Friday, January 23, 2004

Quality of Life

There's a "quality of life" survey at the IGDA. I filled it out, but it bothered me. It is not the best survey ever written. For one thing, there were too many multiple choice answers where the choices were all so specific I would have chosen "none of the above" if it was an option. For example, do I work at a big studio with 4 or more projects or a mid-sized studio with 2-3? I work at a big studio with 3. Better to have two questions: number of employees, and number of projects. Another thing: I checked the box for "my friends and family think I work too hard and want me to spend more time with them" but my wife would think that even if I worked a forty hour a week job. (She's going back to work in February, though, so there'll be some more slack in the near future.) And sure, I seem stressed out all the time. But that's just my pessimistic nature. And no, I don't think I'm paid enough. But who does? And no, I don't look happy. But I am. Really. And yes, everybody here is involved in the crunches, whether they want to be or not. I personally think that's a good thing. (I talk more about it in my latest Manager In A Strange Land article.) And yes, we're too god-damned ambitious, and it means we always end up cutting and crunching...but that's our own damn fault. It's not like upper management is breathing down our necks and saying, "More! More!" Basically, this survey is going to make us look more miserable than we really are.

Thursday, January 22, 2004

Do Me A Favor?

I was talking to Chance Glasco the other day - he's been doing some pick-up work for Treyarch on a really cool project TBA, and got some of the Call of Duty story from him. The real impressive part of the story is this:

- 23 developers

- 18 months

- 92% on

This alone would put Infinity Ward in the bang-for-buck hall of fame along with Factor 5 and Epic - but they're even more insane than Factor 5 and Epic, because Rogue Squadron II and Unreal Tournament were at least done with their own engines. Infinity Ward had to take the Return From Castle Wolfenstein engine, a totally unfamiliar engine, and run with it. (Of course they replaced the renderer, and many other things, which is a necessary technique for making a great game in a hurry, which I've mentioned a couple of times before.) I can't imagine adapting to someone else's engine that quickly. Also, although Call of Duty is a spiritual sequel to Medal of Honor, I think it shows more originality then the other two bang-for-buck top contenders.

So I'm going to say it. Infinity Ward in the highest bang-for-buck developer of our time. That I know about, anyway. (Although Neversoft is up there, too, there's less innovation in a Tony Hawk sequel than in this.)

Which means...they should win Rookie Studio of the year. Unfortunately, I already filled out my nominations, and I forgot to nominate them. I'm sure somebody else nominated them...but just in case nobody did...could you do it for me? There's only one day left...


In other news, I know my comment system totally sucks big piles of ass. I will switch to typepad one day.

Tuesday, January 20, 2004

Notes on Deus Ex: Invisible War

Somebody in the comments wanted to see my notes on Deus Ex. We aim to please. I'm going to be biased, though, because the Ion Storm guys are kind of my heroes...even though I never finished Deus Ex, I so admired it, and I admired how much they managed to deliver in a fairly short time frame with not that many people on the team. Hey, I think even they admit that the first Deus Ex was too long...

Choice. Choice. Choice. Deus Ex still owns the word choice. Even more so, this time around: now you have more control over the narrative, and more or less side with whatever faction you want. This may be the first game that really pulls this off. Although plenty of games have alternate endings, usually the endings are based on how many points you get (True Crime, if you count your cop karma as "points") or how well you play (Silent Hill). In Deus Ex, decisions you make affect the story. Decisions you make can even end up in the news. Talk about feeling like you're part of a world. Also, unlike the first Deus Ex, where you were railroaded into siding with the terrorist faction, this Deus Ex leaves it up to you. One of my big complaints about Neverwinter Nights: if I was really evil, then I'd join up with the enemy. Deus Ex makes it possible, in a way. (But what I'm really looking for is joining the enemy and then supplanting their leader and becoming master of evil.)

Another example of the Deus Ex freedom of choice: they have universal ammo. Your ammo magically works in any kind of gun, from a rocket launcher to a poison dart shooter. (Ahh, nanotechnology. Can explain anything.) This was a ballsy, scary move on their part, and I'm sure some of them lay awake at night wondering if it would work. One of the things that makes it easy to make sure your player has to use the entire repertoire of toys you've provided is to have limited ammo. Ratchet & Clank, for example, has two dozen guns and the thing that makes the player try all the guns is because they run out of ammo. But Deus Ex isn't about making the player do anything: it's the player's game and the player can play it however they want. On my first play through, I favored the sniper rifle. I also liked killing with the energy sword and then picking up the dead bodies and hurling them thirty meters into the air. Maybe that's not your cup of tea. But that's okay, you don't have to play that way.

I'm on my second play through now, seeing how far I can get without killing anyone, hard level this time. One great improvement over Deus Ex 1 is this one is shorter, so you actually can go back and experiment with different play styles, and different character advances. Choices are only meaningful if they have consequences: once you've chosen a character upgrade in a given slot, you're fairly committed. You can also upgrade your weapons; once you've made your choice, there's no going back. The shortness of the game allows you to actually "explore the game space" as Will Wright might say.

One of Harvey Smith's things is "orthogonal units". We see this in Deus Ex in a number of places: the enemies are well-differentiated, from the heavily armored Templar trooper--which explodes when killed--to the stealthy Illuminati soldier--who gives off a cloud of poisonous gas at death--to the robots (more vulnerable to EMP weapons than conventional). Color coded trip lasers--gas, alarm, death--different upgrades get you past the different kinds. (Although I usually used grenades.) The player verbs are also orthogonal: sneak, kill, hack, circumvent. Weapons in an FPS are usually orthogonal, and it's no surprise that this game sticks fairly close to the standard FPS model.

People bitched about the demo. This is no surprise, because Deus Ex has to grow on you. Those moments of emergence, when you lead one group of enemies to get killed by another group, or you discover that you can use the explosive crates that are lying around as weapons, or you lure guards away from your target and hide as they run past...because Ion Storm is letting those moments come naturally out of the system instead of crafting set pieces, those moments are rare and precious. It's only after you've had a handful of those moments that its hook is in you and you discover you can't put the game down. How do you demo that?

I played the PC version. Halo is the only tolerable FPS on a console. I wanted my mouse and keyboard interface. I have an okay machine...GeForce 4, 2.4 GHz P4, 512 megs...but it wasn't until I downloaded the patch and turned all the graphical thingies to minimum that I was happy with the framerate. I wonder about Ion Storm's choices, here. Although the beautiful lighting definitely has wow factor, I think players respond on a possibly unconscious level to good framerate. I'm living in a glass house here...once I've worked on a game that manages to maintain 30 FPS I'll have a right to complain.

I'm afraid that Deus Ex is one of those games that teases us and then doesn't meet our expectations. You throw a coffee pot at an NPC and he says, "Quit it." You're impressed. Then you throw a dead body at him and he doesn't seem to notice. It's things like this that made Tom Chick lambast the first one: because this game is a step ahead of the others in the AI department, we expect perfection. Give an inch and we take a mile. For example, it's never bothered me before when an RPG had a handful of irrelevant sidequests. "Great!" I said. "If I want to do a little extra exploring/playing, I have that option. And if I don't want to do it, nobody's holding a gun to my head. And it makes the world seem more rich and detailed!" But this time it bothered me. In the city of Seattle, there are a few sidequests that have almost nothing to do with the main storyline, and I ended up having to do them just to make the money I needed to advance the plot. (Mark Nau calls this, "You're saving the world and you're on a budget.") It's just not exciting to find proof of a nightclub owner's corruption when you're supposed to be getting to the bottom of a global conspiracy. (Actually, come to think of it, the same thing bothered me about Knights of The Old Republic but I hadn't put my finger on it until now.)

Maybe a trick from screenplay writing could help us out here. In the Hollywood formula screenplay, the subplot is tied to the main plot. Take Matrix, for example. The love story between Trinity and Neo is a subplot. The resolution of that love story is inextricably tied with the resolution of the main plot. I'd like to see more of this kind of plot construction in the side quests in RPG's, where they grow out of and tie into the main throughline.

Ending a game is tricky. Zelda has good endings. Metroid Prime had a good ending. Everything else in recent memory?...enh. Most games try to change up the gameplay at the end, give you something different to make it memorable. This leads to things like Xen from Half-Life, the race at the end of Halo, the 'shoot the tiny targets' endings from the two Max Paynes, and a collection of not terribly interesting bosses from many other games. So it seems like sacrificing your core gameplay at the end is a mistake. On the other hand - Deus Ex 2 sticks to its core gameplay - bringing us back to Liberty Island, the starting point of the first game, *awesome* closure, sense of unity, damn good storytelling - and...well...the gameplay feels anticlimactic. Maybe because it's more of the same...maybe because it's a "we ran out of time at the end and had to punt" deal...maybe because none of the choices you can make in the end are very satisfying? One of the endings I saw was depressing, the other -- and I think it was the 'good' ending -- creepy.

Tomo Moriwaki has a rule for final bosses: he says that the whole game is just training for the final boss. Metroid Prime was a great example of this, where your entire repertoire of moves is necessary at the end. But if Deus Ex went this route - in the final level, you had to use stealth AND shooting AND hacking AND so on - it would have violated the spirit of the game. Deus Ex is about OR. I don't know what the Ion Storm guys should have done here.

The first Deus Ex made me realize the value of a continuous world. (Although it wasn't the first to have one.) A little bit of jargon: I would define a "continuous world" as one where you're not artificially constrained. Games which have traditional levels, where you can't backtrack (You came in from this door but now it's locked? - we employed this device a couple of times in Spider-Man 1. Ew.) are not continuous. Games like Zelda, Deus Ex, Quake 2, and Half-Life are. I'd further define a "seamless world" as continuous worlds that have no visible load times: GTA3, Dungeon Siege, Prince Of Persia, Ultima VI, and Metroid Prime are all mostly seamless. The storyteller in us does not like continuous worlds! The storyteller wants to get to the action, and skip the boring walking around parts. But the game designer loves continuous worlds. The game designer wants to immerse the player and give them control. It is wonderful that Deus Ex is continuous. Hopefully the next Deus Ex will be seamless, because the load times were too long.

Oh, and I should mention that the music is excellent. Combined with the lighting, it feels like you're in Blade Runner.

I ended up bagging on the game a lot more than I intended to. It's easy to find fault but hard to put your finger on those intangibles that make you want to keep playing. I'm especially culpable because these 'notes on' articles I write aren't supposed to be pejorative but are supposed to point out cool interesting things about games. I should point out: this is a very good game. It's like Deus Ex but better. This is the first game in a long, long time that I'm playing again.

Monday, January 19, 2004

Speaking of Dvorak

Been using it for a month and still haven't broken 70 wpm, and my qwerty typing is ruined. I'm paralyzed at anybody else's machine. Which is okay: they learn more if I make them type. Still, so far, switching has been a spectacularly stupid idea. Oh, yeah, and everyone at work makes fun of me. I'll give it...two more months...and if I'm not over 100, I'll switch back.

Sunday, January 18, 2004

The Design of Everyday Videogames

Rich Bisso made me read Don Norman's *Design of Everyday Things* and I'm glad he did. There's a handful of stuff in here that applies to videogames. The book really isn't about design in the way I think of it: I think of design as being how to make something look good. When designing in the arty sense, function is irrelevant…that's probably why those graphic design samples have garbage text, so no actual information actually gets through. When a website has a pretty, well-composited layout, I think of it as being 'designy'.

What the book is about is usability. And for geeks like me, that's more important than design.

But rather than make you read the whole book, I'll cover the essentials right now, quicklike.

1) Affordances: a handle affords pulling. A knob affords turning. A plate affords pushing. Some ways this can apply to videogames: you have a gate operated by a switch. Make it so pushing the switch up raises the gate, and vice-versa. It applies to the controller as well. Which button should be the jump button, and which crouch? The button at top should be jump, and the button at the bottom should be crouch. Unfortunately, this violates another, more important usability rule:

2) Don't fuck with conventions: at this point, so many games have put their jump button in the same place that it is foolishness to violate the convention. There's a little demon in us that makes us want to be different for the sake of being different, or marginally better. Ignore that demon. Don't try to make people learn Dvorak when they're happy with Qwerty. A good example of doing this right is Simpson's Hit And Run. I could pick that game up, and my Grand Theft Auto skills translated perfectly. (On the other hand, although True Crime changed things around, it was pretty clear they had to in order to express the game's additional complexity.)

3) Knowledge exists in the world. Each key on a keyboard has a label saying what it does. Nintendo does this excellently with their games, often having an iconic representation of the controller right there on the screen. (The water nozzle in Mario Sunshine, the usable items from Zelda. The tutorial points of Prince of Persia work similarly, giving the player the knowledge they need when they need it. Having a controller layout accessible from the pause menu is a workable alternative, sacrificing usability for designyness.

4) People are intimidated by a large number of controls, but it's much better to have a control-per-function than to try and compress many functions into a single control. A particularly bad example would be to have a button do one thing when you tap it and a completely different thing when you hold it down. Peter Molyneux got this absolutely wrong when he went from Dungeon Keeper to Black & White. The large number of icons in Dungeon Keeper intimidated people, sure. So he concealed an even more complicated functionality in an invisible interface: less intimidating, yes. More usable, no. With console games, we frequently have more commands than buttons, so we're tempted to use chords and modes and double-taps to express the additional options. A good move is to put those extra options on the D-pad, like Splinter Cell does. And how do we avoid the intimidation factor? Make the new commands available slowly, after the player masters the ones you've already given them. (Side note: I've occasionally gotten into arguments when working on RPG's about whether we should make unavailable menu items greyed out or invisible. The answer: before they're introduced they should be invisible, but after the player knows what they are and is no longer intimidated, if the option becomes unavailable for whatever reason, then it should be greyed, which says to the player "This option isn't available right now, but may become available again at some point in the future." And if you really want to put some knowledge in the world, let the player choose that option anyway, at which point you can put up a dialog box explaining why the option isn't available.)

"Wait a minute, Jamie," you say. "What about the transition from Ultima V to Ultima VI? They went from having one control per function (each letter on the keyboard did a different thing) to a streamlined iconic interface where there were only a dozen or so controls! And it was clearly better!" They did an interesting thing here, where they abstracted several different commands (such as 'unlock' and 'open') into one abstract 'use' command that does the appropriate thing depending on context. This is another avenue you can take: instead of finding more buttons, reduce the number of verbs. Which is not without its drawbacks: suppose you abstract Punch and Kick into one Attack button, which does a punch if the enemy is close but a kick if he's far away? You've taken away some of the player's opportunity for choice. (But it's probably worth itÂ…in both Neversoft and our Spider-Man games, for example, the punch and kick were virtually identical, giving the player what they would eventually discover was a meaningless choice. Better to use one of those buttons for something else, something orthogonal. A defensive button, for example. Oh, wait! We're doing that!)

I think that's it. I'm not doing justice to the book but I do think I've hit some of the most important take-away lessons for game designers.