GameDevBlog

You Fiend

Tim Schafer almost tricked me into pre-ordering Psychonauts even though it doesn't come out until April.

I don't like pre-ordering.  It seems like a good way to guarantee that all your friends have the game before you do.

Also, I can't add *double fine action news* (or *beard beard action beard*, as they're calling it these days) to Bloglines.  This is bad.

Quality Sells?

Oh yes, let's not forget theory three:  *Wow* sold well because it's a really good game.  Coming in at 93% on gamerankings, there's even something of a semi-objective measure there...it's the most highly rated MMO of all time. 

I'd love to believe that's all there is to it...but articles like this leave me little hope.  And then there are games like Ico: high ratings, poor sales...and then there's Spider-Man 1, considered marginally above average by the critics, one of the top ten sellers of all time.  Mark Nau once ran a correlation between gameranking scores and sales, and although there is a correlation (yay!), it's terribly faint.

Still, quality helps, and if you're the best in your class, it probably helps a lot. 

One trap I'm falling into is the "what's so special about X" trap:  frequently, when a game is succesful, we try to analzye it, and look for the one factor that made it such a success.  When Myst was succesful, it was "This kind of graphic adventure must be hot!"  When Tomb Raider was succesful, it was "You have to have a hot chick as your main character!"  When Grand Theft Auto was succesful, it was "You have to let the player be a bad guy!" or "You have to have a free-roaming environment!"  (The industry still hasn't gotten over the "Why is GTA so succesful?" phase yet, and we probably won't until GTA gets unseated by something bigger.)  The problem is, when you're dealing with these insanely popular games, these games that redefine our ideas of what good game sales are, you can't boil it down to just one factor.  It's usually a combination of factors, and together that combination is crazy.  Combo!

Boy Was I Wrong

Way back when I predicted that *World of Warcraft* would fail.  Probably nobody remembers.  And Google failed to index that page of my old blog, so I could just keep mum, and even if somebody did remember would they be able to find the link?  No.  But here I am, coming clean:  way down at the bottom of the page, here http://gamedevleague.blogspot.com/2003_09_28_gamedevleague_archive.html, I say that *World of Warcraft* will weaken the *Warcraft* brand, because people think *Warcraft* means RTS.

So:  how come an *Everquest* RTS tanks but a *Warcraft* MMO goes ballistic?

Two theories.

One:  *Warcraft* doesn't mean RTS to people, it means games with cool art.

Two:  there are places where line extensions work.  An obvious example is politics.  Actors have incredible mindshare and are pretty much unstoppable forces when they turn to politics.  Even though a lot of people say "I'm not going to vote for an actor!  What do they know about politics?"  so many people say "I haven't even heard of this other guy" that the actor wins.  On the other hand, could a politician go into acting?

Likewise, millions of people play Warcraft.  Even if 4 out of 5 of these people say, "I'm not interested in an MMO by these RTS people," the other 1 out of 5 means 500,000 people, which is huge by MMO standards.  But only 400,000 people play Everquest.  If 4 out of 5 of these people say, "I'm not interested in an RTS by these MMO people," then the Everquest RTS does under 100,000 units.

I like that theory.

In other news, I've been catching up reading other people's blogs, and I can't believe I never came across this one before.  He even quotes me!  There's an ego boost.

And finally, *Spider-Man 2* was nominated for "Outstanding Achievement in Gameplay Engineering" by the AIAS.  I am pleased as hell by this.  That's a much, much cooler nomination than the "Best Voice Over" we got from SPIKE TV;  the category is cooler, the organization giving it is cooler, and it's kind of what we were all about - we were really focused on a new kind of game mechanic, made possible by some clever engineering.  Thanks AIAS!

Various Stuff

Good to hear from Raph in the comments!  By the way, anybody who hasn't checked out his slideshow on small-world-networks and gaming should:  http://www.legendmud.org/raph/gaming/index.html, go to Essays and scroll down to the one on small worlds.  It gave me a new lens with which to view the world.  Sorry if I misinterpreted you, Raph, but didn't you say something along the lines of us having to discard dressing if we truly wanted to approach game development as an art?

Speaking of terminology:  I'm told that board game guys have been using the word "color" for what Raph was calling dressing since the beginning of time.  And, as for the "salad" - I always used to call it "gameplay."  That's what I said in *Difficult Questions About Videogames*, and a few others seemed to agree with me.  Others, however, think of "gameplay" not as the formal abstract core systems, but as an activity.  The act of engaging with the game. Down to the feeling of the controller in your hands.  At first I thought those people were just wrong.  But then I decided we should throw "gameplay" out as an overused term.  (One of the guys in *Difficult Questions* - I forget who, I don't have the book on me right now - said that we all use the term, thinking everybody knows what we're talking about, but we all have our own definitions.  He was right on the money.)  Instead, how about "game mechanics" as a term for the salad?

One other thing about save-anywhere:  dozens of hugely popular games don't have it.  Let's take an irresponsible look at the Best Games Evar from gamerankings:  Zelda - nope.  MGS - I forget.  I don't think so.     Metroid Prime - nope.  Resident Evil - nope.  Halo - nope.  Grand Theft Auto - nope.  Half-Life 2 is, in fact, the first game on the list that does have save anywhere.  Now I'm not one to say that just because all these great, massively popular games don't do it, it must be wrong, but it can't be the enormous issue that some of you have made it out to be in the comments section.

Speaking of irresponsible looks at gamerankings, I was just wondering:  how important is variety in games?  Some games are basically the same thing over and over (driving games) whereas others strive to give you mini-games and the like to change it up.  My first instinct when thinking about having mini-games is to say NO!  If people wanted to play Battleship they'd play Battleship;  they wouldn't play the Battleship minigame in Wind Waker.  If they wanted a rhythm game they'd play Parappa or DDR or Amplitude rather than the rhythm missions from San Andreas.  If they wanted a stealth game they'd play Splinter Cell instead of the stealth missions from True Crime.  And so on.  But maybe I'm wrong about people.  Maybe your mass market game consumer only wants to buy a few games, and he wants those games to be swiss-army knives:  you can get your racing, shooting, stealth, rhythm, dating-simulation, extreme sports tricks, role-playing, and territory conquering strategy all from *San Andreas* - why spend more money on other games just because they do those subgames better?  Plus, from a production point of view, Erik Bethke's argument for a lean, tight feature set - so those features can be high-quality, high-polish - is pretty compelling.

So I'm looking at top games on gamerankings and asking myself: How many sub-games do these games contain?  What constitutes a sub-game is tough to define (each boss fight in some of these games could almost be considered its own mini-game), so your mileage may vary.

Zelda, Ocarina of Time:  fighting, puzzle-solving, nonlinear exploration, horseback riding, stealth, boss fights.  6+

MGS:  stealth, shooting, boss fights, nonlinear exploration.  4

Metroid Prime:  shooting, jumping, nonlinear exploration.  3  (Correction:  boss fights, ball-modes, both side-scrolling and straight ahead.  6.  Good point.)

Halo:  shooting, driving.  2   (or does flying / tank driving count as a separate game?)

Half-Life 2:  shooting, driving, ant-lion/squad herding, "Don't walk on the sand" stealth, physics-puzzle solving, automated turret placement.  6.  (Does the gravity gun count as a separate game?)

San Andreas:   I counted eight up until my PS2 threw a rod the other day.  I'm not sure if I want to buy another PS2 just to finish this game...

Wind waker:  Ocarina of Time plus some more mini-games and sailing.  8+

Pro Skater 2:  Tricks.  Collecting.  2.  (Side note:  the highest rated of the Tony Hawk games has the fewest sub-games:  Tony Hawk 1 had racing, and then Tony Hawk 3 and on started adding new kinds of challenges.  Not that this anecdote proves anything.)

Half-Life:  Shooting.  Xen.  The "get past the tentacles" puzzle.  The final boss fight.  4.

Vice City:  Driving, shooting, rail-shooting, remote control vehicles, buying property.  5.

Prince of Persia:  Fighting, navigating terrain.  2

Knights of the Old Republic:  Fighting, nonlinear exploration, leveling up, talking, space combat, racing, a couple mini-games here and there.  8+

So...my new stance...a game needs to contain at least two games to be popular, and more sure doesn't hurt.  And if those games aren't polished (San Andreas's aren't, although they're well tuned) it's not the end of the world, either.

Notes on A Theory Of Fun

I mentioned yesterday how I'd just read this.  Will Wright has an introduction to the book where he says that he doesn't like the title but despite that it's an okay book.  There's an important word in the title, and that's "A".  If this book was called "The Theory of Fun" or, simply, "Fun", I would have probably thrown it away in disgust when Raph gets to his definition of what makes a good game.

Trying to define what makes a good game is a fruitless task, and yet so many people try to do it.  All you can do is define what makes a good game for some people at some time.

Still, Raph Koster admits that it's just "a" theory, and when viewed in that light, it is a useful lens for looking at games.  I've always known that games have an abstract, formal core and dressing on top;  the kind of chess pieces you use to play chess do not affect the underlying game, and the look of your character in a fighting game does not affect the underlying game.  But I've never dwelled on the fact.  One of Raph's most important points, that he keeps coming back to, is this fact, and after I read the book, I did find that I was looking at games in a new light.  For example, in my mind, a fighting game is no longer a fighting game, but a game of realtime tactical rock-paper-scissors with a whole bunch of different kinds of rocks, papers, and scissors and a handful of other bells and whistles thrown in as well.

Raph discounts the "dressing" as just "dressing", which is a strange stance to take for someone so heavily into role-playing.  Story and simulation are both parts of this dressing, after all, and role-playing wouldn't exist without them.

Actually, I'll go out on a limb and say any game wouldn't exist without them.  Thought experiment:  you can get closer to the formal abstract game part of your game if you take the names of your units and replace them with significance-free letters and numbers.  Instead of moving pawns, knights, and bishops you're moving A Units, B Units, and C Units.  Instead of playing a fighter or thief you're playing Class A or Class B.  The dressing is so deeply marbleized into gaming I'm not sure it could exist without it.  I like the feeling of a wooden chess piece in my hand, I like to see my character do a gory finishing move in Mortal Kombat, I'm not just increasing a stat, I'm earning experience.  Etcetera, etcetera.  Not to mention:  why did I find *Alpha Centauri* boring but not *Civilization* - they're the same game, but one has dressing that appeals to me and the other does not.  Oh yes:  when *Grand Theft Auto* went into three dimensions it's still essentially the same game as *Grand Theft Auto II* - at the formal, abstract level - but the dressing changed.  And that made all the difference.

Just yesterday, as we were meeting about the new combat system for our yet-to-be-announced project, I suggested (probably because I'd just read this book), that we could get started on working on it even though our animator is busy on another project.  "We could use our old animations in new places, just to see if the basic system works."  Tomo and Mark disagreed:  the animations are such an integral part of how the combat system plays and feels that it wouldn't be a valid or useful test.  (I still think it would be a valid and useful test of the underlying game mechanic--we could check for exploits, see how deep it is, look at the learning curve, for example--although I agree that it would not be a valid test of how people will respond to the system.)

Raph Koster does admit that dressing is important for mass market games.  Which is what I make for a living, so maybe his book--a bit like Chris Crawford's--just isn't for me?

Okay, so enough dissing on his book.  It really is valuable to separate which parts of your game are game and which parts are dressing.  Think of all the game design rules you either explicitly or implicitly apply as you design:  how many of those rules apply to the dressing and how many to the game?

Here are some of my game design guidelines:

"Make The Player Feel Like A Hero":  dressing, although a well-tuned game that always teaches the player something when he's ready for it can at least make the player feel clever.  "Have a Sniper Mode":  although the mechanics are formal and could apply to a camera game, the reason it's fun is because you shoot somebody in the eye, right?  "Find The Emotional Heart":  dressing.  "Don't Simulate a Simulation":  dressing.  "Make things behave as you'd expect them to in the real world":  dressing.  "Immerse the player/ suspend their disbelief":  dressing.

So, while I have a lot of guidelines that apply to the abstract game-game-type-game as well:  "meaningful choices", "keep it simple", "perceivable consequences", "intermittent schedule of reinforcement", it seems like most of the time I'm thinking about the dressing rather than the underlying game, and it may be high time for me to change that.

Save Anywhere

Over on damnedmachines they have a question for me.  The answer goes something like this.  In general, I'm not supposed to talk about Spider-Man 2 without going through Activision PR, but I think this is innocuous enough...

The main reason there used to be no save-anywhere on consoles (and Spider-Man 2 doesn't really have save-anywhere;  it's like Zelda, although you can save anywhere, when you restore your game, you find you're at the beginning of the mission you saved in the middle of) was technical;  there wasn't enough space on the memory card.  It's also a major challenge to code save-anywhere, so now that there is enough space on your typical memory card, most of us are stuck with engines that don't support the feature.  Rewriting the Spider-Man 2 engine so it could handle save-anywhere would take a lot of effort, and would seriously impact the way all the designers built their missions...a steep price to pay for something that might, as Amit puts it, ruin the gaming experience.  If all things were equal, though, I'd want save anywhere.  If people want to ruin their own gaming experiences that sometimes should be their prerogative;  much better than having someone quit your game altogether because their last save was too far back.

Still, Amit is right that some people will ruin their own game experience with abuse of quicksave, cheats, or other exploits.  Once someone said to me, "I don't get what everyone thinks is so great about Doom.  I dunno, maybe I shouldn't play it on God mode."  Which reminds me:  I read Raph Koster's Theory of Fun last week, and one of my favorite parts was where he talks about players tendency to bottom-feed in various ways, to find exploits which suck the very fun out of the game.  This tendency of people is yet another thing that makes game development difficult.  (I had other thoughts reading the book, a handful of stuff I'd like to argue about, maybe I'll write about it later.)

I've also met people who don't like a high density of checkpoints.  These people grew up on console games, and, as they see it, part of the challenge should be to get through a whole level, a whole serious of challenges, without needing to save.  The videogames of today are for pussies, in their opinion.  They scoff at save anywhere.

So you can't win.

Notes on X-Com

Mark Nau couldn't believe that I hadn't played X-Com.  He started wandering around the office asking people, "Can you believe Jamie hasn't played X-Com?"  They couldn't.

I remember giving it a try when it first came out;  it did seem like the sort of thing I would like.  I don't remember why I gave it up so quickly.  I do remember other people in the office - these were my days at MindCraft - totally absorbed in it.  Why didn't it get its hooks into me back then?  I have some theories:

* I wasn't interested in the resource management half of the game.

* It was too hard:  most of your squad tends to get wiped out quickly in the early missions when you don't know what you're doing, even on the beginner levels.

* I didn't bother reading the manual:  my philosophy usually is that if you have to read the manual the game is flawed.  And the X-Com manual has a tutorial that's fairly necessary to get up and running.

* Bad luck?  Maybe I got randomly screwed by the early random missions.

Well, whatever the reason, I've been hearing enough about X-Com (the whole Mark Nau thing, the comment in my blog a couple weeks ago, Tom Henderson reminiscing about the wonderful balance between the macro game and the micro) that I finally took the effort to figure out how to make it work on a modern machine.  I'd tried before and failed.  Googling X-Com finally led me to the information I needed to make it happen:  http://www.xcomufo.com/

Well, it has probably been some of the best gaming pleasure I've had all year. And I mean 2004.  That's right, the best game for me in 2004 was originally published in 1993.  The curmudgeons are right, games aren't actually getting any better.

One of the things that interests me the most about is the random content.  Since I just finished a game that had a lot of semi-random content, it's interesting to me to consider why the random content worked for X-Com but didn't work for Spider-Man 2.  There are a couple factors, I think:

* The macro game carries the micro game.  Rumor (http://www.gamerankings.com/itemrankings/launchreview.asp?reviewid=155318) is that Julian Gollup is really only into the tactical portion of X-Com, that the macro game was MicroProse's intervention, and that explains why Laser Squad Alpha is just the tactical game.  But it seems like Julian Gollup might be missing the boat here:  without the resource management portion, where you're building your bases and squads and moving down the tech tree, the micro game would become boring sooner.  Despite the fact that resource management doesn't sound fun, it is fun.  It gets to the point where missions are a distraction taking you away from building up your dream team.

* Intermittent schedule of reinforcement.  The reinforcing stimuli, in this case, is a new mission.  It doesn't take much for a mission to feel 'new' : some terrain you haven't seen before, an opponent you haven't seen before.  I played probably around a hundred missions before I won.  Most of them felt like repeats.  But I kept going because every now and then there was a mission that was entirely new:  my base being invaded, for example.

* Light at the end of the tunnel.  After a point, I would have stopped playing except I knew that there was going to be one final mission to win the game.

I scanned through GameRankings looking for games with random content in the top two hundred.  They're incredibly rare.  In fact, I'm not sure there are any.  The Diablos aren't in the top.  There's an immediate problem with random content:  some of your players are going to get randomly screwed - a mediocre experience will be randomly generated for them.  If a reviewer is one of those players it'll pull down your rating.  The main advantage of random content is you can create a longer game with a smaller team.  If you're making a game that isn't meant to be disposable -- play for ten-twenty hours and put away -- it's a very tempting option.  X-Com might have been more lucrative if it was sold on a subscription model.  There are some hardcore fans who might have kept it alive to this day.

Addendum

Yes, to be fair, although Quake 2 had the best AI we'd seen to date, it was a marginal improvement over what had come before, and when Half-Life came out, the bar had been raised to a whole new level.

Someone noticed that I hadn't said anything bad about HL2.  I'm trying harder than I used to to avoid dissing on games - it's kind of lame for somebody in the industry to dis on other people's games in a public forum, for one thing, and it's too easy, for another.  Get a couple game geeks together talking about a game and the first thing they start pointing out are its flaws.  We just can't help ourselves.