Let the Stupid Rats Die. For Science. You Monster.

Clear-headedness under pressure.  It's a skill that should come in handy if I ever find myself in a life-or-death situation where I need to keep my wits about me to survive.  Am I driving a FWD, RWD or 4WD car?  That will be key to knowing how to regain control of the vehicle and not slide off the cliff into the canyon below.  Will water extinguish this fire, or cause an explosion?  Should I put my hands out to stop my fall, attempt to roll with it or take the full force of the fall on my stomach so as protect my head?

That last one is a real situation I found myself in where I did indeed take the fall on my stomach.  With a metal bar acting as the delivery system.  I couldn't roll with the fall and putting my hands out would have catapulted me off a ledge and onto my head.  The stomach option was very painful, but I take pride in the notion that it was actually the smart choice.  The details mattered.  In the best games, the details matter, because then my decisions have meaning, and I have a chance to develop this crucial survival trait that I may need some day.

Portal 2 is Well-Designed
I've played through Portal 2's single-player campaign, and the co-operative campaign with my wife, and I've just played through the single-player again with the developer's commentary enabled.  The  game's developers, Valve, are big into testing-based design.  That means that they get a bunch of playtesters to go through their game and they observe where the testers get stuck, what's fun for them, what's frustrating, and what just goes completely ignored.

It's a great tool in game design.  I remember wandering around old Doom levels for ages just trying to find the right key-card or door.  It was that sort of thing that made me give up the game after a while.  Test-based design, when utilised well, does wonders to correct problems like this.  Valve in particular is pretty good at pointing you in the right direction so you don't get lost in levels that are essentially linear.  Of course, this is a little too frequently expressed in the form of an arrow painted on the wall, literally pointing you in the right direction.  Baby steps.  At least they're addressing the issue.

Just like this one.

Portal 2 has been tested and honed to a tee, and the "drawing arrows on walls" problem is ameliorated somewhat by the fact that the entire setting places the player in a metaphor as a rat in a maze, with an omnipotent, indifferent scientist observing and manipulating their every move.  The controlling AI, GLaDOS, can reconfigure the entire facility to reroute you to where she wants you to be, so the sense of being railroaded into a particular course is perfectly sensible, and not at all immersion-breaking.

Reconfigure the Phlebotinum Field!
Anyway, this all brings me to the developer's comment that inspired this post (don't worry, no spoilers):

In this excursion funnel ride, playtesters would often end up shooting the wrong Portal at the critical moment and killing themselves.  This ruined the moment for players who often didn't quite understand why their excursion funnel hadn't been redirected.  As a solution, we now detect when the player places the wrong portal in hopes of saving themselves.  We help them out by moving their other portal under the excursion funnel source.  This effectively makes this section foolproof by allowing the player to shoot either portal to save themselves.

You don't really need to know what an excursion funnel is.  Replace every instance of the term with "[phlebotinum]" and the quote can still be understood enough to get the gist.  Basically, [phlebotinum] is needed to succeed.  The player has control of two portals and to correctly succeed, they must use the correct portal to manipulate the [phlebotinum].  If they use the incorrect portal the [phlebotinum] disappears and they fail.  Incorrectly.

Okay, fine, this is an excursion funnel.

The thing is, the [phlebotinum] can be anything.  It can be a "hard-light bridge", paint, the player, a box, even the player's point-of-view.  It's often a slightly more abstract concept like velocity or rotation.  Usually several of these must be manipulated at once.  This is the essence of the game - manipulate thing with portal placement.  The excursion funnel is just the newest addition to the roster of things that can be manipulated.

Throughout the entire game the player is trained to solve puzzles by placing portals.  One lesson they should have learned by this point, very near the end of the game, is that if they shoot the wrong portal at the wrong moment it can ruin their day.  I developed a habit of making a quick mental note every time I placed a portal: "Orange portal is at [phlebotinum] source; therefore blue portal delivers [phlebotinum] to destination."  "Orange portal has been placed at [phlebotinum] destination; therefore blue portal must be placed at [phlebotinum] source."  "Blue portal is at point A; orange portal is at [phlebotinum] source; [phlebotinum] is now at point A; move orange portal to [phlebotinum] destination."  It can get quite complicated if you're not on the ball.

I was going to link to a video, but this is more succinct.

It was All a Dream
The situation that the developer is describing is a tense, time-pressured escape where any mistake means death.  The antagonist is yelling at you, distracting you with hilarious one-liners and trying to mash you to death with crusher panels.  If ever there was a moment that this skill I've developed should pay off it's right now.  So I remember that I've got the orange portal at the [phlebotinum]'s renewable source, and wherever I need to use the [phlebotinum] I just place a blue portal.  I can't get back to the source anymore, so I keep my finger off the orange portal button.  If that orange portal moves, I lose the [phlebotinum] and I'm dead.  With this strategy, the solution is completely turnkey.  Place blue for [phlebotinum], repeat as needed.  If I remember my rule, I'll succeed.  This is a detail that matters.

So I got through that section myself without any problems (surprise surprise), and I thought it was because I'd done it right.  Now the developers are telling me that they rigged it.  There was only one place to put a portal at that moment, and it didn't matter which one.  So in fact the only thing I could've done wrong was nothing at all. I have no idea if I got it right the first time I played, so the intrinsic reinforcement of success that test gave me was in fact false.  They didn't want to ruin the flow of the moment, so they removed all meaning from it instead.

Now, a friend of mine who is very experienced at running tabletop role-playing games says that he does that sort of thing all the time, and the developers using this trick isn't really the problem.  Their mistake was telling me about it.  The thing is, there's a difference between using sleight-of-hand in a computer game and using it in a live situation.  There's a rule among magicians that you should never repeat a trick for an audience, because they'll get a chance to analyse the trick and figure it out, which will ruin it for both of you.  In a computer game, the audience has infinite do-overs.  Even if the devs hadn't revealed that trick, it would've been noticed by someone at some point.  Also, you don't get do-overs in a live situation, so you need the tricks.  A ruined moment is ruined forever; it can never be fixed.

So What's the Problem? 
I'll tell you what the problem is:  I'm playing a puzzle game.  If a moment in the game is dressed up and presented even as a trivial puzzle, I expect it to be an actual puzzle, with some kind of potential for failure.  When Yahtzee reviewed Portal 2, he bemoaned the fact that instead of being a puzzle game with a really good story framework, it was instead an interactive story with some puzzles thrown in because, well, it's supposed to be a puzzle game.

What I've described here is only a single example of the developers placing story and accessibility above the actual puzzles in this game.  They've done a lot of things right, like removing time or dexterity dependent puzzles from the game, thus allowing the main story to be experienced by a broader spectrum of players.  That's a good thing, because it's certainly a fun ride, one that's helped my wife to get into the first-person genre a bit more, for which I am eternally grateful.

The thing is, this is near the end of the game.  I can't imagine a player who would be critically discouraged by a minor setback at this point, so audience members wouldn't be lost here.  In the developer's words, they just didn't want to "ruin the moment".  Okay, sure, it's a good moment, but it's only ruined for people who made a mistake, and since there's a 50-50 chance of getting it right even if you've left your brain at the door, those people are almost certainly in the minority.

I Do Not Understand How They Could Not Understand
But here's the real infuriator for me - apparently the moment was doubly ruined for players who "often didn't quite understand why their [phlebotinum] hadn't been redirected."  Seriously?  They didn't get it?  Often?  They've been playing this game for 6 or 7 hours if they're good at it.  I can guarantee you that if they're still confused about why they failed, it took them a lot longer than that to get here, and they should be made to relive their stupid, pointless death over and over until they do understand.

Don't get me wrong here; I'm not saying that stupid people should be punished.  Far from it.  If you've made it this far, you're obviously not stupid enough for this lesson to be beyond you.  If you've made it this far but somehow managed to avoid such a fundamental piece of understanding, then you are clearly complicit in your own ignorance.  This is supposed to be a puzzle game, and while it is inevitably beyond the infirm of mind, one of its primary functions should be to cure the lazy of mind.

This is the reason that I believe that test-based design can't solve all our gaming ills.  When I heard that quote, I realised why the game was so easy.  It seems that Valve really lost perspective in their pursuit of a smooth and enjoyable experience for all their test subjects.  If the worst players are given veto power over anything that's too difficult for them, then the whole game will be brought down to their level.  Looking back over my time with Portal 2 I realise that it wasn't satisfying at all, because I had learned very little.

One of GLaDOS's many foibles is her incessant compulsion to test, and test, and test.  Apparently she's doing science, but it's never really explained what kind of science she's doing.  If they're just testing the portal device, they don't really need to put humans through a deadly gauntlet.  The tests seem more suited to psychological research, but even then it's never really revealed what it's all about.  It's kind of a subtle running joke throughout the whole story.  I find it ironic, then, that Valve has failed to learn the moral of their own story by literally testing to a fault.  If there's one thing that we can learn from Portal 2, it's that GLaDOS would make a poor game designer.

This is a jump.
Because I needed something to separate the article from the epilogue.

Post Scripting
Hopefully when Valve releases their new DLC for the game in the middle of the year there'll be some stuff in there for the hardcore puzzlers.  But if not, I'm okay with that.  If there's one thing that putting my ideas out on the internet for all to see has done for me in this case is to make it clear to me what I want:  I want to make their game better.  It's for people like me that Valve helped pioneer the release of developer tools to their gaming audience.  I'm gonna use the free Portal 2 Authoring Tools to make hard puzzles, and I'm gonna try to make them right.

The solutions will all be applications of well-known Portal mechanics in clever and complex ways, requiring some real head-scratching to pull off.  One thing I'm not going to do is require (spoiler warning) silly, elaborate, esoteric, glitch-exploity, waste-of-damn-time tricks to make it through the levels, like some user-made maps.

There's a map contest being held at the moment, and I'm doing a level design class at college in the trimester after this one.  Maybe I can get credit for this.

Hang on a Second...
I've just had another thought.  The physics in first-person games have been developed around games in which navigating the environment was secondary to the challenge at hand.  In fact, if you missed a ledge while attempting to defeat the invading horde of evil things then it was a cheap death, and it detracted from the game.  So a bunch of helper physics were written into early FPS games to reduce this problem.  The most obvious one allows the player to alter their trajectory mid-jump.  Obviously this is absurd, but you can see why they did it.  Unfortunately, it's exactly that kind of physics that has made it into the Portal franchise without a whole lot of modification or even thought, and it's allowed a lot of cheap tricks to bypass the real puzzles at hand.  You could argue that these tricks require a lot more skill than the original puzzle, and in fact they wouldn't occur to you the first time you tried the game, but I wonder if the same physics would've been implemented if they were made from the ground-up for a physics-puzzler like Portal.  My guess is not.