Tuesday, November 2, 2010


Many decisions made in the process of creating a game will be between two equally valid, but irreconcilable, viewpoints. It's interesting watching ebb and flow of those types of opinions in the industry at large. Or quite simply, a whole lot has changed since 1992 and 2010.

Quinns over at RPS wrote a pseudo-retrospective on KGB, an adventure game from '92. I haven't played KGB (although after reading this post, I want to track it down), but what stands out most in Quinn's piece is how content the game is to mislead you or at least provide you with very incomplete information. Back in '92, you could still fail at an adventure game and apparently, you can fail quite often in KGB. Except it might be twenty or thirty minutes before you're fully aware you've failed.

This is something that you'd never see in a contemporary game. An unrecoverable fail state that doesn't resolve for twenty to thirty minutes? Most would criticize that as poor design, and there's some truth to that. But at the same time, Quinns describes what has become an utterly alien sensation in games- not knowing the outcome of your decisions.

The reason we moved away from consequences like this is pretty obvious. It's very frustrating to make a decision, damn yourself and not realize it for half an hour. And it can feel very unfair if a game offers you a choice, but something completely unexpected happens that makes the decision feel meaningless. Even though we're making decisions through an avatar, the decisions still feel personal. If a character in a novel or film makes a decision with incomplete information only to realize the full consequences later, it's less likely to feel frustrating. Or even if the audience has information the character doesn't, it still doesn't feel unusual. But if a game does the same, I think many would feel slighted, like the game was withholding information that should have been provided, regardless of fictional context.

Even inside of a single game series we can see this change in thinking. I've been playing quite a bit of Fallout: New Vegas recently and while I'm definitely liking it, I can't help but notice how explicit all of the decisions are. Many of the decisions are still compelling, but you're seemingly always fully informed of the consequences (almost to a fourth wall breaking degree at times). Whatever the NPCs say will play out almost always will. You know if you choose A, you'll receive X and this group will be mad at you, but if you choose B, you'll get Y and a different group will be mad at you. The outcomes of your decisions are all very well telegraphed.

On the one hand, you never feel like you would have made a different decision in hindsight (because since you can reload and make a different choice, this will encourage probing same & reload behaviour which really diminishes the emotional impact of any particular decision). But you're never really surprised either. Now Fallout: NV does a good job of providing surprises in other ways, so it's not exactly a deficit, but it's certainly not how making decisions in real life works. Decisions in real life are agonizing because you rarely have complete information about the possible outcomes.

This stands in stark contrast to the first Fallout game, where it was possible to lose the game completely. Not in the "I ran out of HP, died and had to reload" sense, but in the complete and utter "you can do nothing to change this, this is The End" sense. And there were actually two different ways this could happen. I guess some spoilers follow, but come on, Fallout came out 13 years ago. You should have played it by now.

One you're aware of at the beginning of the game and is totally explicit. Your Vault has 150 days of water left and if that counter goes to 0, you lose the game. But even after you fix the water chip, another invisible counter will begin later. After a certain number of days, Super Mutants will find your Vault and kill everyone inside. I have no idea if the timeline is always fixed or if other things can influence it, but once its course is set, it cannot be altered. And the game does not tell you when this is going to occur or what it will even occur at all.

I actually had one game of Fallout (not my first playthrough, thankfully) become completely unwinnable. I had two save slots that I'd alternate between and had made some expedition to The Glow. One save was made going in, the other was coming out. Maybe one or two days travel out from The Glow, I hit the game-ending date. Even reloading my original save, I couldn't get back to civilization in time to change anything. Without any other save slots, the only thing I could do is start the entire game over again.

Imagine a game shipping with such a feature today. It would be lambasted! The thinking of both creators and player is that players should be provided with nearly or fully complete information and allowed to make a well informed decision. And while I'm amenable to this thinking and believe it's good for players to have more agency, I can't help but read Quinn's piece on KGB and think that "I'd love to make a decision and be surprised by the outcome." I'm not exactly sure how one could adopt that sensibility into a modern game, but I think it could be done. It wouldn't be through failing half an hour after the decision was made, of course. But the feeling of tension and suspense in making those kinds of decisions can be very satisfying. We just need to figure out how to remove some of the frustration without removing the unknown. Do that and I think one would have something very compelling.

Labels: , , ,


Blogger thequickbrownfox said...

I think there is a sense in which games are less emergent than they once were. With increasing focus on developer authorship, players are led by the hand and the routes they take are more contrived. The outcomes of "important" decisions in most modern games do not emerge from the rules and properties of the world. They are simply among several fixed paths conceived by the developer.

I think you're right that this is partly an attempt to become more accessible and less frustrating. But aside from that and the tendency towards authorship, another large part of the reason for this is to do with technology working *against* us. Our hardware has the ability to render massively detailed environments and assets, so devs create such assets, but then they feel the need to restrict the possible combinations of events so that the player never sees anything "go wrong". But things going wrong is also an indication and an inevitable consequence of a game that has a rich possibility space. If events are emerging that surprise the developers then the players will definitely be surprised! And as you point out, surprise is very important.

Examples of being limited by technology/high-production values:

1. Graphics: If you're restricted to tiles/blocks you can quite easily randomly generate environments and allow players alter them on the fly with weapons/tools. These things increase the possibility space (See Minecraft). You can also do this in the high-tech Red Faction but the possibilities are much more limited and therefore less interesting.

2. Voice acting: it massively restricts the number of possible NPC speech, and removes the possibility of generating speech dynamically.

3. Suppose you create an open world, you will need to restrict where an enemy can be to a certain space. If you had smaller closed off-levels, you might be free to let enemies wander through them randomly without breaking the entire design. There might be a small chance that an enemy ends up somewhere the player really didn't expect them: SURPRISE!

November 3, 2010 at 5:20 AM  
Blogger Christof said...


What you're saying brings to mind the excellent analysis of Metro 2033 vs. Bioshock Christopher Thursten did.

In there is the crucial sentence:

It’s a game that could only have come from an eastern European development culture whose holy grail is absolute simulation as opposed to absolute cinematic integrity, where implementing ideas that a player may never experience is not a failure but the defining characteristic of interactive media.

It's one of the reasons why games like S.T.A.L.K.E.R.feel so different, so much more life-like: They do not shun away of confronting you with the unknown - even though it's not so much about holding back information as it is about throwing you into a system which is so dynamic and complex that even the creators don't know everything about it.

It's... about freedom, and the infinite space of possibilities, I guess. One of which it is to fuck up, but that's a small price to pay.

November 3, 2010 at 6:04 AM  
Blogger Mr Durand Pierre said...

You and I are thinking along the same lines, my friend.

This was definitely my biggest gripe with the Mass Effect games- that you knew what dialogue options were "good" and which were "kinda jerky, but still good" without having to actually navigate a conversation or pay attention to what the other person was saying. The game actively encourages you to pick a side and stick with it too (since you get more dialogue options that way), so once I went Paragon I had to really force myself to choose a Renegade option knowing it wouldn't benefit my character as much. For shame.

Oddly, ME2 contained a non-telegraphed moment that was both brilliant, yet frustrating in how it was presented. Some spoilers may follow...

Once you complete a certain mission your crew will get kidnapped and you'll have the option to go straight to the endgame. I was a completionist, so I stuck around doing all the optional "loyalty" quests. By the time I got to the endgame I'd waited too long and my crew was dead. I was totally fine with this as I figured a pure happy ending would have felt cheesy and I liked that I had to make some sacrifices along the way and that it was my own fault. It was unexpected, but not illogical.

The problem with this is that if you do all the missions before triggering mission x where your crew is kidnapped, you can have the best of both worlds. I found this a bit silly that you're still procrastinating from saving the galaxy by the exact same amount, but if you do it in an arbitrary order you're fine.

I think it would have been infinitely better if you couldn't have the best of both worlds and you did have to make the tough call I did.

As long as the consequence seems to stem naturally from the choice I think it's fine to unfairly lure a player into a trap. I actually have an article going up very soon about just that...

November 3, 2010 at 2:47 PM  
Blogger Nels Anderson said...

@thequickbrownfox Yeah, this is definitely one those cases where you're boxed in by the desire to provide more fidelity/immersiveness. There a sort of spectrum, I guess, with freedom on one end and fidelity on the other. Tabletop RPGs sit on the far side, because in the latter, the bounds truly are only what you imagine and agree upon socially. And then more fidelity, VO, etc. keeps moving you further and further toward the other side. Totally agree with that you're saying there, although I think there are ways to deal with this without falling into a huge content sinkhole. Certainly ain't easy or well established though.

@Christof I love that Eastern Europe seems to have taken up the mantle of Immersive Sims, even if they sometimes produce incomprehensible madness (e.g. Pathologic). And I'll have to check out Chris Thursten's piece, thanks.

@Mr Durand Pierre Any game with a universal good-evil/karma/morality scale falls prey to this. Instead of actions having any ambiguity, you'll immediately informed of their absolute moral integrity/wickedness. ME2 made that even more explicit, but I'm a little disappointed Fallout: New Vegas added faction standing (great) but still kept the global karma too. I guess it's heritage now (but even that's a paltry excuse), but it still makes it sorta hard to feel as if there's any grey area in your decisions.

November 3, 2010 at 10:44 PM  
Blogger Dagda (Brooks Harrel) said...

I think there's a second side of things to consider here, one element of modern game design that's still rather primitive: The consequences of failure are almost always going to be a short-term setback (losing some regenerating health) or "go back and try again".

Very rarely is a game with strong goals or narrative designed so that the player will live with the consequences of their imperfect decisions & actions (rather than just taking them as a cue to load an old save). The exceptions are well worth examining:

-At the start of a new game, Mount and Blade always strongly recommends that the player choose the game mode that forfeits the option to reload an old save. You spend the game working to build up a band of fighters, with numerous ways to make a profit and/or fight for a cause. If you choose to fight in person and get cut down, the game describes your men carrying you from the field of battle (with the rescue effort and lost morale increasing the number of casualties). If your band is crushed entirely (typically via multiple engagements with a superior force that's managed to corner you) then your character is captured, triggering a time-lapse sequence as you're dragged around the map until your captors are defeated, you manage to escape or an ally pays your ransom. You won't quite have lost everything, though- your character's skills remain intact, as do the personal bonds you forge with named NPCs.

-King of Dragon's Pass throws random dilemmas of all kinds at you, many of which play out in multiple parts over months or years. Choose to exploit a nearby village of weak beastfolk and you'll be in hot water when their hulking friends come to visit. Decide to try raising the strange lizards are hatching in the nest you found, and you might get a notice on how your clan now rides triceratops into battle. Different options can have different results based on luck and the abilities of your clansmen; I was ready to string my clan's Trickster up after he ruined our relations with a neighbor, then forgave everything when he volunteered to talk to a dragon our hunters found & got it to gamble away half its hoard.

-X-COM gets very interesting whenever you're in territory you haven't played through several times before. The guesswork involved in researching technology with uncertain benefits to keep the pace against the as-yet-unknown ways in which the enemy threat will escalate is a fairly unique experience even after all these years.

November 8, 2010 at 11:18 AM  
Blogger Nels Anderson said...

@Dagda That's a big ol' issue, no doubt. It's tricky, because (and I think rightly so) I've been very frustrated playing games where certain failures were permanent and irrevocable (Fire Emblem comes to mind). The difference between engaging/tense and frustration is both fine and subjective. There are definitely ways to enforce permanent consequences, but do so elegantly in a way that actually enhances the game is non-trivial.

Probably the way it feels the most organic and satisfying is to have consequences be deferred. I still love the narrative "post-mortem" you get at the end of Fallout 1&2. You see the long-term effect your decisions had, but at that point, it's much too late to reload to change them. Bioshock 2 did that a bit as well.

Of course, those are just narrative solutions. How to do it in a more systemic way, I'm not sure. Obviously you could have delayed consequences there too, but it's a bit harder to make that readable and good gameplay, you know?

November 9, 2010 at 10:57 PM  

Post a Comment

Subscribe to Post Comments [Atom]

<< Home