Column: 'The Interactive Palette' - Puzzle Design in the Myst Series
February 20, 2009 8:00 AM |
['The Interactive Palette' is a biweekly GameSetWatch-exclusive column by Gregory Weir that examines the tools and techniques of the digital games trade with a focus on games as art, using a single game as an example. This time - a look at puzzle design in the Myst series.]
Puzzle. The word has many different meanings in the context of video games. The term "puzzle game" can refer to a game in the mold of Tetris; there are many shaped or colored blocks or jewels or bubbles that must be cleared by manipulating them within a time limit. In this context, the "puzzle" tests the player's coordination and reaction time.
However, the word "puzzle" is also commonly used in a broader and more conventional context to describe an intellectual obstacle in any video game. When the developers of Half-Life 2 break up the shooting-and-driving gameplay to have the player assemble a makeshift ladder, that's a puzzle. So are the ubiquitous sliding-block puzzles of Zelda infamy. In these cases, puzzles are intended to serve as pacing devices, allowing a moment to relax and think in the middle of all the killing.
However, there are categories of video games where the game is entirely composed of puzzles. One group is games like Chip's Challenge, 3D Logic, or Portal, which all present a sequence of similar puzzles differing only in complexity and scale. However, to the true puzzle connoisseur, the height of the art is to be found in the adventure game.
The Monkey Island series, the Infocom interactive fiction games, and escape-the-room games portray worlds in which puzzles are a fact of life. The developers try, with various degrees of success, to incorporate puzzles seamlessly into the game world, so that players can inhabit a character who thinks her way around difficulties rather than shooting her way through them.
In the adventure game category, few games are more maligned than Cyan Worlds' Myst series. Spanning five single-player games and one repeatedly-resurrected online game, Uru, the Myst series is often blamed for bringing about the death of the adventure game. It popularized the concept of a silent, faceless protagonist exploring an uninhabited game world, and led to a myriad of copycat games where the atmosphere was spooky but the puzzles were arbitrary and banal. These concepts were anathema to fans of the character depth and humor of the Sierra and LucasArts adventures.
However, it really is the puzzle-solving where the Myst series shines. By looking at the best (and worst!) puzzles of the series, we can gain an understanding of how to construct truly compelling intellectual challenges. For a puzzle to be effective, it must be three things: fair, novel, and integrated.
A Fleeting Glimpse
First, let's look at my three favorite puzzles from the Myst series. These are Riven's marble puzzle, Myst III: Exile's Amateria Age, and Uru: The Path of the Shell's Ahnonay Age. All three require observation and logic, then reward the player with an "a-ha!" reveal. Below, I'll discuss the setup and solution to each puzzle.
The marble puzzle is the trickiest challenge in Riven, the second game in the series. In order to restore power to the mystical linking books that lead to the villain's haven, the player must enter a combination into a device on Dome Island. Luckily, the villain has an unreliable memory, and has based the solution on the geography of the game world. The player must place five marbles in a grid, with each marble's color and location matching that of a small dome encountered over the course of the game.
Part of this puzzle's genius is that it brings together all of the challenges surpassed so far in the game. In order to know the locations of the small domes, the player must have solved all of the earlier puzzles which provide access to the domes. Discovering the associated colors requires the player to have determined which symbol matches each color, and which dome matches each symbol.
At a few points in the solution process, a process of elimination is required to account for damaged machinery. The solution requires a full exploration of the game world, over five leaps of associative logic, and the use of several devices to process and sort the proper information.
Trial and error would get the player nowhere with this puzzle; there are thousands of possible combinations, and only one correct one. The puzzle is fiendishly difficult, but it isn't unfair; every aspect of the solution is prominently clued somewhere in the game, and all the player needs to do is put the clues together. Many players, including myself, required more than one hint to solve the puzzle, but my reaction to the solution was "Why didn't I see that?" rather than "How could I have been expected to see that?"
Amateria is one of three main puzzle worlds in Myst III: Exile. It is an island with hexagonal geology and enormous crystal spheres that roll along metal tracks. In a series of preliminary puzzles, the player must get spheres to traverse short patches of tracks, using simple logic to disable obstacles. This provides access to a central control station, where the player discovers that the three short stretches of tracks are connected, and must arrange a route so that a sphere can roll through all three tracks in sequence.
Finally, when everything is set up properly, the player discovers that the tracks are more than puzzle elements; they are a transportation system. The reward for solving the puzzle is a roller-coaster ride through each of Amateria's puzzles, finishing at the part of the world that the player has been trying to reach the whole time.
Like Riven's marble puzzle, Amateria's final puzzle brings together each of the previous puzzles of the Age. The player must have solved each puzzle and understood its solution in order to complete the challenge. Beyond that, the puzzle is based on real-world principles. Amateria does not just ask for a secret password which magically opens a door; the player is constructing a path for herself to take to her goal. Beyond that, the reward of the roller-coaster cutscene is one of the most exciting and good-looking sequences of the series.
Finally, Ahnonay is one of the worlds from the Uru expansion, The Path of the Shell, as well as being included in a modified form in the short-lived, GameTap-sponsored Myst Online: Uru Live. This Age was constructed by a charlatan named Kadish to fool people into believing he had power over time travel.
When the player first enters the Age, it has been abandoned for some time; she must use a description of how the Age used to work in order to experience Kadish's illusion of time travel. By experimenting with objects in the world, the player finds that Ahnonay is nothing more than a soundstage that is somehow being transformed from one "time" to another.
Diagrams within Ahnonay suggest a fourth "time" that is not part of the usual tour. By combining the player's knowledge of how travel between worlds works with the controls found behind the scenes, the player can reach an observation room outside of the entire structure, where the true nature of the world is revealed.
There's nothing mystical about Kadish's trickery. He simply built a huge device which rotates each "time zone" into place in preparation for a group's arrival. There are four separate soundstages, each one able to be rotated in and out of the active position.
The player's progress through Ahnonay works on two levels. As the player solves small puzzles within the Age, her understanding of the mechanisms behind the world grows. At the beginning, Kadish's illusion is powerful, and the puzzles within seem impossible. By the end, the player has solved all of the puzzles and understands the mechanism behind the Age's trickery.
Each of these puzzles — Riven's marble puzzle, Myst III: Exile's Amateria Age, and Uru: The Path of the Shell's Ahnonay Age — is memorable and compelling because of certain qualities they share. By looking at these qualities, we can better understand how to construct equally effective puzzles.
Not Yet Been Written
For a puzzle to be a good one, it should be fair, novel, and integrated. A fair puzzle is one which can be solved using information and tools provided in the game, without depending on guesswork, specialized knowledge, or a walkthrough. A novel puzzle is one which is unique to the game, not a reskinned version of one the player has seen before. An integrated puzzle is one which can believably exist in the game world, and which has appropriate in-game consequences.
For a puzzle to be fair, it should be possible to solve it using only information provided in the game and accompanying materials. It's fine for a puzzle to require things like basic math and logic skills or the fundamentals of gravity and electricity, but more specialized knowledge should either be avoided or provided in an in-game text.
In Graham Nelson's "The Craft of Adventure," he cites Zork II's diamond maze as an example of an unfair puzzle; it requires knowledge of baseball, which is uncommon outside of America. This is a mild offense, but imagine a game which assumed knowledge of medieval heraldry or Indian Carnatic music. In the worst case, a player might not even notice a puzzle existed, which would mean she wouldn't know to look up the required information in an external source.
Additionally, a puzzle can be unfair because it requires undue guesswork. A puzzle solution that only makes sense after the fact is unfair, as is an unclued, arbitrary solution like Dare to Dream's use of a fish tail as a key for a door. Generally speaking, if a walkthrough is the only way to figure out a solution, then the developer is being far too clever. There should be a continuous trail of logic from the clues to the solution.
Novelty is also important in puzzles. At some point, the boxes-on-a-slippery-surface puzzle was novel. Not anymore. Any experienced player who encounters one of these puzzles knows the way to solve it; she just needs to carry out the solution algorithm. There's no intellectual challenge, which makes the puzzle nothing more than a time-waster. The same goes for a sliding tile puzzle (or "fifteen" puzzle). As a general rule, if a puzzle is included with the default installation of your operating system, it's not novel.
Novelty is important because the fun and challenge of a puzzle is in discovering its solution. Just implementing a solution the player already knows is pointless. Nor is it enough to just dress an old puzzle up in new clothes. A Tower of Hanoi puzzle is not made new again by making the disks precious coins or UFOs. It might be acceptable to do something clever like make the disks actual floors of a tower that could be navigated in different configurations... but only as a last resort.
The final requirement is integration. Players are perfectly willing to accept a world full of absent-minded folks who write down the combination to their luggage in code, but there are limits to their credulity. Take 7th Guest, which is famous for its puzzle in which rearranging soup cans opens a door. This is a game where ghosts are evidently murdering other ghosts, and the puzzle is still hard to believe. At the very least, a puzzle should make sense in the game world.
More than that, though, a puzzle should fit. Most of Myst's puzzles suit the game world; there is a reason for them beyond the game designer's need to challenge the player. The series' use of Ages, or dimensions, helps with this by providing a strong theme for sections of the game. Channelwood's puzzles are related to the water-based power system used there, while Kadish Tolesa's puzzles all feel like they were created by a conceited and slightly crazed con man.
Fairness, novelty, and integration are all important for making good puzzles, but what makes the puzzles discussed here stand out is their payoffs. Each of them has a moment of epiphany, where the odd structure of the game world suddenly makes sense, and all the pieces fit together. Because they are especially tricky, the reward is sweeter, and the player feels a greater sense of accomplishment.
It's this puzzle with a twist that is the greatest strength of the Myst series. By starting with a solid puzzle and then adding a twist, developers can make their own puzzles stick in players' memories. This doesn't just apply to adventure games, either; any kind of game that uses puzzles will be better-served by a cleverly crafted obstacle rather than just another game of Nim.
[Gregory Weir is a writer, game developer (The Majesty Of Colors), and software programmer. He maintains Ludus Novus, a podcast and accompanying blog dedicated to the art of interaction. He can be reached at Gregory.Weir@gmail.com.]
Categories: Column: The Interactive Palette








12 Comments
Very nice read. I love the Myst series. But your intro suggested that we'd also be looking at some of the worst puzzles of the series, and I'm curious which ones you would put on that list.
I love the game's puzzles for the most part, so when it comes to bad ones, two terrible puzzles really stand out:
Bad: The final "Dream" puzzle in Myst 4. It's like a massively convoluted version of the Tower of Hanoi with such obnoxious and arbitrary rules that even after I had figured it out it still frustrated me.
Worst: The firefly puzzle in Uru. Check the link I put as my URL, I created a detailed Q&A walkthrough for the puzzle to illustrate its absurdity.
LuigiHann | February 20, 2009 10:21 AM
Yes... the article was long enough that I left out some of the worsts, and forgot to change the intro.
Other worsts: the incredibly fiddly organ lock for the rocketship in Myst, the sound-maze in Myst, Revelation's carry-the-elements puzzle, and Uru: Path of the Shell's do-nothing-for-fifteen-minutes puzzle.
I got most annoyed, I think, at the puzzle in Revelation's Haven where you have to transport cannonballs. It arbitrarily restricts you from leaving a cannonball at the upper floor, which would make the solution much less tedious.
Gregory Weir | February 20, 2009 12:44 PM
"[I]magine a game which assumed knowledge of medieval heraldry or Indian Carnatic music"
Do you mean this would be a strong offense? If the game environment contained clues that enabled a player to learn these details, even by looking them up elsewhere, then such puzzles would be fair and appropriate -- even if obscure.
Dennis G. Jerz | February 20, 2009 8:09 PM
The Dream puzzle, and indeed the entire existence of Serenia and Dream in Myst IV, is a blight upon the face of the Myst series.
The firefly puzzle isn't so bad, I don't think. It at least would be easier to figure out (and make more sense) if the fireflies stuck to your avatar, as they were originally intended to (time constraints made that not happen).
The Selenitic organ isn't so bad, either, because even if you're tone deaf, you can count the number of half-steps from the bottom of the slider.
I don't think the sound maze is bad, either, because it teaches you what the sounds mean. The very first time you move you can only go north and it plays the north sound. The next time you can only go north and (say) west, and it plays the west sound. It teaches you at the beginning so you don't have to do a whole whole bunch of trial and error.
The do-nothing-for-fifteen-minutes puzzle is pretty terrible, you're right.
I don't remember the cannonball puzzle, but I know I generally didn't like Myst IV very much.
One of the puzzles I hate most is the "stand on volcanic vents and let the boiling-hot steam propel you up in the air" puzzle. I suppose that's not "bad" the way these others are bad, because it is novel, integrated, and (I suppose) fair. But it's also ridiculous and stupid. Back in beta testing I defended that puzzle against people who said what I'm saying now, but in retrospect, they were completely right.
75th Trombone | February 20, 2009 10:08 PM
Dennis: Gregory allowed for the very exceptions that you mentioned... if it is well hinted at and available for reference you can use all sorts of obscure knowledge for the basis of puzzles. You can fold that back in with Gregory's mention of the Fire Marble puzzle. One of the facets of that puzzle was the system of symbols used to mark things in the game. You could get into the obscure technical details of how and why the symbols work, but the game also provides several good opportunities to learn them and even still mostly makes sense if you don't catch most of the subtle "ah-ha!" stuff.
(Which several of the later games went on to test and re-test, offering wonderful integration across the series as a whole...)
Gregory: This was a nice article to read. Thanks.
The organ lock puzzle in Myst was one of the last that I ever solved on the main island. I quit Myst 4 during its own organ puzzle, and haven't made it back to seeing any of the rest of the game yet.
Also, which waiting puzzle are you talking about in PotS? I think the oven puzzle was fair and unique (but then when I first played PotS I managed to do it first, so I was busy beating Ahnonay while the pellets were in the oven). As for the other one, it served one purpose: reminding us that the puzzles were originally built for multiple players. It needed a stronger redesign (rather than, "oh, we'll just bend time") or should have provided for the old adventure stand-by of swapping between avatars.
Max Battcher | February 20, 2009 10:23 PM
I've long loved the Myst games, and certainly don't think they're the reason the adventure genre has suffered over the last decade. If they are, it's because other game designers struggle to match the quality, internal consistency, and production values of the Myst games.
But I too despaired of the Dream puzzle. Not one for the record books. Unlike most Myst puzzles, there was no sensible way to go back out of the environment to look for more clues as to how to solve it. No way to observe from a distance. Too obscure, even for Myst.
Ben | February 21, 2009 4:11 AM
Max: The waiting puzzle I'm referring to is the one that requires standing in one place for an extended period of time just to open a door. The one in the Great Tree Pub.
Gregory Weir | February 21, 2009 9:03 AM
Yes yes YES!
Everything you are saying about what makes a good puzzle in a game is spot on. I think the factors of immersion, logic and novelty are what set truly innovative games apart.
Plugged in "oh we need a challenge here" puzzles do not an adventure game make. Half the games plugged as 'adventure' these days are like that. Tangental peeve: hidden picture games are not an adventure!
And dear ghu, make the items make sense! If it could work, it should! I've played too many games where the hammer in my inventory won't break the crumbling wall, but the broken pipe I find later on will. Grr.
Bad:
~The berry/lemming puzzle in Syberia 2.
~The Build Your Own Trap puzzle in The Dig.
Utterly Unsung Puzzle Game That Everyone Who Loves Puzzles Should Play:
The Stone. All the puzzles are still there, even if the underlying Metagame aspect isn't.
>>If the game environment contained clues that enabled a player to learn these details, even by looking them up elsewhere, then such puzzles would be fair and appropriate -- even if obscure.<<
I think that would be ok, as long as there's enough clues in game, that the solution is a quick dash to Google.
Eleri Hamilton | February 21, 2009 10:13 AM
Gregory: Oh wow, I forgot that third variation on that. I have no excuse at all for that one. It didn't even make sense from the multiplayer standpoint thing... That area didn't even reappear in that form in MO:UL, which is probably why I seem to have forgotten it.
Max Battcher | February 21, 2009 3:39 PM
Eleri: hidden picture games are a perfectly valid form of gameplay, and I like how the Mystery Case Files series has mixed in more complex puzzles... but I agree that they're not quite the same.
And, yes, the Stone is awesome.
Max: Sadly, it didn't have a chance to appear in MO:UL, regardless of form. I can only assume that we would have been able to do that puzzle in some way, but the door to that room never opened before the game was canned.
Gregory Weir | February 21, 2009 4:06 PM
*nodnod* the MCF games are exampled of Hidden Pictures Well Done. Reminds me of the back of my old Highlights magazines when I was a kid. I like the hidden picture games where they've at least tried to integrate the objects into the theme of the game, rather than just randomly scattering objects around a background.
I'm one of those creepy people who didn't think the Waiting Game was that icky in Uru. I'm betting it would have been very different in Uru Live. *sigh*
I'm also a heretic in that I was irritated by the puzzles in Riven. Lots of back-and-forthing, actions that didn't have a result until much later in the game. If I do something, I wanna know that it *did* something.
Which, I suppose, is utterly contradictory to not minding the waiting puzzle.
So, do you think there's anything out there now that comes close to having puzzles that meet the 'good puzzle' criteria?
Eleri | February 23, 2009 12:47 AM
I don't remember any canon balls either....Hmmmm....
I would like to say that I found the sound maze quite interesting in Myst and although I found a few of the other puzzles in the series to be borederline in fairness, I had to draw the line at the fifteen minute wait in Uru. I never found any very obvious clue, even in retrospect and was so unlike anything else in the series that that I thought it was a bit much.
As for carrying the elements about in that stone maze, the less said the better. I was doing it on a Dell laptop on which it was quite impossible without the downloaded fix, after which it was fine and that leads to one other criterion, perhaps inevitable with changing hardware, the puzzle has to be possible with the available machine. I have run across a few that I am convinced simply cannot be done on the machine I was using.
James Stewart | February 25, 2009 9:40 PM