Wednesday 15 April 2009

What's the Story?

At this year's GDC, Jeff Kaplan of Blizzard talked about quest design, outlining nine problems with quests in WoW. Lum picked it up at Broken Toys (which is where I saw it first, though it's all over everywhere now), and now Timothy Burke has posted an interesting response at Terra Nova, which has prompted me to post my own thoughts on the subject.

Kaplan's problems with quest design are much the same as ours, which is to say, the same as most everyone who thinks about these things (but not, apparently WoW players). To summarise his nine points, I'm cribbing from the write-up at ShackNews:

1. The Christmas Tree Effect: quest hubs, where every quest-giving NPC has a shiny light over his or her head, and you can 'vacuum up' a lot of quests in one go. Players don't bother to read the quest text, they forget which NPC gave which quest and what order to do the quests in. Apparently, WoW players love this.

My experience of this is that it's the way WAR was implemented and the way the quest system in DAoC has evolved. It's horrible. It's overwhelming, distracting and immersion killing. The lights-above-heads work as attractors, but the goal becomes 'turn all the bloody lights off' rather than 'experience fun quest content'.

2. Too Long, Didn't Read: Quest texts are too wordy, and players don't read them.

True, but why? Well, firstly, and I'm not the first to suggest this, quest text is not universally well written. So you start off reading them, wanting to learn more about this new world you find yourself in... and after a wee while, you start thinking ho hum, I know as much about the lore as I actually care to, and listening to yet another NPC jabber on about his lost dog/sister/sword is way less interesting to me than getting the money and xp points I've been promised for this quest. Just tell me: How many foozles and where are they? Quests simply aren't entertaining enough. And it may be too late to make them entertaining, at least as long as they're presented in this same format. The playerbase has given up on them already and I doubt designers will be given another chance.

Secondly, and I haven't seen anyone talk about this anywhere, there's a procedural problem with quest text in an MMO. This is massively compounded by Christmas Tree quest hubs, but it's a problem in any game where a) the prevailing culture is one of acheivement rather than exploration, and b) players are grouped.

When I'm questing, I have three modes of play. One is solo, questing-for-fun. Under these circumstances, I'm questing because I've got nothing else to do, and I've got my Explorer hat on. I might read quest dialogue, if I think it's going to be interesting, and I'm not offered too much of it at once. Another is solo, questing-for-advancement. I've identified that a particular quest allows faster progress than farming and is repeatable, so I go farm the quest. (An example would be Dousing the Flames in DAoC / Albion.) I've got my Acheiver hat on, and I'm doing the same quest multiple times, so the last thing I want is to have to click through pages of dialogue, and the whole experience is mind-numbingly dull. Finally, there's questing-for-loot. Often, the good loot comes from the good quests; developers have put a lot of effort into these ones, knowing that the reward is an uber item lots of people will want. But they're usually tough quests that you need a group for. And the more of you there are, the less you feel able to take the time to read quest dialogue. The purpose of the group is always to bash through these things as quickly and efficiently as possible, because that's the prevailing culture. Sadly for the Explorer in me, I know of no game where this isn't the case.

So it seems to me that storytelling needs to be decoupled from rewards, so that the Acheivers can get their fix without being forced to wade through a lot of irrelevant narrative, and the Explorers can consume narrative content at their own pace, without being driven through the content at breakneck speed by (external or internal) reward-seekers. This could be done by only offering small rewards for story-heavy quests, or by putting all the texty stuff in books that the player can read later, if they want. Oh and make the damn things entertaining, please.

3. Medium Envy: Quest writers would rather be writing books or screenplays. "We need to stop writing a fucking book in our game, because nobody wants to read it."

This is the main point Burke addresses. I'd like to quote all of it, but this is turning into a long post already. I'll assume you're not a text-hata - Go and read it!

Essentially, it seems to me that Burke is saying you can't effectively crowbar controlled, linear narrative into a static multiplayer world, and we need a new paradigm. This is exactly the problem we're trying to address with Apocalypso. One of the commentators points out that EVE is an evolving world in which players generate their own narratives, though this player-history is not 'easy to see or well documented'. There's a sweet spot here, that has to involve presenting both designer-created and player-generated histories as an integrated and accessible part of the game, and I think I have some ideas on how to do that, but I'll save those for the design documents!

4. Mystery: Apparently quests that make you think are a bad thing. "We wanted the action... to come from the gameplay, not in figuring out 'What am I supposed to do?"

OK OK. Experience may show that within the WoW context, people hate the mystery quest. But see point #2 on quest text. You have to understand what's motivating the player at the point when you present them with this content. If someone just wants to get the Shiny Hat of Pwnage because they've heard it's the best Shiny Hat in the game, do they want to have to figure out which levers in the Hall of Confusion open the gate to the Dungeon of Muddle and then find their way through the Maze of the Lost? No. They'll look that stuff up online and curse your name the whole way through. Figure out which players like solving puzzles. Offer them puzzles to solve, at appropriate stages. Don't try and make the min/maxers do it too.

5. Poorly Paced Quest Chains: Don't give your players long quests spanning lots of levels that will suddenly get too hard. It pisses them off.

6. Gimmick Quests: I think what Kaplan's saying here is don't write something in that isn't really part of this game, because you wish you were working on GTA.

7. Bad Flow: Don't group your quests up so the player has lots of kill quests and then lots of collection quests etc. Seems like common sense.

8. Collection Quest Mistakes: Um, pitch the difficulty appropriately for the reward and don't ask people to collect more stuff than they can carry. (I'm paraphrasing massively here but there didn't seem to be more to it.)

9. Why Am I Collecting this Shit?: I.e., what does the Mayor of Questville want 18 gnoll paws for. And why doesn't every gnoll drop 2 paws?

Kaplan should have put this under point 8, and it's obvious, apart from the last question. Why doesn't every gnoll drop 2 gnoll paws? In a thread I was reading on Raph Koster's site recently, a reader suggested that in their ideal MMO, every wolf would drop all the things a wolf should have about their person, i.e. wolf teeth, wolf hide, wolf eyeballs, and not money, torn jerkins or toyboxes. You can extend this, and say that every bandit should drop what that bandit is wearing. It reminded me of how bizarre I found the DAoC loot system when I first started playing it. Horses with money, bandits with nothing but a muffin, when I could clearly see armor and a shield, not to mention the sword he'd been hitting me with just before I killed him.

I don't know how much extra load it would be for the servers, but wouldn't it be nice if loot all worked like it does in Morrowind , where the NPCs have exactly what you expect them to, and you can loot all of it?