Carl Muckenhoupt on Best Puzzles 2012

Carl Muckenhoupt writes on Counterfeit Monkey, Bigger Than You Think, Dinner Bell, Endless, Nameless, In a Manor of Speaking and Shuffling Around.

Counterfeit Monkey, by Emily Short

Counterfeit Monkey, by Emily Short, takes a single puzzle and turns it into a world. The puzzle it’s based on is the one with the “T-remover” from Steve Meretzky’s Leather Goddesses of Phobos, a device capable of removing the letter T from objects, turning a twig into a wig or a poster into a poser. Take that and extend it: give the player a device capable of removing any letter at all, provided that the result is still a valid word. Then — the most difficult trick — instead of simply making a nonsensical puzzle environment around it, try to take the idea seriously. Give some thought to the sort of effects this technology would have on the world, the politics of control over it, the economics of making cars out of cards, the resurgence of interest in language reform as a matter of not just social engineering but immediate practical effect. That’s what Counterfeit Monkey gives us.

But I’m here to discuss the puzzles, not the story. The puzzles, though, are those of a storyteller. Just as there’s an arc to the plot, there’s a discernable arc to the puzzle content.

It starts, naturally enough, with tutorial-like ease. One early gating puzzle can be solved with a severed body part, and there’s an abundance of things in the vicinity that can be turned into body parts by removing a letter (a pear, some tomes, etc.) The game starts requiring some very specific transformations immediately after that, but it generally allows multiple solutions wherever it can; throughout the game, there are puzzles that will accept anything from a general class of objects, such as when bar patrons bet that you can’t create something smaller than a pebble. Sometimes you already have the object you need from just playing with the device earlier, transforming objects simply because you can. Other times, the sudden need for a thing is what makes you notice that it’s an option.

As you progress, you get more options, both through your growing pile of transformable objects and through measured access to new devices to transform them in new ways. These devices help to keep the puzzling fresh by periodically endowing your inventory with new potential, but there are usually limits on their use. In one place, a bartender uses a homonym paddle to wow the tourists by making drinks from literal screwdrivers and rusty nails. You can ask her to paddle objects for you, but she can refuse if she finds the request dangerous or unethical, and besides, the mere fact that you have to bring her the objects limits you to things you can carry. Elsewhere, there’s a machine for combining two words into one. It doesn’t have the bartender’s ethical constraints, but isn’t any more portable. So the letter-remover remains your primary weapon simply because it’s the one you can bring with you wherever you go and apply to immobile room features.

And that’s part of why it’s such a big moment when the letter-remover’s constraints come off. There are two such constraints from the start of the game: you can’t use it to create sentient beings, and you can only use it to create physical objects rather than abstractions. The thing is, these are both reasonable as technical limitations of the game engine. Convincing NPC behavior is difficult to produce at the best of times, so when we find out that we can’t just create random persons on the fly, we cut the author some slack. And putting “Tao” in the game as an object, or a “par”, or simply “it” devoid of antecedent, well, how we can complain about the absence of something we can’t even imagine? (As if a text-based work needs to confine itself to the imaginable!) The point is, it’s easy to see the sense of these as necessary limitations, so it comes as a shock when we reach the point, rather far into the game, when we learn they’re not. And not just a shock: it comes as an immense liberation. Wacky ideas we thought of earlier, and which were rejected, suddenly become pursuable.

And I think it’s worth emphasizing that “wacky”. Despite what I said about taking the premise seriously, the wackiness does inevitably come out. In the endgame, there’s a point of no return in the bowels of a secret compound where the government keeps the really powerful word-manipulation equipment, such as the experimental T-inserter. Some of the scenery there is wacky as a security feature: an unusual object such as a solid gold lock or a ladder made from human teeth can’t be adequately described in a single word, and that makes it harder to transform. But then you get the anagramming gun. Longer decriptions are more likely to be anagrammable, not less, and the anagrams are very likely to be odd things like a “monk corpse bonnet” or a “rodeo beaker”, because that’s how anagrams are.

Now, the whole progression here is one of increasing power, but this doesn’t make things easier. On the contrary: pretty much everything you’re able to make in the intro area is useful, but the set of things you can make grows far faster than the set of things you can solve puzzles with, drowning intent in possibility. The anagramming gun is a turning point in this regard. Throughout the game, you’ve been making deliberate changes to objects. Even when you zap objects just for the hell of it, you generally have a target word in mind, because removing letters from things at random usually doesn’t work. The T-inserter is a little less predictable, because sometimes a word can have a T inserted in more than one place to produce different words (and in fact there’s a puzzle around this), but still, you generally have some idea of at least one place where you expect it to insert a T. The anagramming gun, on the other hand, you mostly use with no idea of what you’re going to get. So, for the first time in the game, I was just zapping stuff willy-nilly, not even knowing if it was going to have any effect or not.

In fact, the endgame takes this even further. There’s one device with effects you don’t know at first, and which you have to figure out by experiment. Another device takes this to an even greater extreme: the umlaut punch. This is a device that adds umlauts to words. Since the game isn’t in German, there is no object that becomes a valid word when umlauted. It will, however, temporarily turn any object whatsoever into a heavy metal band, which is useful for reasons I won’t go into here. This is something the player has to discover by trying out a device, not just without knowing what the result will be, not just not knowing if it will work at all, but with a very reasonable expectation that it will not work. Such is the puzzle arc of this game. The deeper you get, the less you’re in control.

Beyond that, it’s worth noting that the game is large, and has a considerable amount of content off the critical path, including some optional achievements. One in particular is notable for the way it works into the puzzle themes I’ve just described: an achievement for performing successful removals on every letter of the alphabet. The last letter you can remove is Q, from a squid that shows up during your escape from the endgame complex. I didn’t know there was a word “suid” in the English language when I encountered this. I just knew it was my last opportunity to nab a Q.

Bigger Than You Think, by Andrew Plotkin

Andrew Plotkin’s Bigger Than You Think is an iterated cave-crawl. Although capable of receiving typed commands, it’s essentially hypertext in form: it consists of a tree of nodes, each node having some highlighted keywords that you can select, by clicking or by typing, to advance to a different node. The one enhancement it puts on top of this is an inventory, which is to say, a set of keywords that stick around and which you can attempt to apply at any node to effect permanent changes. Some of the inventory items are physical objects (a rope, a crowbar, a mysterious crystal), others are more abstract (a word, knowledge of a system of runes), but each is, figuratively, a key that opens a door, opening access to more nodes.

So its nomination for the Best Puzzles award is a little surprising. Every single puzzle in the game is a “use inventory item in the right place” puzzle, without even any way to specify how you want to use it. Additionally, every object you can pick up is the solution to exactly one puzzle, and every puzzle is solved with exactly one object. This is very old-school design, the sort you see in games from the 1970s by authors who hadn’t yet figured out that there were other options. But Bigger Than You Think pulls some tricks that add a little interest.

It has to do with the work’s form. It is, as I said, a tree. No loops. You start at the beginning, you choose various branches to go down, and before long you hit a leaf node and have no choice but to start over. Except you don’t start over completely from scratch: you keep your inventory, even if the story has to come up with a new explanation for why you have these items at the beginning of your adventure. In one branch, you meet a man who tells you stories, nested within the frame-story. These are also interactive, and you keep your inventory through them as well, bringing items both into and out from the sub-story. This is contrary enough to expectation that the realization that it’s possible at all is an “Aha!” moment.

Furthermore, that bit about having no option but to start over when you hit a leaf node was a lie. There are nodes that seem like endings, or even death messages, but which still afford the opportunity to use an inventory item to save yourself. This is really just a consequence of the nature of the work. Formally, there’s no distinction between “rooms” and “events” in hypertext, like there would be in a conventional text adventure. It’s all just nodes. The only thing that makes an end node an end node is a lack of forward links. But because the content is so adventure-gamey, it’s easy to forget this and just restart whenever it looks like the current run is over. The strongest example of this, a node that sees you falling into a deep pit, was so effective at fooling players with its apparent finality that it won the award for Best Individual Puzzle. The illusion is helped by the way that the interface keeps the inventory links at the top of the page, away from the new story text, where it can escape attention.

Above all, though, what the puzzles in this piece do is demonstrate that this merging of hypertext and old-school inventory puzzles is possible at all, and can work about as well as they did in the cave-crawls of yore. The solver’s thought process is no different; it’s all either “Here’s an obstacle I haven’t seen before, let’s take a look at what I’m carrying to see if I have a way past it,” or, conversely, “Here’s a thing I can take with me, wait, I remember a place where this would be useful! Let’s go back there!” (even if “go back there” means starting the story over). For puzzles such as these, perhaps the parser was always an unnecessary distraction. This is a particularly surprising conclusion for a game by Plotkin, who has stated (in an interview just a few months prior to writing this piece) that IF is “all about the parser” and “the suggestive, but not delimited, range of choices that a free-text prompt offers”. But then, he’s also an author known for challenging assumptions; try to tell him what IF can’t do or must be, and he’ll treat it as a dare.

Dinner Bell, by Jenni Polodna

In 2012, a group of IF enthusiasts celebrated the 20th anniversary the the They Might Be Giants album Apollo 18 by creating an IF “tribute album”, Apollo 18+20, with one game for each song the album. Jenni Polodna’s Dinner Bell was among them. The song that inspired it is one of the quirkier and more enigmatic songs by a quirky and enigmatic band: its lyrics mainly consist of a litany of unwanted foodstuffs — “I don’t want a pizza, I don’t want a piece of peanut brittle, I don’t want a pear” — to an upbeat tune while a less-discernable voice mutters ominously about Pavlovian experiments. How do you turn this into IF? Apparently you do it by making a treasure hunt out of it.

Set in an absurdist laboratory mock-up of a kitchen, the game makes you find each of the sixteen food items referenced in the song. Although you’re desperately hungry, your operant conditioning makes you automatically deposit each item into a grocery bag immediately on acquiring it. Because the entire game takes place in a single room, the hunt consists mainly of ransacking fixtures (such as the oven and the refrigerator and various drawers) and examining things. There are a few puzzles beyond that: you need to find a light source and a means of handling hot items, figure out what to do with a substance called “shiptogar”, and — the most puzzlesome bit — identify the sole real pear in a bucket of wax pears. None of these things are difficult in themselves; the pears are the only part that really requires insight rather than diligence. Nonetheless, it’s possible to get stuck simply because there’s such a wealth of detail in the room that it’s easy to neglect bits, either by losing track of what you have and haven’t examined or by mistakenly assuming that a detail is unimportant. An admirably subtle in-game hint system helps with this, telling you what part of the room the experimenter keeps glancing at like Clever Hans’s trainer.

The important thing for a game like this is that it can’t afford to let the process of thoroughly searching everything in the room get boring. Keeping it short is a good first step, and filling the room with amusing detail also helps: even when you don’t find anything useful, you’ll discover things like tiny fire-breathing dinosaurs battling it out with the scrubbing bubbles in the sink. So when you actually discover a food item, it feels like a bonus. Better yet, you might find two or three together! This might have just been a consequence of the song having more food items than the game has hiding places, but it’s effective nonetheless, making the moments of greater discovery seem more special, and giving the game a sense of lavish abundance (somewhat ironically, given the starving protagonist). It’s a short game, but a full one.

Endless, Nameless, by Adam Cadre

Adam Cadre’s Endless, Nameless is a little bit nostalgia and a little bit Hindu mysticism and a little bit roman a clef about the Interactive Fiction community, and also something of an exercise in redeeming the terrible. This a game that occurs at two levels. First, there’s the inner game, Nameless Quest, a simulated 1980s fantasy text adventure on a simulated BBS, with parodically amateurish writing and some vestiges of a D&D-like stat and level system from when it used to be a combat-oriented MUD. Whenever you die in this game, you wake up in the outer game, an afterlife — or, as one character calls it, “interlife” — which is sort of a Wreck-It-Ralph-like backstage area where you can hang out and talk with other iterations of your hero character from earlier revisions of the game before re-entering it. These two layers differ in their writing style, level of detail, and even their set of available verbs: the inner game initially supports UNDO but not TALK, the afterlife is the other way around.

But the biggest difference between the two levels — or at least, the one that’s relevant for this writeup — is the puzzle content. Aside from the plot-level “what the heck is going on and what does it all mean” aspect, which is a major part of the game as a whole, there’s basically only one puzzle to be solved in the afterlife area, and that’s how to get back to the inner game, where the real puzzles are. And those puzzles you’re trying to get back to? Taken at face value, they’re mostly pretty bad ones. You can lock yourself out of victory on turn one (and on quite a few turns thereafter), crucial items aren’t noticeable enough, and most importantly, solutions sometimes require unreasonable leaps of intuition without adequate cluing. Sometimes a puzzle requires more than one such leap: there’s one puzzle that requires you to realize that (a) you can use two spell-granting potions simultaneously to create a combined spell and (b) a spell that only works on “beings capable of communicating by sound” can be cast on a machine that beeps. I’m talking real read-the-author’s-mind stuff.

Except that you don’t have to read the author’s mind. You have an entire afterlife full of alternate selves with varying degrees of experience with the game just waiting to help you. At first I was reluctant to seek their advice, regarding them as just an elaborate hint system that would spoil the experience, even though that doesn’t make much sense given how much effort had been put into making the experience of the inner game unpleasant. But ultimately, these people really are part of the game, and asking them about all the unfair puzzles really is part of the intended experience of solving them. And anyway, you don’t need to worry about spoilers when most of them have inaccurate information.

That’s the crux of the puzzle-solving here. The alternate heroes are all from older versions of the game, when it was quite different. Some of the things they remember have survived revision, some haven’t. But that doesn’t mean that their out-of-date intel is useless. It just means you’ve got a different puzzle to solve, one of extracting the pertinent information from what sounds like a dead end. For example, when one guy talks about scooping liquid fuel from a machine and carrying it home in his shield from the mountain kingdom of the dwarves who aren’t actually in the game any more, your take-away should be “my shield can be used as a bowl”. This is the level of cluing that makes those intuitive leaps possible, and it hinges on the realizaton that cluing is taking place at all.

One other thing I think is worth mentioning: the use of time limits and inventory limits. The game’s spellcasting system involves temporarily charging yourself up with magical energy by essentially injecting yourself with a syringe-like magic wand. The effect only lasts a few turns, and in some cases the timing is tight, because have to fiddle with your inventory in order to do something like pick up a bulky item that you couldn’t carry at the same time as the wand. Inventory limits are passe these days, mainly because they’re a bad idea for most stories; they force the player to pay attention to something irrelevant to what the game is trying to be about. But this game really is to some extent about inventory limits, because it’s to some extent about text adventures from the 1980s. The great part of this is that here, for once, the inventory limit isn’t just an arbitrary inconvenience. It’s part of the puzzle content, something that you have to think your way around at some of the most dramatic moments in the game.

In a Manor of Speaking, by Hulk Handsome

2012 was a big year for games based on wordplay, a theme that text-based IF is singularly well-suited for. Fully half of the Xyzzy Best Puzzles nominees were wordplay games, including the winner. But unlike the others, Hulk Handsome’s In a Manor of Speaking centers on semantic rather than syntactic play. The whole thing is based on puns, double meanings, and figures of speech taken literally, from the names of rooms to the randomly-occurring flavor text. They just pervade the thing. It’s all very jolly and whimsical. The world is disjointed and nonsensical in the best way, and the story is little more than a vehicle for pun delivery. It is, in short, an eclectic puzzle environment of the old school.

In fact, it’s striking how old-fasioned the interactivity and puzzle design is. Just as in Bigger Than You Think, every object you can pick up has only one use, and using it successfully uses it up, removing it from your inventory. There are numerous unsignalled one-move deaths, although they’re entirely without consequence because you can just UNDO them. There’s even a Scott-Adams-style per-room hint system, made possible by the way that there’s never more than one puzzle per room or more than one room per puzzle.

Now, I say that the game is pun-based, but it would be exaggerating to say that puns are the focus of the puzzles in the same way that anagrams are the focus of Shuffling Around. Contextually, yes, the puzzles are grounded in wordplay-oriented situations, but that doesn’t always make them wordplay puzzles. When a tourist asks you to take his picture, but there’s no sign of a camera anywhere, the realization that he doesn’t want you to photograph him but is literally trying to give away a picture is something that requires the player to think about the words. But selling the picture to an art dealer later on does not involve word-based thought, even if the art dealer does give you a thousand words in exchange for it. There are eighteen actions that increase your score, and by my count, only six or seven of them require you to supply the pun yourself, rather than slipping it into the output afterward. Most of the rest are straightforward application of inventory items, such as slotting a bar into a groove that’s clearly there to receive it or giving a plot token to an NPC. One bit almost seems like an anti-word-puzzle, a subversion of the premise: when a stranger yells “Duck!”, he really is just saying you need to duck to avoid being struck by something, rather than making a pun about waterfowl.

On the other hand, even if not all of the puzzles require punnish thought, it’s important that, from the player’s perspective, enough of them do that any of them could. This is enough to keep the solver’s mind in wordplay mode.

Anyway, puzzles in IF are made of more than just the process of solving. The consequences are also part of the act of discovery that a puzzle represents. Or at least, the consequences definitely have an impact on how enjoyable the puzzle is, and that’s where this game shines. The puzzles here may not be the deepest or the most difficult, but they’re fun to play with, simply because it’s fun to successfully interact with this pseudo-illogical world and see what it comes up with next, whether in response to your valid solutions or your failed attempts.

Shuffling Around, by Andrew Schultz

Shuffling Around, by Andrew Schultz (under the pseudonym Ned Yompus), is another wordplay game, this time focused on anagrams, and focused on them with singular laser-like intensity. The content is basically built out of anagrams like bricks in a wall. All significant outputs from the game, and many insignificant ones, contain offhand anagram pairs like “buckets be stuck here” or “you’ll startle a starlet”, sometimes multiple pairs per sentence. Even room names are mostly formed from anagram pairs, like “Sacred Cedars” and “Cruel Ones’ Enclosure”; in the few cases where they’re not, they’re anagrams of something else, like when you’re tossed in The Nick and escape to the Kitchen.

In this world, you alone have the strange ability to transform any anagrammable object into its anagram simply by typing its new name, creating a door out of an odor or a missile out of some smilies. This is the game’s primary puzzle mechanic. Many of the puzzles involve some object manipulation in addition to the anagramming, and in a few cases the uses of objects are themselves non-obvious (for example, an empty ketchup bottle in a monster’s lair is supposed to hint at its fondness for tomatoes), but every single puzzle in the game has some anagramming component (the tomato itself has to be formed out of a motto). Graham Nelson famously described text adventures as “a novel at war with a crossword puzzle”, but the emphasis here on creating words by thinking their letters makes it feel especially crossword-like.

Perhaps surprisingly, there’s still room for different solving techniques. Sometimes an anagram was so obvious that I transformed the object as soon as I found it, without knowing why I needed the transformed version. Sometimes I only figured out anagrams by knowing what sort of object a puzzle required, such as when I needed a weapon and had some noughts in my inventory. Sometimes it took an extra nudge from oject descriptions, like when a statue of a spearman is described as “cheesy”, hinting at the solution “parmesan”. I suppose it’s all analogous to the techniques used to solve the more typical adventure-game puzzles — doing things because you can vs. following narrative hints vs. reasoning from goals.

The game is divided into an intro area, a hub from which you can enter three main chapters in any order, and an epilogue. Each of these subsections is a self-contained puzzlebox environment, relating to the others only through the structure I’ve just described. You’re only required to complete two of the three main chapters before entering the epilogue, but I’d be surprised if many people took that option, as anyone who’s enjoying the game enough to complete two chapters will no doubt want to tackle the third. The intro area is particularly notable for guiding the player towards the realization that you can transform objects without stating it explicitly. Discovering the game’s central mechanic is its first puzzle, and it’s one of the parmesan-like lots-of-contextual-hints ones.

There are a great many anagrams to be found. The maximum possible score is 79 points, awarded one point at a time, and all but a few of those points are for finding anagrams. The game politely makes the completist’s hunt for last lousy points more manageable by reporting on the score for each sub-section separately: “You have scored 4 out of 5 points total for the Intro region. You have scored 16 out of 16 points total for the Forest region…” This is an excellent way of providing extra information to those who want it.

And in fact “extra information to those who want it” could be the game’s motto. It’s got the usual adaptive hint system, giving you measured prods in the right direction on request, provided you ignore the warning that they could spoil the fun. It’s got a one-use item that will just outright unscramble one word for you, which feels entirely like cheating. But best of all, it’s got impressively sophisticated in-game anagramming assistance in the form of gadgets you can use to scan objects for partial information about their transformations. You get to choose one of two such gadgets during the intro chapter: one shows you the positions of the first and last letters in the transformed name, the other shows you which letters are already in the correct positions, Mastermind-style. Either one will let you know whether the scanned object can be transformed at all, which can be a considerable help all by itself. The great thing about these gadgets is that, although they’re every bit as much a source of hints as the HINT command, they don’t feel at all like cheating. I think this is because, unlike a normal hint system, there isn’t even a possibility of accidentally getting more information than you want from them. The final realization of what order the letters need to go in is always left to the player, and with it, the solver’s satisfaction. This is an effect that other puzzle-based games would do well to imitate in their hint systems, if possible. Unfortunately, it seems like it only works here due to the puzzle content’s extreme regularity of form.

One thought on “Carl Muckenhoupt on Best Puzzles 2012

  1. Pingback: The Stack » Xyzzy Awards Best Puzzles Writeups

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>