20 replies to this topic

[Beefster]

I just wrote another post on Linearity. I've gone even farther into depth of theory and avoided ranting (for the most part). I classified types linearity within two subgroups: physical and objective.

Here, I'd like to talk about why linearity worked for Half Life 2, but not for Portal 2.

Half-Life 2 was very story-centric and had a point A to point B mentality. The linearity was justified for the most part. You didn't have any exceptional mobility or freedom, but it didn't feel unnatural. It kinda made sense to ask "where do I go next?" Not the best application of linearity, but not horrible either.

But Portal 2... It just doesn't feel right. Linearity works wonders for the test chambers, since it's really just a first-person puzzle game in that context. But when you're in the areas between test chambers, it's rather obnoxious. YOU HAVE A FREAKING PORTAL GUN! You should be able to shoot portals onto just about any nonreflective flat surface, but unfortunately, the gun got nerfed between the two games. The game had so much potential to allow for INSANE player choice, but they floundered their opportunity.

Just imagine: at a point early in the game, there could be three possibilities "offered" and some relatively early point: continue through the test chambers, try to escape, or try to re-destroy GLaDOS. Or take the existing setup and add several branches from it that allow similar choices.

[Mitchfork]

Well, I think what you have to remember about the Portal series is that it's not an adventure or shooter game; it's a puzzle game. You start at level 1, you solve the puzzle, then you move on to level 2. I was very aware that this was the structure and I didn't mind. Linearity is not a bad thing.

[Nicholas Steel]

What? In Portal you can't shoot portals on every surface either, that pretty much justifies it in Portal 2 but in Portal 2 they went a step further with explaining that portals can only be created on surfaces coated in Moon Dust.

If you could shoot Portals anywhere then it would be a very dull puzzle game, may as well just make it so moving the mouse insta wins the game at that point.

[The Satellite]

Yeah... I don't get what's wrong with Portal being linear. It makes perfect sense, honestly, because, well, like Eb said, it's a puzzle game.

[Beefster]

Portal being linear works until you're not in the test chambers anymore. Once you're in a situation where it makes sense to have freedom, it's aggravating to not actually have it.

Half of Portal 2 is not test chambers- and it's silly that the game is linear for those sections.

[Nicholas Steel]

Well how would you implement the story in a non-linear environment and have various effects etc. be visible and attract the players attention?

[The Satellite]

Actually the entire game is nothing but test chambers, it's just half of it are in very old-school, dilapidated ones.

Sure, there're segments between each facility through the years, but those are just meant to take you to the next area anyway.

[Giggidy]

Just imagine: at a point early in the game, there could be three possibilities "offered" and some relatively early point: continue through the test chambers, try to escape, or try to re-destroy GLaDOS. Or take the existing setup and add several branches from it that allow similar choices.

Let's say, a quarter of the way through the game, the player DID have a choice of either killing Glados, trying to escape, or continuing, and there is no more branching and no re-used content shared by the three paths.

Congratulations! 60% of the game's content won't be seen by the player on the first playthrough no matter what. How many players will go back and replay the beginning to see all three paths? Let's be extremely generous and say 50%. This means half your playerbase will only see 40% of the effort you put into the game. Which means you wasted tons precious development resources, and for what?

Don't say "replay value" because the entire concept of replay value is a trap. What would you rather have, a 10-hour game you have to play 8 times to see everything, or an 80-hour game where you see everything on the first go? Even for the players who do go back and replay the game twice more to see everything, you've also done them a great disservice: Making them play through the opening 25% of the game *three times*, the latter two of which almost certainly being pure busywork. You make everyone happier if you just put all of your resources into the best first-time playthrough possible rather than making players do things multiple times.

But Portal 2... It just doesn't feel right. Linearity works wonders for the test chambers, since it's really just a first-person puzzle game in that context. But when you're in the areas between test chambers, it's rather obnoxious. YOU HAVE A FREAKING PORTAL GUN! You should be able to shoot portals onto just about any nonreflective flat surface, but unfortunately, the gun got nerfed between the two games. The game had so much potential to allow for INSANE player choice, but they floundered their opportunity.

The real problem is most of the puzzles in the middle of the game only have one or two possible places to put portals. When you can put a portal almost anywhere, you're surrounded by choices that are wrong. Distinguishing the incorrect choices from the correct ones is the object of the puzzle, but when there's very few choices, the puzzle ceases to be meaningful.

[Beefster]

Let's say, a quarter of the way through the game, the player DID have a choice of either killing Glados, trying to escape, or continuing, and there is no more branching and no re-used content shared by the three paths.

That's a rather clumsy assumption. Nonlinearity is often way more complex than that. I was giving a heavily simplified version which probably wouldn't do much good on its own.

Congratulations! 60% of the game's content won't be seen by the player on the first playthrough no matter what. How many players will go back and replay the beginning to see all three paths? Let's be extremely generous and say 50%. This means half your playerbase will only see 40% of the effort you put into the game. Which means you wasted tons precious development resources, and for what?

That's a rather selfish attitude. When it really comes down to it, it should be about the players, not the developers. Some simple writing and slight game shortening would mitigate that problem anyway.
But think about it: if you make your players happy and your game replayable, they won't resell the game because it's worth playing again. (admittedly, in this age of digital distribution, this isn't as applicable as it used to be...) They like your games and want more. If you want to make more money from your games, you need to make them better.

Don't say "replay value" because the entire concept of replay value is a trap. What would you rather have, a 10-hour game you have to play 8 times to see everything, or an 80-hour game where you see everything on the first go?

The 10-hour game that you have to play 8 times to see everything. Starfox 64 is a VERY short game, but it's hundreds of times more replayable than other games. It's made clear that there are choices to make (well, based on skill level), so players will easily play the game a hundred times before getting tired of it for a few years.
Of course, it depends on the structure of the game. If a world map is there, taunting me with the possibilities, it will make me want to play the game over and over. And since it's nonlinear, the game is a little different every time because of my choices.

Even for the players who do go back and replay the game twice more to see everything, you've also done them a great disservice: Making them play through the opening 25% of the game *three times*, the latter two of which almost certainly being pure busywork. You make everyone happier if you just put all of your resources into the best first-time playthrough possible rather than making players do things multiple times.

Or in my particular example, you could make the alternatives somewhat obscure, leaving the "normal" ending for the typical first play-through. Or you could make it clear that the other content is there.
You could even make the occasional branchoffs really short and mock the player for being a "quitter" or something. It gets them back on the rails without ever feeling railroaded.
And another solution would be to have multiple routes to the next section of tests. Probably not something you'd notice, but you might figure it out on a second playthrough.

The real problem is most of the puzzles in the middle of the game only have one or two possible places to put portals. When you can put a portal almost anywhere, you're surrounded by choices that are wrong. Distinguishing the incorrect choices from the correct ones is the object of the puzzle, but when there's very few choices, the puzzle ceases to be meaningful.

This is exactly the part I don't like about the "in between" sections of the game; There are no meaningful choices, so you don't really have to think.

[Giggidy]

That's a rather clumsy assumption. Nonlinearity is often way more complex than that. I was giving a heavily simplified version which probably wouldn't do much good on its own.

If that's not what you wanted then it's not the example you should have put forward.

That's a rather selfish attitude. When it really comes down to it, it should be about the players, not the developers. Some simple writing and slight game shortening would mitigate that problem anyway.
But think about it: if you make your players happy and your game replayable, they won't resell the game because it's worth playing again. (admittedly, in this age of digital distribution, this isn't as applicable as it used to be...) They like your games and want more. If you want to make more money from your games, you need to make them better.

> Implying this complaint had anything to do with money.

Look, no matter what kind of game you're making or how many resources you have, there are opportunity costs. Working on one thing means giving up working on another. Whether you're just putting a hobbyist open-source game together in your spare time or you're working on a hundred-million dollar project, you have to decide what's important. My central argument is that branching story paths provide a very poor ROI no matter what your goal is.

The 10-hour game that you have to play 8 times to see everything. Starfox 64 is a VERY short game, but it's hundreds of times more replayable than other games. It's made clear that there are choices to make (well, based on skill level), so players will easily play the game a hundred times before getting tired of it for a few years.
Of course, it depends on the structure of the game. If a world map is there, taunting me with the possibilities, it will make me want to play the game over and over. And since it's nonlinear, the game is a little different every time because of my choices.

I think we're just gonna have to agree to disagree here: I couldn't even bring myself to play Starfox 64 for an hour before I never wanted to play it again. I couldn't imagine myself still being entertained by that *years* later.

You could even make the occasional branchoffs really short and mock the player for being a "quitter" or something. It gets them back on the rails without ever feeling railroaded.

Isn't that what Portal 2 already does?

[Beefster]

The idea of a long play time is a trap. You could make a game 50 hours long and it still won't feel fulfilling because you never really had to use your brain the entire time. On the other hand, you could have a 1 hour game that makes you feel like you really accomplished something. I generally don't criticize games for their length; I criticize design.

Look, no matter what kind of game you're making or how many resources you have, there are opportunity costs. Working on one thing means giving up working on another. Whether you're just putting a hobbyist open-source game together in your spare time or you're working on a hundred-million dollar project, you have to decide what's important. My central argument is that branching story paths provide a very poor ROI no matter what your goal is.

Linearity is still (often) the easy way out. They could have at least made the world as a whole physically open. I would have been perfectly happy if the entire game were like Co-Op- where it's puzzle-centric. My one and only criticism with Portal 2 is the parts between test chambers.

Isn't that what Portal 2 already does?

Not really. Other than a few secrets here and there (such as the singing turrets), there is nothing meaningful off to the side to explore. There are no optional puzzles. Like I said, this would be perfectly reasonable without the in-between parts.

SOOO... Anyone actually read my blog post? I was hoping more of the discussion would come from that...

[Mitchfork]

That's a rather selfish attitude. When it really comes down to it, it should be about the players, not the developers. Some simple writing and slight game shortening would mitigate that problem anyway.
But think about it: if you make your players happy and your game replayable, they won't resell the game because it's worth playing again. (admittedly, in this age of digital distribution, this isn't as applicable as it used to be...) They like your games and want more. If you want to make more money from your games, you need to make them better.

The problem then becomes that puzzle games aren't very replayable as a genre. Once you've solved the puzzle once, solving it again is sort of trivial unless you tack on challenges (the Portal achievement chambers, for example). If I have to play through the first half of the game again to get the content I missed, I don't think that's very good design.

Granted, Portal 2 had chapter select, so maybe I'm overstating a bit on that. The linearity really didn't detract from the experience in my opinion though.

The 10-hour game that you have to play 8 times to see everything. Starfox 64 is a VERY short game, but it's hundreds of times more replayable than other games. It's made clear that there are choices to make (well, based on skill level), so players will easily play the game a hundred times before getting tired of it for a few years.

Well yeah, that's an arcade shooter.

This is exactly the part I don't like about the "in between" sections of the game; There are no meaningful choices, so you don't really have to think.

That's a good thing. While I rarely struggled with any of the puzzles in Portal or Portal 2, I also recognize that repeated puzzles grow monotonous no matter how good your game design is, especially since every puzzle essentially has the same solution (shoot portals). Think about Half-Life 2; usually right after you'd solve a puzzle (for example, shooting the energy balls into the receptors in the citadel) there would usually be some enemies pop out that you have to fight. Contrast to, say, Call of Duty, a shooter that has immersive graphical realism (for the most part), excellent smoothness and control (debatable, but I think that Call of Duty has the best player movement mechanics of any shooter I've played), and can't figure out how to do anything other than throw brown people at you.

I mean, it's largely a matter of taste, but I rather enjoyed the "in between" sections of Portal 2. They gave a chance for players to take a break between the puzzles and offered exposition while still doing something interesting; it's neat to see Aperture's underbelly (and there are a lot of cool Easter Eggs), and Wheatley's/GLaDOS's dialogue at these junctures was very entertaining and a key reason why people enjoy the games so much. If you have dialogue constantly playing while the player is solving puzzles, you risk them tuning out a key part of the game's experience, or stopping to listen at which point you get boring talking heads. In Portal 2, almost all of the dialogue was either at the beginning of a puzzle, exiting a puzzle, or in an in-between corridor; it's an intentional design choice, and I think it's a really solid one.

Portal and its sequel (really a lot of Valve games) are really notable for selling a compelling, entertaining story while keeping the player in control of the camera at almost all times. It's a great accomplishment. Neither are perfect games (the first comes close in my opinion) but I think criticizing it on its linearity really misses what the games set out to do.

[Giggidy]

The idea of a long play time is a trap. You could make a game 50 hours long and it still won't feel fulfilling because you never really had to use your brain the entire time. On the other hand, you could have a 1 hour game that makes you feel like you really accomplished something. I generally don't criticize games for their length; I criticize design.

Of course: My statement made the presumption that all the playing time is equally intensive. Of course a 10-hour game with no filler is better than a game where you have to spend weeks doing pointless grinds to get anywhere, but an 80-hour game with no filler is strictly better. In most games (star fox included) replaying levels *is* a grind because it's the same content you've already cleared before.

SOOO... Anyone actually read my blog post? I was hoping more of the discussion would come from that...

I did, but just had nothing to say about it: It's not so much of a discussion of linearity vs. openness as it is a description of it. A rather surface-level one, at that. If you're interested in a serious game design post, you should talk about things like:

1. How to make backtracking through an area you've already been through (necessary in sandbox games) interesting.

2. Thoroughly connected game space designs (Skyrim) vs. Hub-and-Spoke designs (Super Mario 64).

3. Reasons why you would want to seal off some parts of the world from the player until they accomplish some arbitrary goal. (e.g., GTA: San Andreas only allows you to roam around in Los Santos until you get sufficiently far into the story.)

4. The difficulty of effectively hiding secrets and collectibles in open 3D environments, and why this is much easier when the player is basically guided down a hallway.

5. How to make an effective, worthwhile sidequest.

[Beefster]

Good ideas. I'll make sure to revisit it again and go into more depth. I don't have much experience with the design aspect, though. But I'll see what I can put together.

I kinda touched on #4 once. I don't think I really did it justice, though.

I felt my surface analysis of linearity was important because I've realized that it's more than a simple sliding scale. There's a physical aspect to it and there's an objective aspect to it. You can't really say your game is nonlinear if the objectives always follow the same sequence.

Personally, I prefer convergent objectives in an open world, but other combinations have been known to be good (IMO).

Man. That gives me a lot of future topics to talk about... I should also start incorporating pictures into my posts and put more than a couple hours per weekend into it.
Topics I want to cover at some point:
-Your 5 topics
-Endgame quests
-New Game Plus
-Challenge runs (such as minimalist runs)
-Play Time and Filler

By the way, thanks for giving me a month's worth of things to post about... On top of the month's worth of stuff I already want to do.

[Giggidy]

I kinda touched on #4 once. I don't think I really did it justice, though.

Really? Where?

I felt my surface analysis of linearity was important because I've realized that it's more than a simple sliding scale. There's a physical aspect to it and there's an objective aspect to it. You can't really say your game is nonlinear if the objectives always follow the same sequence.

Yet, can something be said to be physically linear and objectively non-linear, under your definition? It's clear they aren't orthogonal.