GDC09: Casting a Pod

***Coletta Factor: Spoilerish Discussion of Resident Evil 5 ensues***
So, the always-gracious Michael Abbot had me on his podcast last weekend to chat 'bout the GDC with some eminent bloggers-- Ben Fritz of Variety's the Cut Scene blog and Duncan Fyfe of Hit Self-Destruct. I don't exactly remember what I nattered on about into my USB rock band microphone (it was 11 AM on a Saturday and I was, naturally, quite drunk), but if for some unexplainable reason you'd like to hear it you can pick it up here.

One thing that came out of the conversation is that we all took very different things from the conference, though we were all people who write about games on the Internet. Ben's one of the very few really good industry reporters, so a lot of his time was devoted to interviewing publishers and publicists and gamesmakers-- hunting down the newsworthy. And Duncan talked about how the main business of the conference-- the panels and the awards-- weren't really useful to him given the way that he writes about games.

For me, the real benefit of an event like the GDC (aside from getting to meet all these great people from the Internet) was coming into contact with a new language. All of us games writers who hanker after a better critical discourse on games stand in need of more vocabulary-- if not a common set of concepts or a shared jargon, at least a common discourse that we can draw on when we talk about the kinds of irreducibly subjective things that games do to their players.

And it turns out that game developers are fellow partisans in this struggle. For the betterment of games, they've faced down the formidable task of explaining their practices to their fellows. They've salvaged elements of their craft from inarticulacy, because they need to explain to each other what makes a good level and what makes for satisfying combat mechanics and how to encourage cooperative play. All this is pretty downstream from the user-end experience of the game in motion, but my fond hope is that I can poach some of these ideas and use them to explain how and why games are fun.

The value of this language for the ordinary games-player is that it would allow you to see things you didn't see before. We can spill a lot of ink asking the function of criticism, but one thing that this secondhand enterprise can do is offer insight into how artworks function. You can go back to the same thing you've experienced and appreciate it in a different way.

This is one thing I mentioned on the podcast-- when I came home and played through Resident Evil 5 with my ladyfriend, I felt like a had a more expansive grasp of what the game was doing. Randy Smith's talk at the GDC was about the design of environmental puzzles, but when we ran into some crazy frustrating boss encounters later in the game his talk was the first thing on my mind.

Just like puzzles, your classic Zelda-style boss encounters in Resident Evil 5 require the player to exercise a new set of techniques. They require a different tack than the inexplicably-multiethnic African zombie mob. And this is why it's so important for the designer to provide the player with some tools to understand how that puzzle works-- what its moving parts are, how they operate, when the player is on the right track and when they're not. (My all-time shortcut for this idea is this character in Prince of Persia: Sands of Time who just screams admonishments at you as you navigate a complicated disc-sliding puzzle)

A late boss encounter really illustrated one of the main ideas from Smith's talk. To simplify, one of his central points is that the moving parts of a puzzle should have clear affordances-- that is, you should be able to understand how the elements of a puzzle can be used by looking at them. Like, if you need to sever a dragon's head by dropping a portcullis, that portcullis should be jagged and mean-looking as hell; the rope that's holding it up better look very severable.

It's a fundamental unclarity about affordance that had us stuck on some of the later boss battles. RE5 leans heavily on its context-sensitive button prompts to inform you about the environment-- whenever you're in the vicinity of something that can be used (pulled, pushed, operated, swung, cut, uppercutted), the X button appears at the bottom of the screen. That's how you find out something is usable.

The problem is, when you're faced by some homicidal ex-partner who's flipping around and unloading clips into you, getting some proximity is the last thing you want to do. Nothing signals to the player that this enemy can be used in a totally novel way when you're both at close range. We spent a lot of time hung up on the wrong solution-- shooting from a distance-- before we accidentally ended up at close range. And it was only then that the context-sensitive menus popped up and the game telegraphed the correct solution to us.

This basic issue recurs in a suite of late-game boss encounters-- these enemies have unique affordances that you need to know, but the only way you discover them is by approaching really close under select conditions and seeing the X button pop up at the bottom of the screen. This is bad puzzle design.

Anyways I could nerd out about Smith's talk at length-- it was strangely appropriate and fitting that a talk about how you teach things to players was a model of pedagogical clarity and insight-- but you can hear me nerd out on this very subject on the podcast.

Oh, and I was about to tell you what I asked Will Wright...