn't just take the new smooth retargeted clips and replace my old animations with them.

And there's still some bugs in my spaghetti system. Surprise. The joys of stateful programming - there's a state in the code representing the character's, well, state, and a state in the animation system, and communicating back and forth provides plenty of opportunities for screwups.

The code, in theory, can tell what state the animation is in with something like this:

if( swingAnim.IsInTransition(0))
{
    AnimatorStateInfo asi = swingAnim.GetNextAnimatorStateInfo(0);

    // do NOT forget the space in Base Layer. (I did.) 
    int jumpFlyHash = Animator.StringToHash( "Base Layer.JumpFly" );
    if( asi.nameHash == jumpFlyHash )
    {
        lastJumpUpTime = Time.time;
    }
}

But I was seeing mysterious delays when I used that method to trigger state changes that I never figured out.

One last plus, though - is being able to look at the state graph while the game is running is a pretty cool way to visually debug. You can't do that with Unity 3. Very cool.

Here's what I've done so far:

 

Big In Japan - Playstation Mobile Sales Data

Got my first month's sales data for sixty second shooter Deluxe. Without further ado:

Australia Canada France Germany Italy Japan Spain United Kingdom United States	23.13 30.08 10.99 12.56 9.42 322.83 7.85 51.68 235.53
	704.07

That's in Euros, bringing my total up to 910 - about 500 units. 

sixty second shooter Deluxe had a number of reviews in the low twenties at the end of November - so, looks like about 1 in 25 players gave it a review. If you assume the same ratio for other games, you can get an idea how other titles on the platform are doing.  

It's about what I expected so far. At the moment I don't want to make another game for the platform, but maybe sssD will have a long tail, or maybe Sony will promote it some (in the North American PSM store, anyway, the same five titles have been in the top bar ever since it launched...what's up with that?), and maybe PSM will grow as it gets more adoption on Android.

Surprised that so many sold in Japan! It wasn't even localized into Japanese.

In other news, I'm hearing that sales in the Windows 8 store have been horrible. Worse than this. So it seems I backed the right horse. But it also makes it look like my whole "launch titles are awesome" theory is wrong: the Chrome store, PSM, and the Windows 8 store have all been disappointing. I'm really surprised by Windows 8 - if anybody had the muscle to make a store decent it was Microsoft, I figured. But I never intended my 'make launch titles' strategy to mean 'make one launch title and then give up' - I knew it was a numbers game, and you'd have to make a bunch and maybe one would hit big.

But the platforms on the near horizon don't seem like they're going to sell a lot of games. I love the idea of Ouya but even if they sell ten times as many units as they have during their Kickstarter it's still a fraction of what one of the Big Three consoles do. The GameStick and nVidia Shield, probably similar deals. Steambox will be huge, I bet, but launch titles will be competing with Steam's entire linux back catalog.

Hmm. If I had to bet on another platform that'll help some new titles go big, I'd bet on Apple TV...

The New Plan - Delaying The Kickstarter Yet Again

When I first got the idea to do Energy Hook I planned on Kickstarting ASAP so I could see just how brilliant or dumb an idea it was. But in the process of working on the pitch and prototype I've fallen in love with it. I can't not finish it now. So I'm going to. Finish it. No matter what, even if I don't make a cent off it.

I want to do the Kickstarter anyway. Why? To find out if I should make the smallest, minimally viable game possible - or if we'll have the resources to do something grander.

That being the case, Paul urged me to wait more on the Kickstarter, so we can make the game look as good as possible before we pass the hat around. I really didn't want to wait. I've already postponed it once.

So I did that thing that people do when they have tough decisions. I made a list of pros and cons. It looked something like this:

Should I do the Kickstarter now?

Pros:

• It's eating my brain. I want to get it over with.
• Slipping again is embarrassing.
• We'll find out now how much resources we have - we might have to redo less stuff. (For example, right now we're making everything fairly flat-shaded and colored-in - with a lot more funding we could have higher surface detail, but we'll have to redo those flat-shaded models.)
• Once the Kickstarter is over, we'll focus more on making a good game instead of making a good kickstarter pitch.
• Kickstarter fatigue. It's getting harder to have a success on Kickstarter as time goes on.

Cons:

• The longer we wait, the better the game will look, and more people will want to fund it.
• The longer we wait, the more we can build awareness and our social networks.
• And if we wait long enough, we can have a first playable to show journalists before the launch and to give as a reward to higher tier backers.

Although the list of pros seems longer than the list of cons - the pros are mostly psychological issues of my own. Yes, I'm devoting too much brain glucose to thinking about the Kickstarter, but I don't have to be. Yes, I'm working harder on making our Kickstarter pitch look good than making the game good, but I don't have to be. Yes, I'm embarrassed about slipping again, but I don't have to be.

And yes, "Kickstarter fatigue" seems to be setting in; there are fewer successes on Kickstarter these days. But I don't want people to fund the game because Kickstarter is hot, I want people to fund the game because they genuinely think it's worth funding.

The only semi-reasonable pro is the one about redoing work, but if we raise enough funding that we have to redo work, that's ... actually a really good problem to have.

So, the cons have it! We're slipping.

Our new plan is to get something playable going, which could take anywhere from 1-3 months, all the while Paul making more better art. He's finished his teaching gig at Digipen and has more free time to make stuff. Then we'll have something to hand out to a few select journalists and players. (Which is another reason to Like us on Facebook - we'll give away one or two to our FB followers.) And then we'll do the Kickstarter.

What do you think?

Measuring The Purpleness

I have two incompatible beliefs. One is that we should make remarkable games. Another is we should just do it: write some code, lay out some assets, make some game, whether we feel we have a remarkable idea or not. Because perfectionism is an unhappiness habit. Because if we wait for that one perfect idea we might never make anything. 

They're not that incompatible, actually. We can make a lot of games that we think aren't-that-remarkable, and maybe that's our insecurity talking. Maybe they're more remarkable than we think. They didn't expect Canabalt to be as huge as it was. Sage LaTorra just commented that he didn't think Dungeon World was particularly remarkable, but it is, in a variety of ways.

So how do we know? If our game is unremarkable or we're just being insecure?

If you're working on a game, you can get an idea how remarkable it is by sharing early. Post a youtube video. What percentage of likes do you get? How many views? You can compare that to other games. Put a link on reddit. How high does it get voted up? How much do you get retweeted on twitter? Etcetera. Far from an exact science. But it gives you an idea. Some say it's dangerous to show crap early, because people will think and say negative things about your game, but the real problem there is just that it hurts our feelings. It's not necessarily going to hurt our success down the road, and may even help as we fix and address those issues that they complain about.

And supposing our game is not that remarkable yet. Almost nobody Likes. Nobody comments. What do we do then?

Well, we could abandon it and start over. Try another idea. Shoot our baby in the crib. That's probably the right choice - stop trying to polish the turd, stop throwing good resources after bad. We've learned a lot and hopefully enjoyed the process and we'll probably just have to accept that's all we're getting out of it, this time.

Or we could keep working on it, trying to make it remarkable.

Nobody can help us with that decision - even if we write a long list of pros and cons, assigning point values at each one, we're still going by our gut at some level.

One thing we shouldn't do is keep shilling our unremarkable game. Let's not bother the press and our friends anymore until we've made some kind of significant improvement. 

So here's my question: can you make an unremarkable game remarkable? How would you do it? And let's hear stories about games that were or seemed unremarkable at first but turned a corner.

Purple Cow Theory Applied To Schizoid

Continuing my thoughts on Purple Cow as applied to games.

Just to reiterate, you've got to stand out, be a purple cow, to get noticed in this cluttered world. "Be remarkable," is one of Godin's key phrases.

You might get the idea from my last post that game-jam-sized games can't be purple cows. Of course they can. Schizoid was one. Schizoid actually was pretty remarkable - not only was it aggressively about the co-op, it was weird. Hard-to-describe weird. There was no good elevator pitch for it. There was no shooting! It was kind of like Pac Man in the way you'd eat dots, but analog - you could move any direction, and then there was the whole color coded thing on top of that. 

But to this day, when I talk about Schizoid to people, if they've heard of it, they say, "Oh yeah, wasn't that a twin-stick shooter?"

To which I usually say, trying to keep the annoyance out of my voice, "Well, I guess, but you couldn't shoot."

But this isn't their bad; this is human nature. We see something and we immediately want to put it in a category and feel like we understand it. If your game even looks slightly like another game in its screenshots and trailers, people are going to assume it's more-or-less a clone, lump it in the category, and move on. People looked at our screenshots and said 'glowy bloomy saturated colors with spaceships - must be another twin-stick shooter.'

So it's not enough to be remarkable, you still need to get the message out why it's remarkable, and overkill is probably the order of the day there. Maybe we should have named the game something like "Co-operate!" (looking at Loadout! and Incredipede as examples there - some of what makes those games stand out is right in their titles) and maybe our look should have been something like this:

Something like that anyway. Something that would clearly show the homage to Pac-Man, clearly show it's a two-player game and color-coded, and not remind anyone of Geometry Wars or Flow, the games du jour.

And maybe our trailer should have explicitly said, in big letters, "Red can eat red. Red cannot eat blue." and vice versa, so people who weren't paying that close attention would notice it.

How is your game remarkable and how are you making that clear?

How purple is your cow?

I read Seth Godin's Purple Cow last year and it was a revelation: it made me realize I had lost my way as a game maker and artist.

The main thrust of the book - in a sea of white cows, a purple cow will stand out. Don't try to be like everyone else, do something remarkable. Literally "remarkable" - as in, people will make remarks about it. News about your product will go at least a little bit viral.

When I read it, I had been making game-jam-sized games like We're No Angels and sixty second shooter, and wondering if there was anything I could do to make sixty second shooter more succesful, but I had to be honest with myself: sixty second shooter is fairly unremarkable. It's just a mash-up of a couple other genres with programmer art. Sure, there are a few remarks one could make about it:

• it's actually really fun (but fun isn't enough, of course, there are thousands of really fun games out there)
• it used new technology (but, hey, the players don't care)
• barriers to entry are low - as long as you're running chrome the load time is nearly instaneous, it won't take too much out of your day, you don't need to sign in and give me an e-mail address (but so are a thousand games on Kongregate)

Which is a big part of why I'm stepping up and making Energy Hook. It's the most remarkable project I can think of that we can do with almost no funding.

Also, Purple Cow has given me a lens with which to look at other people's games. I often will give people feedback on their games and almost always the first thing the games have me asking is, "What is remarkable about this? Why should I play it, and talk about it, when there's a thousand other games out there?"

Because, wow, the game market isn't a sea of white cows. It's a technicolor ocean of all shapes and colors. To stand out, your cow has to be a frickin' ultraviolet light.

So - here's something we can do. Put a link to the website for the game you're working on in the comments, and, in just a few sentences, tell us why it's remarkable. What makes it purple? Go!