Sunday, May 24, 2009

Systems with limits

"Sim City's a bit too loosey-goosey for me"

I mentioned that in the last post I wrote, but I wanted to explore it a bit more, so I will.

See, when you build a system, you build it to do something. In a situation where you're building a giant war robot, it's pretty clear what it does: it kills other giant war robots. Most of the fundamentals of the system are usually precreated for you - your robot can walk around, target things, communicate, produce energy, and so on. Your part of the deal is to create a system on top of that for addressing the particulars of the challenge and how you want to approach it. Equip it with lasers or missiles, make it nimble and lightly armed or a heavy... whatever your approach calls for and your resources can manage.

The fun comes in taking your system and applying it to the challenge. Your skill at making a system combines with your skill at playing the game. If you imagine it topologically, it's like the game world is a bumpy surface, you're building another bumpy surface, and then trying to mesh them together.

Anyhow, my comments on Banjo & Kazooie: Nuts & Bolts. It's the same idea, but more so. You're building the whole structure from the ground up. Where a mech game provides you with the basic giant robot frame, N&B asks you to build the basic frame as well, making the play focus more on basic movement than the weapons output.

To me, the fact that the system faces a challenge is an important part of the game, because it gives me a feedback loop. I build the system to accomplish a goal and, based on how well it works, I build the next system a bit different. That kind of iteration really sells me on the game.

Sim City doesn't really do that. It lets you build a system, but the primary challenge that system faces is to simply get bigger and more complex. Contrast that with Evil Genius or Dungeon Keeper, where the challenge is to resist enemy assault. Once you've done that, you start again on a new base.

So, in many ways, Dungeon Keeper is more like Nuts & Bolts than Sim City, because it involves creating iterative systems that face various challenges, while Sim City does not. To me, this makes them fundamentally different genres.

The other thing to keep in mind is that when the basic movement is not taken for granted, some of the gameplay may be more difficult to plan out to as advanced a level. It's far wider in scope, so the narrow missions and rewards you would normally hand out are easy to upset and do not cover the real scope of the game.

I wonder how best to deal with that?

Does it make sense? What are your impressions?

Saturday, May 23, 2009

The Incredible Machine...

... and a bird with a wrench.

Recently, I've been playing Nuts and Bolts, the most recent Banjo-Kazooie game. The previous games aren't really my style, being platformers. This one isn't a platformer. Its a build-your-own-vehicle game. It's pretty much right up my alley.

One of my friends feels the opposite - he likes platformers and not build-your-own-vehiclers. So I'm not making any judgments here, just talking about me. It's what I do best.

Build-your-own-vehicle games are an interesting and fairly rare phenomenon. It doesn't surprise me that I like them a lot, because they are basically the part of RPGs I like refined down into a pure form.

Even when the customization is fairly weak, such as in the various giant robot games, I still love it. Building a working system is great. I also, unsurprisingly, love all the various simulation games such as The Incredible Machine, Evil Genius, Sim City, etc. Sim City's a bit too loosey-goosey for me... but I'm falling off the subject here.

I'd like to talk to you a bit about the way this Banjo & Kazooie game handles the progression of play. Don't worry, there aren't any important spoilers. I'm not sure it's possible to spoil the game, as it's a farce with no plot twists or reveals.

The developers decided to start the game off really slowly. It was a good thing I picked it up during a slow week, because if I had tried getting into it on a normal schedule, I would have given up: it takes five hours before the game gets interesting.

Those five hours are spent slowly accumulating very generic parts. Basically, you get wheels and engines and some guns. This really limits the kinds of things you can do, and it's actually deadly dull. I imagine if I started over knowing what I know now, I could probably power through it in two hours, but still...

It's important to note that in the hub world (the "real" game world) you are forced to use a trolley cart, a very basic vehicle, so navigating in that world is extremely dull. In the mission worlds you can use whatever vehicle you want, so it's a bit more interesting, but not a lot when your self-made vehicles are basically just souped up trolleys.

At some point, things slowly pick up. You get the ability to fly, which is a lot of fun, not only in challenges but just in general flying around the mission worlds. Your trolley gains new abilities, allowing you to reach new locations, and exploring the hub world becomes fun and interesting if you're the sort who likes to explore. Which I am. Which is good, because as build-your-owns go, it still wasn't much fun: it's just that the exploration became a bit more fun.

The build-your-own options keep increasing, and it's gotten to the point where it's a lot of fun to build custom vehicles for whatever mission I'm attacking. It's a lot of fun to invent weird approaches, such as using the sticky grapple for air races, or building a glider with no engines or power source.

But should it have taken thirty hours to get here?

I'm all for the gradual curve: give the player more and more options over time. However, that gradual curve can backfire when someone comes in with a lot of skill inherited from other, similar games. It's important to provide a ramp for unskilled players, but it's also important not to apply brakes to skilled ones.

Actually, although I say that, I'm not sure it was the wrong decision. The play is not as deep as I would like, and I spend most of my time trying to figure out ways to do interesting things that the designers never intended (or, at least, don't bother to mention or require). I'm not sure the game would actually be thirty hours long if they let me accrue a lot of interesting parts right near the beginning, because I'm running dry on the building challenges included in the game and beginning to invent my own.

I keep wanting some things that aren't there, like a big ocean so boats are vaguely useful, or a 45-degree angle piece, or the ability to map things to directional presses and buttons in a far more carefully-specified way.

I guess what I really want is a lego-style universe to build and explore. Like Boom Blox, except without the pointless blowing-stuff-up-all-the-time parts and a lot more exploration. Ideally a MMORPG, so I can explore other people's worlds...

Anyhow, those are my thoughts on the matter. How about you?

Tuesday, May 19, 2009

The Limits of Mice

One of the things I've been playing around with more and more recently is different input types. I've used touchscreens a lot recently, but I've also been toying around with XBox controllers, augmented reality gear, and accelerometers. Nothing worth talking about in particular, but it highlights something I really want to talk about: what you can do is largely limited by how you can do it.

For example, one of the things I frequently want to do is adjust some other element of my interface temporarily. Right now, I want to turn down my music so I can focus on this paragraph a bit better. I did so, but it involved stopping writing, going over to the mouse, running the mouse over to a window on the other screen, and turning it down. In other words, it was distracting. Ideally, I could have just glanced over, or made a quick hand gesture.

This is just one of a thousand tiny details I wish I could tweak. It's a little hard to see it now because we can't do it: it's like someone using visicalc and trying to imagine the kinds of features a million times more powerful hardware will allow.

Trying to imagine the kind of work environment we might get out of more aware computers is a similarly difficult task. Let's do some thought experiments.

If you've read this far, take a moment to try to imagine a future computer. It can read you like a book. What kind of interface features does it have?



Come up with anything?

One of the big features we'll see almost immediately is some kind of "broad vs precise" gesture set. The ability to affect environmental programs without having to hunt them down is already becoming very common. Keyboards and laptops are often equipped with volume controls, even play and stop buttons. My keyboard actually has more than a dozen buttons I've never used ever on any keyboard, and that's in addition to the useless OLD buttons I never bothered using.

But the addition of an audio environment is a fairly recent one. The idea that a computer would constantly play music as you worked is relatively new and weird, so it's only now that we realize we want to be able to control it without interrupting whatever we're actually doing.

There are a lot of other features that we might see added to this environmental backbone.

For example, I can imagine my other monitor having a pastiche of imagery popping along it. The idea would be twofold. 1) to give me "microbreaks" and 2) to produce a kind of visual "white noise" to cushion me against any actual visual distractions. Such a program could easily integrate various alerts and email notifications, and I can imagine a lot of really interesting low-impact ways to handle it (a Twitter map would be fun).

However, this kind of thing is only good if the computer can tell fairly well when you've gone into "work mode", and even what kind of very precise mood you're in at the moment. The "sideline" stuff would be highly irritating and even distracting if it assumed you were in a mode you weren't in. This could be easily controlled by reading eye movements, even if they couldn't read precisely. But this kind of system can't be controlled using a mouse or keyboard: it needs to be largely controlled by a passive system, something the user doesn't need to activate.

It may sound like this sort of thing would be distracting. But... that's not really the case. It's very hard to see how much something would get used until it's being used. There have been many famous quotes by experts that radically underestimate adoption rates - a certain amount of RAM should be enough for everyone, or someday there will be a phone in every city in America...

Imagine further. Today I'm reading a textbook on investing. Sometimes the book is very dense. Imagine that a delicately configured program can tell not only what text I'm reading, but my comprehension level (by reading my pupil dilation, skin tension, or similar). As I read, some kind of halo follows my center of attention around. When I hesitate or backtrack, it highlights related parts of the sentence and runs through it again.

Wow, that sounds really distracting!

Your children won't think so...

But none of this is possible until we get something more than a mouse and keyboard!

On First Person Shooters

I stumbled across this today: Doomed to Invent Our Mistakes

The most interesting thing about it, to me, is that the comments by readers are spectacularly bad, short-sighted, and petty. It makes me glad I've never been noticed by those kinds of armchair experts.

The basic idea of Saltsman's essay is that modern shooters have focused on a few specific bits of gameplay to the point where they leave off the parts that are fun. In particular, Saltsman dislikes pixel-precise aiming, hiding, reload minigames, and strafing. He prefers those kinds of things to be looser and more forgiving so that we can focus on the parts of the game that are fun.

I largely agree. As games have gone ever more hardcore, we see them becoming very skill-based, but the skills in question do not sit well with me. I'm not a big fan of the reaction-speed skills that most modern FPS games focus on. I don't think all games should ignore those skills, but I would like to see some that focus on other things.

Because I prefer tactical, strategic, and exploration challenges. And I don't play FPS games much any more because they don't bother thinking in those terms any more. You can argue that, say, Gears of War or Halo have strategic or tactical elements, but the truth is that they take back seat to the ability to aim at a tiny, rapidly moving block of pixels while simultaneously strafing at high speeds. Even if you are a tactical god, you're not going to win if you don't have teenager-on-Mountain-Dew reflexes.

I never did, even when I was a teenager on Mountain Dew. But I could hold my own in the FPS games that existed back then, because they had a much larger tactical component. You could think your way through Quake deathmatches.

Anyhow, I agree with Saltsman, and I think that majority of the replies are by people who are so self-absorbed that they react violently to anyone who doesn't have the exact same preferences as they do. Obviously, many FPS fans are going to like FPS games the way they are... that's why they're FPS fans. But some of us stopped being FPS fans for the same reason.

Friday, May 01, 2009

Future of Diseases

So, I've watched this... thing... unfold, watched people vacillate between "OMG THE WORLD IS GOING TO DIE" and "it's completely unimportant". To be honest, I tend towards the latter. But I thought I'd use it to illustrate something about the future.

Biohackers are starting to exist - DIY Bio is not common, but it's going to be, and that's got a lot of people worried. At the risk of using too many snarky capitals, they say "OMG A GEEK IS GOING TO MAKE ANTHRAX 2.0 AND KILL EVERYONE FROM HIS BASEMENT!!!"

Let's use this Swine Flu as an example of how things will really happen. Let's say that this outbreak happened ten years from now instead of now.

In that case, nine years and ten months from now, the biohackers would post about a new strain of flu worming its way across Mexico, putting little flags all over the publicly available "world disease map" they've put together.

Nine years and eleven months from now, the biohackers would point out that it's breaking out in the city. They would also post the genetics of the beastie.

Nine years and 11.5 months from now, Mexico's government would actually notice, and would act just as it became a potential epidemic, rather than a few days after its become an epidemic.


Now why would the biohackers notice?

People seem to think that the DIY Bio movement revolves around manufacturing life forms with fun new genetics in your basement. No doubt there will be some of that. But that's not what we see in other, similar communities.

You see, what amateurs really excel at is collecting a billion tiny, stupid details and cataloging them. The vast majority of people who are involved in DIY Bio won't be doing any significant experiments: they'll just be swabbing bits of city or collecting samples from catches they've set up. Then making a big list of the kinds of things they're finding.

The kind of equipment, expertise, and time required to do that is only a tiny fraction of what is required to actually perform experiments. Hell, you can already collect samples on swabs and send them off to be analyzed for less than ten bucks. There are some very promising technologies coming up that will likely make this kind of basic analysis something that can be done by using a machine the size of a toaster.

The biological health of everywhere is very interesting, and it will become easier and easier for people to collect and share that information, allowing us to create a kind of "weather map" for diseases.


Another huge advantage a modern amateur has is being extremely loud. Classically (even today), a professional will tell a reporter something, the reporter will write it down, and it will get posted somewhere. Usually, the reporter will get it wrong, and the professional has no way to or interest in fixing the error. And even if there is no error, it still quickly becomes obsolete.

On the other hand, as we can clearly see from modern catastrophe response, hundreds of thousands of volunteers will self-assemble to organize the facts. Whatever gets ascertained is analyzed and distributed at maximum speed in a maximum number of useful formats with a minimum amount of error.

Imagine if a biohacker had analyzed Swine Flu. It wouldn't be some two-paragraph little note somewhere: there would be dozens of armchair biologists (or professionals working on their off-time) talking about the specifics of what it means, and running simple little experiments to test other theories.

So if this outbreak had happened in ten years rather than now, it never would have happened at all: the loud, detail-obsessed amateurs would have noticed it and dissected it far in advance of it actually denting Mexico City.


People become obsessed about the idea that biohackers are going to genetically engineer some kind of horrible genetic monster. That's not really a danger.

It's a CONCERN, but not a DANGER. There's a difference. The difference is that a concern is a danger that's being dealt with.

I have no doubt that hundreds of rules and practices will arise around the biohacker community. These rules will probably be mostly to address these theoretical dangers. But the point is that what safeguards need to be built will be built. People outside the realm, people who know next to nothing about biology, shouldn't try to dictate these safeguards. Because they would be, at best, pointless. More likely, they'd force the community underground, at which point the real safeguards wouldn't get put in place.

It's important not to imagine only darkness, especially when you don't know any goddamn thing about what you're imagining. A biohacking community will actually radically reduce the risk of a pandemic, not increase it.