Spector is a fierce advocate of including multiple solutions to any given problem in his games, and I think he puts forward a fascinating case:
Earlier this month, we talked to Valve Software about Half-Life 2: Episode One and all manner of things to do with game development. At one stage, Gabe Newell said that Valve wanted to "get people through as much entertainment as possible".
"This is an argument I have with Warren Spector," he said. "He builds a game that you can play through six different times. So that means that people pay for the game, but don't get to play five sixths of the game, which I feel is a mistake."
Eurogamer: There's a certain irony that you're working with Valve when it's clear that Gabe Newell has an almost polar opposite design philosophy to yours. As you know, he firmly believes that gamers should get to experience "as much entertainment as possible", but that making something where players may miss five sixths of the game is a "mistake". He says, "You spend all this time to build stuff most players will never ever see." That's a pretty wholesale rejection of your company's mantra of allowing players to craft their own unique experiences through in-game choices, don't you think? Who's right here?
Warren Spector: Well, I'M right, of course! No, seriously, there's clearly room for a variety of approaches to game design - god, it'd be boring if we all believed the same stuff and made the same kind of game! Fact is, I've been having this argument with Gabe for, oh, let's see... how long have I known him? Nine years? Yeah, all that time. And before that, I used to argue with Richard Garriott and others at Origin about the same damn thing. I'll go to my grave believing I'm right.
But, really, I find the idea that one design philosophy is "right" and another "wrong" (or even that one is better than another) incredibly odd. I mean, is Star Wars better than The Godfather? Is Lord of the Rings better than Goodfellas? Should Stephen Spielberg make nothing but action-adventure movies because they make more money than his more "serious" efforts? Should we elevate Tom Clancy or Dan Brown to the top of the writing heap and stop reading Shakespeare? Does anyone think all music should be aimed at the top of the Billboard charts? I sure don't want to live in a world where everyone sounds like Britney Spears... oh, wait, I already do... Anyway, you get my point...
It's interesting that Valve's Half-Life series (of which I am a huge fan) is referenced in drawing an almost direct inversion to Spector's design philosophy. For those heathens who never played Half-Life 2, the game is profoundly scripted; you're pushed through scenarios that unfold identically every time. The player's only responsibility is to tick the boxes given to him: step 1, step 2, step 3, all the way to step 50,000 in the finale. There's very rarely more than one or two approaches to any given problem. Interestingly, the recently-released HL2: Episode One expansion features a new commentary mode where someone explained that, in one particular case, where Valve had deliberately included an alternative solution to a problem, those playtesters who exploited it didn't grasp that it was an intentional facet of the game; they thought it was an 'exploit', a piece of lazy game design. The guy said, more or less, that "where we wanted to make the player feel smart for finding an alternative route to success, they just thought we hadn't done our job properly".Eurogamer: Is it really a waste of development time to give player choice, and how do you persuade the player to come back and play things a different way? Rather than miss out on five sixths of the content, tempt them to play the game six times...
Warren Spector: Wow, lots of points to address here - this is going to take some words...
First, and most trivial, I've never said that players should see one-sixth of your content. My "rule" has always been that every player should see about 75 per cent of your content, with another 25 per cent reserved for unique player experience. That's kind of a dopey measure, in a sense, because it implies that the best way to differentiate player experience is to handcraft a lot of paths through a map and a bunch of branching dialogue for NPCs to spout.
There are other ways to get at unique experience that don't require massive amounts of hand-crafted content. But I do believe that generating some content, knowing everyone won't see it, has huge value.
For players, a multipath/multisolution game offers the knowledge that if they're clever they will see and do things no one else has ever seen or done. How can you not want to play a game like that? A year after we shipped Deus Ex, I saw someone solve a particular game problem in a way I'd never seen anyone try before, and I was sitting there with him wondering if his solution would work. I mean, I helped make the game, and I'd played through that part of the game a hundred times and watched probably a thousand playthroughs and I was seeing something I'd never seen before. No game-on-rails or rollercoaster ride can possibly touch that for a thrill!
And check out the forums where people talk about how they solved a particular problem and others respond in amazement that they'd never thought to try that approach. Listen to people debate what one endgame choice said about you as a person, as opposed to what another endgame would have said...
That is so much cooler than listening to people agree how cool it was when they all killed some monster in exactly the same way, or got across some chasm in exactly the same way, or solved some goofy puzzle in exactly the same way.
Beyond that, multipath/multisolution games offer players who aren't great at combat, say, another option (stealth, dialogue, hacking - whatever).
They can keep playing your game instead of throwing a controller or mouse across the room in frustration. I mean, not to pick on Half-Life 2 (which I happen to enjoy immensely), but if I'm not good enough to get past an enemy or a carefully crafted puzzle my only option is to stop playing, and maybe never buy another Half-Life game. In Deus Ex (for another example I enjoy immensely!) if you can't fight your way past a problem, try something else.
Something else will work. (There were no puzzles in DX, so I can't address that - we only had problems, and problems, by their very nature, are open to solution in a variety of ways.) Anyway, giving players ways to keep playing your game seems like a good idea to me. How is that a waste of time and money?
So, to try to wrap this up in something less than book length, I don't believe it's ever a waste of time to give players real choices, rather than fake ones. If all you're doing is putting players on rails and rocketing them through your story, why not just build a roller coaster or make a movie? If the only choice a player gets to make is which weapon to use to kill a bad guy, you've completely wasted that player's time. Roller coaster rides are immense amounts of fun, but really, all they do is provide an adrenalin rush and a moment's distraction from the workaday world. Games can be more. Movies are terrific storytelling devices - I love movies - but movies already exist. I don't need to make them. (Well, I kinda want to produce a movie someday, but that's another matter entirely...) If all you're doing is telling yourstory to players (with the added attraction of getting to pick a gun once in a while), why bother?
If you're not "wasting" development time by allowing players to explore your world, advance the plot their way, test behaviours and see the consequences, I believe you're wasting players' time - and that's a far, far worse sin than wasting some time and dollars on stuff some players might not see.
So what's everyone's take on this? I've got a few half-baked opinions myself-- as I inevitably do-- but I'm going to wait until they reach total fruition before I deploy them. My initial reaction is that the two design approaches, linear and non-linear, offer altogether separate advantages and that neither one can be deemed necessarily 'better' outside of personal preference. But I see Spector's argument. I'm still at war with myself over the use of cut scenes in games; is it not the video game's duty to fulfil the potential unique to its medium-- that is, player interactivity-- rather than echo the techniques that can be found in superior capacities in film? It strikes me that this is a similar argument, because Spector seems to be suggesting that in offering the player a myriad of possibilities, rather than to take Valve's strictly linear approach, the developer can more efficiently capitalise on their medium's unique talents.
Where I think Newell kind of trips up is his fear of the player missing out on a sizeable portion of the game. It occurs to me that if the player's ability to miss portions of the experience-- to make decisions at the cost of other possibilities-- ultimately improves their experience, it may well be worthwhile. But HL2's scripted progression offers its own undeniable benefits; I've never played a game whose balletic precision offers such a breathtaking sense of purpose, direction and, perhaps most of all, place. Such things are perhaps lost with a more open-ended design. Deus Ex's much-derided sequel, Invisible War, for example, took that philosophy too far; its design accommodated player choice to such an extent that no significant decision, in terms of (for example) whether or not to ally with one organisation or another, ultimately impacted anything at all, since in order to meaningfully expand on that decision would effectively necessitate the production of ten different games. If you allow the player's journey to be too open, then there simply aren't enough boundaries to contain him and you end up having to make their 'decisions' completely illusory.
Well?