Feedback’s a bitch

lolcatsdoingitwrong.jpg
Previously on Spectre Collie, I made the claim that the “state of the art” in story-driven games still treat storytelling and gameplay as two completely separate things. In even the best games, the player’s role is that of the guy who pushes buttons and/or shoots guys in the way; the key story moments happen in between my actions, not as a result of my actions. As a result, storytelling becomes increasingly superfluous and marginalized. Worse than that, because my actions don’t really have any bearing on the story, they lose any overall context, which is the whole purpose of having a framing story in the first place.

I went on to claim that instead of treating traditional adventure games as if they were an evolutionary dead end, we should be looking at how those games work to make the storytelling and the gameplay the same thing. The player’s not doing stuff and then watching a cutscene, and the player’s not doing stuff while listening to conveniently-placed audio logs that provide color commentary tangentially related to what he’s doing. Instead, his role becomes that of collaborative game designer. He’s not just doing what the objectives screen says to do, but figuring out what needs to be done in the first place.

The problem with that, says the internet, is that everybody knows that adventure games aren’t fun. They’re exercises in frustration, where you’re stuck trying random combinations of items until you stumble onto the one completely illogical solution the designers happened to choose. That’s the last thing that action games should be doing, you hear, because that’s the exact reason adventure games died out (and good riddance!). And even if they wanted to, action games couldn’t adopt that type of puzzle-solving, because they’re inherently different:

Myth 9: In an action game, all the activity takes place inside the game, but adventure games are solved in the player’s mind.

That’s a very rough paraphrase from Manveer Heir’s “Design Lesson 101” article about Sam & Max Season Two. Mr. Heir is a game designer at Raven who’s been writing fairly in-depth analyses of games on his own blog and Gamasutra. Now, since that article is about a game I worked on, anything I write in response is going to come across as reactionary and defensive. So I should make one thing clear up front: everything Mr. Heir says is a filthy lie, and his posts should be immediately stricken from the internet.

No but seriously: his main observation is dead-on accurate. Adventure games suck at giving feedback to the player. That’s true of every adventure game I’ve played, even my favorites. I think it’s the biggest obstacle for the current generation of games to overcome — not just adventure games, but any game that wants to engage the player on a level more interesting than “cross the finish line to get the next scripted event” or “press the button to trigger the next cutscene.”

There’s a lot I don’t agree with in Heir’s article — in particular, the idea that the adventure game model only works in small doses (in the “golden age” of these games, they were designed to last for several weeks); or that the episodes of Sam & Max Season One were far too easy because he was able to finish in two to three hours (that’s just slightly shorter than how long they were intended to last). But most of the criticisms are common to adventure games, and I would’ve probably agreed with them completely just a couple of years ago — in fact, the start-and-stop pacing of adventure games that Heir mentions was my biggest complaint when I started working in adventure games again.

But I think his observation that it all comes down to player feedback is insightful and extremely useful, for two reasons. The first reason is that it’s not just about difficulty — you can’t just say “Season One was way too easy, but Season Two got too hard in places,” because then you’re trying to hit a moving target, and it varies wildly from player to player. Earlier, I made the case that trying to find some objective measure of difficulty for logic-based puzzles is doomed to failure.

The second reason is that player feedback is something that’s common to all types of games, from so-called “cerebral” adventures to so-called “twitch” shooters. In any genre of game, we’ve got the same basic set of tools, but we’ve gotten so locked into our assumptions about certain game genres, that we’ve forgotten how to use them.

I can’t use these things together.

Go back to one of the most common complaint about adventure games: “I don’t like adventure games because I always end up having to use every item with every other item until I stumble on the one ‘right’ answer.”

Whenever I hear that complaint, my first response is: “Well, don’t do that.” The appeal of these games is in figuring out how to solve a problem and predicting what turn the story is going to take. If you don’t have an idea already in mind when you use the magnifying glass on the dynamite fuse, then why would you even try to do it?

And of course, that’s a lousy response. If you were to ask me what are the most memorable lines from adventure games I’ve played, the first ones that would come to mind are: “That doesn’t need to be painted white,” “I’m not putting my mouth on that,” and of course, “I can’t use these things together.” They’re not memorable because they’re particularly clever, but because I heard them over and over and over again. Because that’s how these games — all games — are played: you do stuff and see what happens. It’s the “interactive” part of “interactive fiction.”

We tend to have this silly idea of adventure games as being “thinking man’s games,” where the ideal player is the cliche of the guy who solves the New York Times Crossword Puzzle in pen. We watch a cutscene, then sit back in our easy chairs and mull over possibilities, then shout “Eureka!” and complete the puzzle. Not only is that insufferably pompous, it’s unrealistic and frankly, not very fun. You want to get in there and poke around and explore.

That’s yet another way Super Mario 64 gets it right: the first thing the game does is drop you into a playground and invites you to just run around and play. Climb trees, swim for a little bit, and learn how things work. It’s also one of the best design aspects of The Sims 2: most of the development time and creativity in those games goes into the failure states, the stuff that you wouldn’t see if you played everything the “right” way. Because learning how the game works is one of the most engaging parts of the game.

Type “HELP” for a list of commands.

There are several reasons why adventure games don’t do negative feedback well:

  • More stuff: In a shooter or platformer, the list of ways you can interact with the world is deliberately kept small. In adventure games, you tend to have a lot of different items at your disposal. It’s easier to cover every possibility of what happens when you shoot something or jump on something, than it is to keep track of dozens of objects and their interrelated uses.
  • Larger “possibility space”: Not only do you have more stuff in an adventure game, you’re usually expected to use it in an unconventional way. And as mentioned earlier, each player has his own idea of “unconventional.” What seems like a perfectly natural solution to a problem to one player might never have occurred to the game designers.
  • More dialogue: Gordon Freeman never speaks, so it’s not jarring when he tries to unlock a door and the game just beeps. It’d be really weird if Guybrush Threepwood tried to unlock a door in the game and didn’t say anything when it failed.
  • Managing difficulty: How do you respond when someone’s done the wrong thing? Do you just say “No?” Or do you say, “That’s a good idea, but wrong.?” Or do you say, “That won’t work, but using the magnifying glass might?” Or “That won’t work, but I bet I could use this magnifying glass to light that dynamite fuse?”

The thing to notice about those reasons is that none of them are unique to adventure games, and none are insoluble. (The closest to being damning is actually the “more dialogue” issue, since that depends on purely practical concerns like production time, voice recording budgets, and download size).

The one that sounds the most damning is the idea of a larger possibility space: how can a game give intelligent feedback when it’s impossible to gauge what the player’s thinking, how close he is to solving a puzzle? That’s the converse of the player’s frustration with adventure games, that it’s impossible to gauge what the designer was thinking when he came up with this stupid puzzle. To me, that doesn’t sound like the death knell of a game genre, but just a simple communication breakdown.

You can use these things together! Ask me how!

Although the hint system in Telltale’s games is almost universally regarded as A Good Thing — and I should point out here that not only was it not my idea, but I was actually against it at first and was proven wrong — it’s still a first step. It demonstrates that it’s not impossible to tell what the player’s thinking; we know the solutions to the puzzles, we know how people play adventure games because we play them ourselves, and it’s actually relatively straightforward to detect when the player’s stuck and what kind of information he’ll need to get moving again.

The issue with that, again as pointed out in Heir’s article, is that players sometimes stubbornly refuse to listen to hints because they feel like it’s “cheating.” The perception — which is unfortunate, but probably unavoidable — is still that the developers have the one “right” answer, and they’re guiding the player through the game, nudging him in the right direction when he’s too dense to figure it out.

I think a logical next step is to take the relationship with the player away from “we’ve got the answers, now you figure them out” and back to that idea of collaborative game design. In a game design session, nobody’s figured out the right answer yet. So the dialogue is one of “well, no, that won’t work for this reason… but what if we tried this other thing?” It’s less like a tutorial, and more like exploration and experimentation.

And the difference between that and something like a strategy game or The Sims is that there is still one right answer. It remains a conversation, instead of a toy or a playset. The developers go through the effort and frustration of coming up with a game that’s guaranteed to have some sort of satisfying resolution, instead of just giving the player a bunch of tools and then removing themselves from the equation. A big chunk of adventure game development is spent just figuring out valid alternative solutions to puzzles and then either implementing them, or explaining to the player why they won’t work in this case. It’s not that much of a stretch to extend that to an overall design philosophy: remembering that we’re not always trying to funnel players towards the one right answer, but keeping an open dialogue with them, guiding them towards the answer that we happen to have generated cutscenes for.

It’s my hope that this would extend past adventure games. If we can figure out how to engage the player in the actual storytelling without getting hopelessly stuck, then we can have more engaging stories in any type of game. And developers of more action-oriented games can include more significant story moments without being so terrified that they’ll frustrate the player and ruin the pacing of their game. And then I can finally play a game that forces me to think about what I’m doing, instead of just telling me to go here and press this button.