Happion Labs GameDevBlog

Loops

So, loops of varying lengths, as described by Cliff B.
Fire the gun;  watch the tracer;  wait for the recoil;  repeat.
A longer loop: After a certain number of bullets, you reload.
A longer loop: Picking up ammo every n people you kill.
A longer loop: Going to a "store" and upgrading.
A longer loop: Making it from one level to the next or one story point to the next.
A longer loop: Completing the game and waiting for the sequel.  (I added this one myself.)
It's weird, my thoughts were headed in this direction already...I was thinking about writing an article on the minigame, the macro-game, and the meta-game, and defining these things, when I realized that games are sets of deeply nested challenges and \defining which level of challenge is macro and which level of challenge is meta really isn't that important.
Your most atomic challenge would be can you hit the badguy?
Your next level of challenge:  can you kill the badguy?  Hit him enough times?
Your next level of challenge:  can you kill enough badguys to get the next powerup / stage / points?
Your next level:  can you earn enough points to buy an upgrade or level up?
Your next level:  can you finish the game?
Your next level:  can you finish the game on a higher difficulty level, or 100% the game?
Your next level:  can you beat your friends?  Get on a leader board? 
So, um, now that we've got "loops", how does that help us design games?  Well, one thing Cliff said was that loops should be "tight."  I know, intuitive.  But there it is - I think you can tell if there's too much slack in your loops - (in hindsight, there was too much slack in the "earn points to get to the next mission" loop in *Spider-Man 2* - a little more dev time would have solved that - and, in *Spider-Man 1*, there was too much slack in some of your combat moves - punch, punch, punch knocks the badguy down and you can't hit him again until he stands up.)  Another tool for our arsenal.

Cliff B points out that repetition is okay, if you're doing everything else right.  I agree.  I think games, like music, thrive on repetition.  You recognize patterns and recognize variations in the patterns.  This gives you a kind of pleasure.  A problem with games these days, though, for me, is that there isn't enough variation.  Killing that five hundredth not-a-zombie in Resident Evil 4 wasn't all that different from killing the first one.  Compare with chess - each game puts you into a new situation you may or may not be able to handle.  Hmm.  Not a good comparison.  Thinking out loud here.  If firing your gun at a non-zombie in RE4 is the most atomic loop in the game, then the most atomic loop in chess would be moving a piece and waiting for your opponent to move a piece.  Ok, let me amend.  Each *position* in chess is a new situation you may or may not be able to handle.  It may look deceptively like a pattern you know, but be different enough that the seemingly obvious move is completely wrong.  Much like the non-zombie may turn out to be one of those non-zombies that sprouts the whiplash tentacle thing out of its neck.  I wish I could make a melee combat system for a videogame that had the depth of chess.
But I probably can't.  So I am willing to accept the repetition.  In fact, if somebody tells me, "Your game sounds like it's too repetitive," I tell them, "It's repeated awesomeness!"
Boy my thinking is fuzzy this morning.  I better stop writing before I say something really, really stupid.

GDC day 2 (or 4)

I realized last night I forgot to really hit day 2.  Day 2 was my day for actually making it to a couple of game design talks.  I saw Marc LeBlanc in the morning, and Cliff Bleczinski (I wonder if I'm spelling that right?) in the evening.  Sort of a Yin / Yang of game design - Marc LeBlanc gave me the apparatus to differentiate them - he talked about intuition vs. intellect.  We do most of our work through intuition - I know fun when I see it! - but sometimes we switch to intellect to get out of a rough spot, like it's a low gear.  And when we give talks (or write in our blogs) we tend to be in the intellectual mode.
After explaining that difference, Marc gave a very intellectual talk.  This was the first time I was exposed to his Mechanics - Dynamics - Aesthetics framework;  it wasn't what I thought it was going to be.  In his framework, Mechanics is your entire game, including the dressing or color;  Dynamics is the experience of playing the game that a player has;  Aesthetics is the player's response to it.  Which means I'm fucked.  Just when I thought we could settle on Mechanics being the name for the formal abstract portion of the game - the "salad" as Hamumu put it - it turns out Marc LeBlanc's already got a different definition for the word.  I said to him, "So now we need another word for the part of the game that doesn't include the dressing or color."  He didn't seem to think so - he said that Whack-A-Mole and Whack-A-Baby are very different games, that dressing is more important than mechanics in that case.  I still think we need a word.  Rules, I guess, is what they call it in *Rules of Play*.
Okay, and then, in the evening, Cliff B., pretty much an anti-Marc, gave his talk.  Very, very intuitive.  "Graphics matter", "Firing a gun with a nice mesh (that has some kind of reflectance shader) and a pretty muzzle flash and a good sound and an obvious effect on the environment is fun", "Story is a motivator", "Game design isn't that hard (we know fun when we see it)", "Trust your instincts (using the goop gun from Unreal as an example)", and - this was the good fortune cookie of the bunch - "End the game when the kids are having the most fun." 
That last one is counterintuitive but obvious when you think about it - for example, in Half-Life 2, I was usually starting to get bored of one sequence before they switched to the next sequence;  if the different kinds of gameplay where marbleized together, so that I'm still having fun with the boat ("Hey, I wasn't done with that yet!") when they give me the car, I'd be motivated to play more hoping for more boat.
Cliff also talked about loops, which converges with something I've been thinking about lately.  More on that later, because Cathy wants me to be in some photos with Sofi now.

More GDC

I'm tempted to name-drop.  I've met a ton of cool people, a lot of whom are semi-heroes of mine, some of whom I've met online, some of whom I've never met.  I'll mostly resist the name-dropping thing, but I will tell this one story:  the guy sitting next to me at Will Wright's presentation, mentions how crowded it is.  We get to talking, he has a French accent, so I have trouble hearing his name.  Then we talk about what games he's done I could swear I hear "world" and "darkness" so I double-take on his name tag.  He's freaking Eric Chahi!  And he sat right next to me!  "Oh my God, you're Eric Chahi," I say.  And proceed to worship him.

Okay, another name:  Jake Simpson seems to be the center of the universe.  People I've only met online know him.  Friends from college I haven't seen in a decade know him.  Does anybody *not* know him?

And he had some stuff to say about the talk Tomo and I did, in answer to questions that Tomo and I didn't actually answer that well.

Question One:  what if you're starting from scratch with your code:  how do you prototype?

My flip aswer was "Don't do that."  Jake pointed out you can prototype in 2D, or text.  Good idea, if you're making a strategy game or a sim or Dungeon Keeper or whatnot.  Wouldn't work for an action game where 3D is king (Spider-Man or Tony Hawk), but it'll work for a lot.

And Jake tested us by asking, "What if you need to prototype a system that's dependent on a lot of subsystems?"

I said, "Then it'll take a lot of time.  Months."  Tomo said, "You're SOL."

Jake later mentioned that you can stub out the subsystems, and although you can't test if the system will be fun, you can at least test if it works or not.

Okay, other highlights:  Paul Du Bois (more names!  I can't stop!) took Chuck Tolman and me to see the Double Fine offices and to play a release candidate build of Psychonauts.  Very promising - still can't wait for it to come out.  The art direction is insane!  It's like a cubist Mario on acid.

Will Wright, in his presentation today, demoed the game he's been working on since, I think, Sims 1.  A friend of mine violated his NDA by telling me that Will Wright was working on Sim Everything.  At the time I was like, well, whatever.  I didn't know what that was or why that would be fun.  Well, now I've seen a demo, and I think I got more pleasure just watching the demo than I got from all the games I played last year.  I'm tempted to quit Treyarch and send EA my resume, provided I could get assigned to that project.

And my powers of prediction are awesome!  Greg Costikyan ranted at the Game Designer's Rant, and what did he say about Allard's keynote?  "I don't know about you, but it made my skin crawl."  His rant made me feel a little guilty about continuing to suck at the Spider-Man teat.  Of the other ranters, Warren Spector was pretty down to Earth (let's stop being Wal*Mart's bitch); Brenda Laurel was like, from some other dimension or something ("videogames are a non-consensual relationship between middle aged men and young boys"); Jason Della Rocca took everybody to task (game developers are blind, xenophobic fools - more on the "fool" part in a bit);  and Chris Hecker had an amusing point to make about the processors in the next generation - for gameplay code, they're going to be slower than the current generation of processors.

Listen, I loved the rant;  they were right on about a lot of things;  but there is some room for argument.  Yes, publishers are getting more and more risk-adverse.  But Activision let us put this wacky swinging system into Spider-Man 2...and EA is letting Will Wright make Spore.  Jason Della Rocca was saying that Steve McConnell was implying that we're fools, becuase software developers who follow best practices shouldn't have to crunch anymore.  Now, I've read and loved McConnell's book.  It's my Bible when it comes to development.  But it hasn't saved me or my team from crunching, though, as hard as we try.  And okay, gameplay code will be just as slow as ever.  The thing is, gameplay code isn't the main burden on our CPUs.  If the CPUs get faster at processing graphics, then we'll have more bandwidth for our gameplay code.  (This point didn't occur to me until later in the evening;  I talked with Chris [last name drop, I promise] a bit after his rant, and I would have asked him about it instead of swapping new father stories if I had thought about it.)

A friend of mine called the GDC a "fucking waste of time" before I went up.  All I can say is, okay, maybe I didn't learn all that much while I was there, but I don't see how meeting, in person, all these people I respect and admire is a waste of time.  Plus I got to see a ton of old friends.  So yeah, that.

Oh, and thanks to everybody who came by my talks afterward and introduced themselves.  Good to meet y'all.  And thanks to everybody who gave props to *Spider-Man 2* or *Die By The Sword*.  I live for that shit.

GDC was awesome.  I'm still buzzing.

GDC Day 1 (or Day 3, depending on how you count)

The last time I went to GDC they served dinner, and had a keynote speech by the editor of Wired.  The crowd got rowdy, people get upset.  It seemed a younger generation of console developers had moved in on the PC developers and were acting like children.  Although the editor of Wired was not a good speaker, I have to admit.

The keynote this time was targeted at that younger generation.  There was plenty to be cynical about in Allard's thing.  I suppose I could leave that for Greg Costikyan.  I will say that when they decided to give away TV's to the crowd I was like, "What the fuck is this?  Oprah?"  (If I had won one of the TV's I wouldn't have turned it down, though...)  And I also have to ask the question, how is a higher resolution TV a paradigm shift?  The answer:  it's not, but it sure was clever to be able to talk about the shift from "3D" to "HD", because they both have two letters and end in D.  Okay, I'll leave the rest to Greg.  No, wait, I've got to say one more thing:  I'm skeptical about Allard's claims Re: XNA studio.  I'd love to have better management software but it's not going to make our management problems go away.  And is the company that brought us Project really the company that's going to give us less painful management software?

Ok, now I'll shut up.  If Allard ever sees that I may have just doomed the team to a real tough time passing TRC two years down the road...

Let's see, what else?  Missed an interview because the Irish pub on the corner took a half hour to get me a hamburger. 

The highlight today was the Game Design Challenge: The Emily Dickinson License.  Both Clint Hocking and Will Wright were damn funny and had cool ideas.  It looked to me like Peter Molyneux strapped an Emily Dickinson license onto an existing engine, though...Clint may have deserved to win, since he came up with an actual game-game, whereas Peter and Will both came up with software toys. 

Tomo and my talk went well, I think.  Good sized crowd, perceptive questions after.  One guy actually wanted to see the slides again later, so I'll have to post them up here somehow, sometime.

I can't say I've learned much today, but it has been a blast.

GDC

So it's GDC week.  I'm going, for the first time in years and years, and Tomo and I are finally giving the talk that we postponed last year due to crunch:

The Early Months of Spider-Man 2: Using Preproduction to Invent Gameplay

That's Wednesday at 4 PM.

Also, Noel Llopis invited me to join in on his middleware round table, Thursday at 2:30 PM.  And I just realized his website isn't in my bloglines.  And bloglines is down right now.  Arg.

This will be the first time I've spoken in front of a group since college.  Or do team meetings count?

So, anyhow, if anybody wants to hang out at GDC, shoot me an e-mail, or stop by after one of those sessions.

Something I realized today - GameDevBlog is one year old - as of last month.  GameDevBlog!  I forgot your birthday!  I'm sorry!

This means I've been blogging for three years now. 

Notes on Escape Velocity: Nova

Boy, am I dumb.

I could have gotten some good sleep last night - my daughter was sleeping for a change - but instead I found myself playing EV: Nova from her bedtime at 8 all the way to 3 AM.

I didn't even get into the combat system!  I spent the whole time exploring, trading, asteroid mining, and trading.

Normally I'm not terribly susceptible to RPGish grinding - maybe it had something to do with the rate at which trading could earn me cash (I found a run where I could make a 50% profit every time) - exponentially increasing your money sure feels good.  Of course, the ships get exponentially more expensive, so it's all illusion.

Like Uplink, the game is dotted or laced with tiny story missions.  "Like raisins in pudding," as Erik Bethke said once.

I'm going back in for more, now.

Notes on Fritz 8

Since the baby came, I usually don't get a straight uninterrupted hour in which to play a game of online chess.  And online chess doesn't have a "Save Game" option or a pause feature.  So I needed a computer program that would play me.  Sounds like no problem, right?  Chess AI is practically solved.

Well, it is a problem.  Even the dumbest of chess programs, as long as they don't purposely blunder, have a rating at least in the 1600s, which is much better than me.  The trick is to find a chess program that feels like a human.  But for the most part, when I play chess programs set on "easy" levels, it feels like this:

I'm getting my ass kicked - I'm getting my ass kicked - I'm getting my ass kicked - the program just made an incredibly stupid blunder, like hanging a queen.  I win.

That's no fun.  It's torment followed by a hollow victory.

Then I discovered Fritz 8.  Fritz 8 has a "friend mode" where it automatically adjusts to your play level (handicapping itself binary-search wise).  At the low levels, it no longer relies on its extensive opening book;  and its blunders are subtle - it opens itself up to attacks, it gives away pawns.  The kind of thing a player about my level would do.  My only complaint would be that it sometimes makes moves that are clearly pointless - like moving a piece and then moving it right back where it was.  It's rare that a human at my level does that - we're too stubborn.

And now I have my save-game and pause.  And I have an indicator of how well I'm doing - currently, my Friend Mode Handicap is 300.  My goal is to get that number down.

On a side note, I do still manage to get in an online game now and then.  I notice that I keep playing against the same people - these people have played hundreds of games and they're still at the same level.  I have a couple of theories:  1) we're all at a plateau that nothing short of dedicating our lives to the game will break us out of.  2) the quality of play, across the board, is getting better.  Imagine there were only two chess players in the world, and they used the Eco rating system.  As they played each other, they would continually get better.  But they're both getting better, so their ratings would stay the same. 

The things I tell myself to feel good about myself.

Oh, I almost forget.  Fritz 8 also has the coolest game analysis feature I've seen.  Trying to decode the output from my previous chess program was murder.  Fritz 8 will have output like "Prevents intrusion on h4", "White prepares the advance of a5", "Doesn't get the cat off the tree", "Does not help much", "and Black can already relax."  And classic chess notation (+/-, +=, etc) for who ends up in the better position.