Saturday, March 15, 2003

Notes On Rockstar

So I'm checking out this Indy Game Thing. Today I spent the morning playing the Uplink demo, and liked it enough to order it, and also played the Rockstar demo, which I enjoyed all afternoon, but didn't give me enough incentive to pay the whole $6 for the full version.

Rockstar simulates being a rockstar, and is more of a vehicle for satire than a game. To play Rockstar you have to fine tune your drug use and holidays in order to balance your creativity, health, alertness, and happiness. You then attempt to record albums when you're at creative peaks and make money touring. After a few hours of this I never quite caught on to the relationship between albums and touring, or even if there was one. I played three times on three difficulty levels. The hardest--masochist--was the only one where it got interesting. It may have been my imagination but there seemed to be some kind of negative feedback as your stardom progressed: vacations became boring, your fans get tired of you, the studios get sick of you.

What made Rockstar cool was its satire; you can get busted for LSD posession, but it boosts your creativity. Heroin can kill you. (That's how I lost my first game.) You can sacrifice happiness for your other attributes, but if you get too depressed you'll attempt suicide, which will lay you up, uselessly, in the hospital for a while, or even kill you. (That's how I lost my last game.) Your mother offers you drugs. You get laid. Your recording engineers swear at you.

But in the end I have to give it a pass, because the gameplay lacked clarity. Exactly what it was that made gigs succesful or not was a mystery to me. It's fine for a game to have randomness but I was completely clueless as to how to have a good gig or not.

I'm told that at Bullfrog they prototyped Dungeon Keeper with an all-text game first, before they went into production. So who knows? Maybe with more work this game could be turned into something very cool. It certainly has potential.

Thoughts On Working On Licenses

Let me get two more cents in on this Spector vs. Costikyan discussion. One thing I've noticed in this thread is the assumption that everyone would prefer to work on an original IP than someone else's franchise. (Although Greg does mention the possibility of dweebs sending our resumes in hopes of working on Spider-Man 3. Hey! That really hurts.)

I've actually mostly worked on original IP: MindCraft's Magic Candle universe and then Die By The Sword and then Draconus. Even ignoring the fact that Interplay owned the Die By The Sword world we created, and that Crave owned the Draconus world, you can probably guess why this wasn't too exciting for me. All of these worlds were Tolkien/D&D derivative shlock. Orcs, kobolds, dwarves, and elves. Fucking elves! Given a choice between that and Spider-Man, I'll take Spider-Man in a heartbeat.

I'm just speaking from my own experience, but I imagine this is a pretty common phenomenon. If you're not one of the few people in the core creative team, what are the chances you're going to buy into their vision? Even if their vision is something cool the peripheral team probably won't buy into it. Freedom Force would be a good the Game Developer post-mortem, I believe it was mentioned that the team in Australia did not buy into the vision that was being force-fed them from the office in the States. And IMO that was some brilliant IP.

Especially if you're a programmer, like myself. Although an artist or game designer can usually put their own unique thumbprint on the core team's vision, a programmer doesn't really have that outlet.

Of course, if I was in the core creative team, I'd rather create my own IP than go with a franchise. (Once I worked on a title called Gryphon Masters that was all me. I decided Celtic myth was an un-mined area that needed to be exploited by FRPG's, did a ton of research, and with the help of the rest of the team [including Ed Del Castillo and Steve Burke] created some really cool stuff. But I knew crap about project management back then, and if MindCraft hadn't folded first I would have had a huge disaster on my hands.) Even with all the problems that would introduce, the need to find some way to get the whole team to buy into an unknown, the lack of support from upper management, and the like, it would be more fulfilling. So my only point with this article is to point out yet another pitfall with non-franchise titles, something to be aware of.

Notes on Jak & Daxter

So Jack & Daxter benefitted from the consulting of Cerny Games. Presumably, they used the Cerny Method. And what did they get for it?

The Cerny Method involves a prototyping phase. In project management terms, game production with the Cerny method starts with a spiral of several iterations, until you have the prototype you don't throw away, and then you enter full production. The prototyping phase, ideally, allows you to take risks, in the search for some innovative game design mechanisms. If the risks don't pan out, you cancel the project. If the risks do pan out, happy happy joy joy.

So I'm a little disappointed that Jak & Daxter offers very little new. It's Mario 64 with the addition of some cool stuff; a mode where you fly 'zoomers', these flying motorbike like things, and some clever moves, such as pole swinging. Daxter, the furry sidekick who sits on Jak's shoulder and occasionally makes amusing comments, is not gameplay but color. On the technology track, loads are well nigh invisible; just as good as Mario 64 (and remember Mario 64 was on a cart), and better than Sunshine. Clever level of detail makes you feel like you're in a world, instead of in a series of disconnected levels.

So I've got to ask; is that all we get? This is all that can be invented with million dollars of prototyping money?

And then I think about it, and really, it's not surprising. The mode of development for most studios, I believe, is to spend eighteen months making a prototype and then ship it whether it was good or bad. So I should have expected that a million dollar prototype was going to, in the end, produce less innovation than the more typical "throw ourselves off a cliff" method.

We could learn the wrong lesson here and say, "Well, then, the Cerny method must be wrong." But stop and think about everything the game does right. Ed Del Castillo once commented that most games that hit the shelves these days don't feel like they were actually finished. Jak & Daxter is an exception to this (although there *is* a Yellow Sage in the final level, which kind of implies they probably intended to have a Yellow Sage World, which didn't make it into the final product). Jak & Daxter is polished. Part of the reason, no doubt, is that they locked down the feature set early in production; after they finished the prototype. The real lesson to learn is that there are limits to how much anyone can innovate, and it's a good idea to have a process in place to curtail further risky innovation later in the project.

Another thing that I love about Jak & Daxter is the number of times I wanted to throw my controller through the screen was very small. This is amazing for me, because normally I hate platformers for that very reason. I never completed Mario 64, and Mario Sunshine definitely set my teeth on edge on occasion. Also, I never once had to get up and consult to figure out how to get through a level. This is no doubt due to another aspect of the Cerny Method: namely, "Gameplay testing is your most important feedback." One thing the best designers say time and again is you've got to watch people play your game. (In the interview sections of Richard Rouse's book it kept coming up: Ed Logg's field tests, Chris Crawford quoting Dani Bunten, Steve Meretzky saving the play logs for his text adventures.) Mark Cerny has made a science out of this, collecting statistics on how long it takes people to get through levels and where they get stuck.

So: prototype and gameplay testing. Interestingly, Treyarch's best game--Die By The Sword--did these things, albeit by accident. If by prototype you mean, start making something, and then change focus several months later and make something completely different. Die By The Sword went from being a fighter to being a level-by-level dungeon exploring game around a year into its development. Also, I think Die By The Sword was the first and last project where we actually watched people play our game, and tried to take steps to remedy the problems that we discovered because of it. (Although we took very few steps; usually by the time we saw the problems someone was having, it was too late and too risky to change anything.) With subsequent games we relied solely on our publishers for feedback, and maybe that's why Die By The Sword, of all the original games we've done, has gotten the best reviews. (I could make some snide comment about publisher feedback here. Consider it made.)

Wednesday, March 12, 2003

I Feel Stupid Now

Just read Raph Koster's Small Worlds slides. Damn. I didn't get half of it.

Something that concerns me: it seems like he's saying most games create a phenomenon where the best player clearly dominates over the second best (and the rest.) Which fucks up my measurement of "depth" (still looking for a better word, please help): it means checkers is just as "deep" as chess.

The Light of Day Brings New Softness

I was in the bathroom this morning and my wife had a copy of Entertainment Weekly open to an interview with Harvey Weinstein of Miramax, and it made me reverse myself on the position I posted last night. No doubt, Harvey Weinstein is the kind of guy Greg Costikyan would like to see at a semi-major game publisher. (Is there a Miramax of game publishers?) He's the guy who funds art movies, is willing to take the risk that they might not be profitable, and it seems like for every two profitable movies they do they also have an unprofitable one. It's thanks to him that we have Lord of the Rings and Pulp Fiction, which they then used to fund some of their flops.

So why can't games be this way? Why can't there be a publishing house that funds art games?

I have a theory: it's because we all want to get rich. We aren't satisfied with just being minimally profitable. People all behave as if a game isn't a megahit then we lose money on it. Let's see if that's true. If we accept Greg's numbers of $3 mil & 3 yrs. (Though at Activision it seems more like it's $5 mil & 2 yrs - I bet $3 mil & 3 yrs would make better games...) And we throw in another $3 mil for marketing, then the game has to sell something like 600,000 if it's a console title and 300,000 if it's a PC title. If we go look at our illegally copied npdfunworld data, we discover; only about a sixth of games do those kinds of numbers. So, theory disproved.

Okay, theory 2: the economics of game publishing make it impossible to make art games. Greg Costikyan is right. We can have art movies because there's so much money in movies--and the movie theaters (or "retailers" and "console manufacturers") get a much smaller cut--that tiny markets of afficianados can support lower budget efforts.

This really is a dark time. Things will change, eventually. As tools improve it will become cheaper to make games, and as games get more mainstream acceptance more money will come in. When the internet eliminates retailers, still more money. Eventually we'll hit a point where there can be a Pulp Fiction of the videogame world, and we'll be able to choose between getting rich off of blockbusters or surviving by our art, instead of having to make blockbusters simply to survive.

That new time won't be all bread and roses. The art house publishers will still have to do semi-shady stuff, like campaigning to win awards for their art house games. Part of where Miramax becomes profitable is by campaigning to win the Oscars. Artists are whores, too.

Aren't blogs great? You can watch people change their minds in real time.

What the hell am I doing still awake? I wanted to go to sleep two hours ago.

I love Greg Costikyan's site but he has a Chris Crawford-ish rant that I enjoyed reading that's implying things I don't agree with. The crux of the thing: now more than ever, game publishers are avoiding risks and thus all you can expect to see in your future are sequels, movie-tie-ins, and me-too titles. The suits at these companies are going to cover their asses and will avoid risks to keep their jobs even if it means the company's bottom line is going to suffer. So indy game developers are getting fucked.

We were a starving indy game developer with an innovative title, once. We shipped late, missed our marketing window, and the innovative title tanked. Then we went into survival mode. We've 'sold out.' So I've been there, I know what Greg is talking about.

Here's what I'm wondering. What, exactly, is Greg's problem?

Is it that not enough innovative games coming out anymore?

I have to disagree. Somehow, innovation keeps slipping through. It's been a few years since The Sims, sure. But there's Animal Crossing and Pikmin; there's Kohan; there's Neverwinter Nights; there's Ico; there's a slew of titles that are broken in various ways but certainly inventive (Europa Universalis, Gothic). Devil May Cry was a new kind of third person fighter. And even sequels innovate: Metroid moves to first person, Mario gets a water gun, Grand Theft Auto gets motorcycles and remote control helicopters and an eighties soundtrack. And Tony Hawk (which Greg dissed on specifically) got it's new mission-based structure which really brought out a lot of the latent coolness of the game for me. If companies are trying to avoid risk, how did this happen? Because companies realize they have to seize opportunities. Because when EGM comes and asks you "Why should our readers buy your game?" the answer, "Because it's exactly like the last game." won't fly.

Is it that starving indy game developers can't make money?

This seems to me like saying we should get paid to make art. I think you should be about as upset that such-and-such struggling garage developer can't get a major label contract as you should be upset that I can't get my novel published. If you want to make a totally inventive, high-risk game, great. Fantastic. Just do it on your own time, with your own money. Make art for art's sake. Suffer for your art. The stories of artists who got rich (T. S. Elliot, Andy Warhol, Will Wright) are anecdotal. Most artists remain unknown. Accept it. As Peter Akemann puts it, you have the choice of being an artist or a craftsman. If you want to get rich, you better become a craftsman.

Maybe I'm just bitter. There's definitely been times I wished I was away from the pressure cooker of the thirty-man team working on the big franchise and in some kind of home office writing Slay.

Tuesday, March 11, 2003

A Better Word Than Depth

Okay, people already don't understand what I specifically mean when I say 'depth', as I was talking about in my post of 2/25 and then again on 3/1. So what's a good word that means 'Having the tendency to evoke complex higher order strategies and patterns from simple sets of rules'? I have the feeling there's already a word out there that means exactly that. Maybe something German. Post your comments, give me some ideas.

I'm thinking out loud for the rest of this post. Bear with me. (Or skip it.)

On the topic of acheiving this quality...whatever we're going to call that first post I reasoned that in the vast space of rules sets there are only a small number of systems that produce emergent patterns...and this is why we're stuck in a design-playtest-iterate approach to making games...but is that really true, I wonder. Maybe some human faculty can be used to narrow in on the cool systems quickly. I used Conway's Life as an example. The question is how long did it take him to find this cellular automata? Basing it simply on population, instead of arbitrary patterns of surrounding cells, seems like an obvious choice in hindsight. And then once you've done that, it won't be long before you've eliminated most of them due to under- or over- populating. So maybe he discovered Life quickly. And when designing games, we can all narrow in on the simple sets of rules that work in just a few iterations.

And here's another possibility: just because Life is the most famous of two-dimensional cellular automata doesn't mean it's the one with the most emergent patterns. The ones that overpopulate create all kinds of patterns, and just because they're static doesn't necessarily mean they're less interesting than the gliders and clocks of Life. (Look at Stephen Wolfram; he's found enough emergent stuff in one-dimensional cellular automata to make him look like a paranoid schizophrenic.) Part of what makes the great games great is the social phenomena associated with them; people talk about them, and study them, and motivate others to study them, and you get this snowball effect of attention massing on this topic, plumbing it as far down as it will go. I'd like to believe that certain games (and Conway's Life) got so much attention because they had the most depth, but who knows? Maybe Chinese Chess is just as deep as European Chess. Maybe those supposedly inferior versions of Go where some of the stones are laid in advance aren't actually inferior. It's hard to believe, but I've got no evidence Game A is deeper than Game B except hearsay, as I've rarely played Chinese Chess and I'm not good enough at Go to be able to tell if the version where the stones are initially placed is in fact more shallow.

Where am I going with this? I'm looking for excuses to consider 'game designer' a nurturable skill. Excuse A was: there might be ways to tell, or at least get a pretty good idea, on paper, if a rule set is going to have the possibility for depth. And Excuse B was: maybe most of the games that have the possibility for depth actually do have depth, and the main reason some games are considered more deep than others is predominantly a social phenomenon.

Excuse B is pretty lame. This is what I get for thinking on paper.

Excuse A has some potential. A game designer can only tell if a game has depth by playing it and playing it and playing it. If a dominant strategy emerges, that's when depth is capped. You could measure the depth of a game: how long does it take to find the dominant strategy? In games like Go and Chess, a dominant strategy, if there is one, has yet to be found. The best game designer will be able to detect the dominant strategy sooner. He might be able to tell, just by looking at the rules of a broken game, what the dominant strategy is, where a lesser man cannot.

Still, even the best game designer cannot tell whether a game has the kind of seemingly infinite depth that Go has or whether a dominant strategy is just around the corner, and just a few more hours of study will turn it up. Which is depressing. When asked "Does this game have the depth to become an eternal classic among the ranks of Chess or Go?" the game designer can only honestly answer either "No" or "I don't know."

Fortunately, in our industry, we're only expected to provide a few dozen hours of play before the dominant strategy is found. And that's easily within our grasp, even for non-genius game designers (like myself) who can't tell if a game is broken until they play it.