I was with a group of game designers and I said, "Probably I'm in the minority here, but I was really disappointed with Portal 2." To my surprise, everyone there (about 10 game designers) agreed, and not a single person was willing to take up the contrary position. I'll explain the complaints, but first I'd rather say what is great. This is all about the single player mode, by the way.
Mechanics Spoiler
I think the mechanics are great. SPOILER of mechanics. I'll just casually mention some here, and you can skip this paragraph if you'd rather not know them ahead of time, in a strange world where you have somehow not played this game yet. Anyway, there are beams that you redirect through portals, bridges made of light you can extend through portals. There are three kinds of liquid goo that you can spatter around the world, each with different properties. I really liked the kind that let you portal off of otherwise unportable material. There are force fields beams that carry you over pits, and you can redirect those through portals. All of this stuff is great. My letter grade is A or A+ here.
Good Stuff
The beginning of the game was probably pretty difficult to design. It has to kind of feel like the old game, also it has to have easy stuff that's like the beginning of the old game so new players know wtf is going on mechanics-wise. It also has to introduce some sort of story and tell us where and when (and who!) we are, relative to the last game. I think it succeeded on all these fronts. Another A grade.
The end (not like the last minute, I mean the last 20% or something) has to put together stuff we learned and give challenges that are more complex than the earlier ones. There's a lot of mechanics here, so it might have been hard to figure out just how to put these together for us in the last several puzzles. I think this was done very well, too.
Subtractive Design is Needed
What's the damn problem then? Everything sounds pretty great! Well, there are three things. The first is that there's this pretty long middle section of the game that kind of throws away what is good about Portal. Instead of small environments where you can portal off of almost anything, it's large environments where you can portal off of almost nothing. These environments look great, so whoever implemented them did a great job, but their very existence makes for a worse experience, in my opinion. Sometimes these parts felt more like a "Where's Waldo" puzzle of just finding the tiny thing that I'm allowed to shoot a portal on.
One designer raised the point, "Whether the environment is large with few portable surfaces or small with many portable surfaces, it's solving a puzzle either way. In both cases, there is usually going to be one correct solution to the puzzle, so does the objection really matter?"
We were all quick to say, "Yes, it matters." The best articulation of this point was that in a small room with nearly all portable surfaces, you are surrounded by choices that are wrong. It takes thinking to figure out what would be a right thing to do. But the feeling is not the same in a large environment with very few portable surfaces. There, you have very few choices and you can solve a puzzle by performing the only real legal moves, not even knowing exactly why it worked. Sometimes not really having to think about it.
Why do these large areas even exist at all? My guess is that there was an executive decision to sell a boxed $60 game, and that game design should just figure out what to do. I hope I'm wrong in that guess. A more sensible approach would be to make the best game possible, and sell it for whatever price made sense. At $60, I am guessing Valve thought people needed to feel a more grandiose experience. And further, that the game needed to be at least a certain length and a series of small test rooms would feel too monotonous for $60's worth of length, whatever that means. So to vary the experience, maybe they thought, "What if you could portal around a Half-Life-like environment?!" Nice idea maybe, but I think some subtractive design would have removed that, for the better. There's something pleasing about the idea that the Portal 1 world is all portable, except for specifically marked black material. The long middle section of Portal 2 teaches us that any old random material is not portable, only the white texture is portable. It just feels sad, like it's not following through with the really bold concept of "if you can portal off everything, how is there still a game?" Portal 1 answers that, but Portal 2 is like afraid to fully embrace it during this middle section.
Objection 2
The above was my strong objection. There were two other objections by other designers. One was that the game is just too easy. I hadn't thought much about that, but when he said that...yeah it did seem pretty easy overall. You could say the market has spoken, and they want easy games. That's not a very satisfying answer to me though. A game like Rez shows how it can be hard (score attack, trying to get 100% shot-down), easy, or even zero difficulty with "traveling" mode. What if I wanted a harder Portal 2? Why can't I have that?
I'll tell you why I can't have it. Because if a puzzle is too hard, it will stop all progression. There is no way around it, given the structure of the game. I think this is where Portal can learn something from Braid. Braid faced that same design problem, but it had a different solution. In Braid, you can progress past a puzzle you're stuck on, you can just skip it basically. Maybe something will click for you later, and you'll go back and finish it (you want credit for all those puzzle pieces after all!). This type of structure allows the designer to make harder puzzles and get away with it. I kind of wish Portal 2 had done this.
Objection 3
This brings us to the last objection I heard, and it's the most subtle one. The objection is that Portal 2's underlying goal is sort of the wrong one. The design goal appears to be to give the player that "aha!" moment (ok, great) but to not care too much about how he got there. The claim is that a better goal would be to primarily care about the transmission of an idea from the designer to the player. This is a pretty interesting point, so let's look at what's really being said here, and how it differs in Braid and in Portal.
Jonathan Blow has said publicly several times that Braid is not about making the hardest possible puzzle. He isn't afraid of having a hard puzzle, and as we discussed earlier, giving the player a way to skip a puzzle, keep going, then come back later allows the puzzles to be harder than they'd be allowed to be in a strictly linear game. Anyway, he says the point is to have INTERESTING puzzles, which is a different concept than hard ones. Braid is trying to communicate ideas about time to the player. When a new mechanic is introduced, the puzzle is usually something that makes the player realize the logical consequences of whatever time-thing is going on. By realizing that, the player can solve the puzzle.
Let's go back to Portal now, and use some more concrete examples. One of the very first things in both Portal 1 and Portal 2 is situation where you see the orange portal in front of you, on the other side of a pit. You have the blue portal gun, but not the orange gun. (Or maybe you do have the orange one in the Portal 2 version of this, I forget). Ok so you shoot a blue portal right next to you and this lets you come out the orange portal, on the other side of the pit. Yay!
But now there's a left turn and another pit between you and the way out. The very first moment I saw this in Portal 1, there was a split second where I thought, "Ok, so I need the orange portal gun." I was thinking that I just came out of an orange portal, so I need to put an orange one on the far wall. But of course after just a moment, I realized that's not true. All I need is the blue portal gun because I can go *in* the orange portal I just came *out* of and it will lead to the room's exit as long as I put a new blue portal past the second pit.
Another way of describing this is that in this moment, I learned the concept of using one central portal to go in and out of, while you move the position of the "satellite" portal. This is almost too easy of a concept to write so much about, but it's an example of the right kind of thing. I don't think this first "puzzle" had any challenge whatsoever, but it caused me to think something and understand something, so it was a very good thing to have in the game.
A more complicated example is concept of momentum through a portal. At some point, Portal 1 shows you that if you put a portal way down at the bottom of a pit and fall into it, then you can come out of some other portal somewhere else really fast. Maybe fast enough to fly across some other pit. Again, that puzzle isn't really about it being hard, it's about communicating that idea to me. That idea is a tool, and I can now use that tool in my arsenal as I progress through the game.
Portal 2 does have moments like this, but I almost think they are a side-effect rather than the entire point of the game design. There are so many other moments that are more like shooting around at whatever available where's-waldo surface, and somehow progressing. Progressing in a situation like this can be "wow, it worked!" and it seems Valvle polished up those moments to be all they can be. But "wow, it worked!" is just not the same depth of experience as "wow, I get it now!" Portal 2 does have those moments, but I think it has way too many of the "wrong" kind of wow.
To put it another way, I felt Braid was trying to educate me, in a way. Portal 2 seems more concerned with entertaining me. Being an entertainment product, it's hard to fault a game for being entertaining, but the worst parts just feel...more shallow or something than the experience from the education moments. In both Braid AND Portal 2, the education moments are are very satisfying, so it seems that should have been more of the goal.
Despite all these complaints, I still recommend the game. It's highly polished, has great mechanics, and is...well...entertaining.