True! I’m actually playing Anchorhead right now as part of my plan to replay all the games with 100+ ratings on IFDB in reverse chronological order. In the new version, I had to LOOK BEHIND something, but I was explicitly told ‘there is something behind _____’. So cluing helps a lot!
5 Likes
I dislike “brain teaser” type puzzles of the sort found in 1970s truck stops. So doing something like a Towers of Hanoi is bad. I’ve been critical of Spellbreaker’s scale puzzle, but the twist (light or dim unknown) makes it better. I realized during my most recent playthrough that it seemed just interesting enough to pass muster.
I don’t really like making rules about what authors can and can’t do. I’d rather emphasize the importance of doing things well. By making inviolate rules, we can give ourselves a reason to dismiss things without trying them or even without thinking about them. It’s good to formulate principles, but in the end, we must consider specifics. Hence, the scale puzzle is fine, while Zork Zero’s multiple towers of Hanoi is not.
No puzzle that requires an action can be good if there is never a reason for the player to perform the action. That could apply to SEARCH, PUSH, TAKE, TABULATE, or even EQUIVOCATE. We could play whack-a-mole all day with individual verbs, but the problem is almost always that the author did not provide a motive (not even a cleverly disguised one) for doing what was needed.
When I learn the solution to a puzzle that I could not solve, I consider the available information. If there was enough information, or if I disregarded valid hints, I give the author their due. I missed it. However, if there wasn’t enough info, I feel annoyed. It seems cheap. I grow suspicious of the author. What else have they been keeping from me?
I like to separate bad puzzles from noble failures. Sometimes an author is creative; they try something new. There is a desire to move the medium forward. If they fail, it is a noble failure. “Bad” applies to the other failures.
- The Bank of Zork puzzle in mainframe Zork was an attempt to move things forward (noble failure)
- The Bank of Zork puzzle in Zork II was no longer new. Dave Lebling didn’t understand what it was. It seemed that nobody did. And yet, it went in unchanged (bad).
Good puzzles usually involve doing cool things. I put magic in Repeat the Ending because doing magic is fun (even in a less fun story). If I had to divide 782873978923 by 13098709187090897 by hand, that would be difficult, but it would not be fun. It certainly would not feel “cool.” A bad puzzle is work. It’s rote. A good puzzle tries to do something new, it challenges not only its player but its author. If the player doesn’t like it, they are at least compelled by it.
10 Likes
To me, a good puzzle (i.e. one that players will enjoy) has two components:
- Finding the solution is fun. This usually (though not necessarily) means an “aha!” moment when you figure out what you need to do. If someone tells you the solution, your reaction should (again, usually) be “ah! that makes sense!” (or “clever!” or “I should have thought of that” or…), not “how on earth was I ever supposed to figure that out?!”
- Implementing the solution is fun. Solving the Towers of Hanoi problem has provided many an “aha!” moment to novice computer programmers, so why isn’t it a good puzzle in IF? Because once you’ve figured out the solution, putting it into practice is mind-numbing tedium.
The infamous cat-hair-moustache puzzle mentioned above fails the first of these. The solution might be fun to play out—going around catching cats and such is good classic adventure-game material—but what you need to do makes no sense even when explained.
Separate EXAMINE, SEARCH, and LOOK UNDER verbs fail the second of these, in many cases. Maybe there’s only one object in the whole game you need to LOOK UNDER to find a vital clue—but unless it’s really well-clued, now you’ve taught the player they need to try this verb on literally everything they see. That just adds a lot of pointless busywork to every room.
The unlisted exits puzzle in Jigsaw (where, in the very first room in the game, you have to ignore the exits listed in the room description and go a different direction entirely to find vital equipment) fails on both these accounts, for me. When I found the answer in a walkthrough, my reaction was “…seriously? Southeast? Why?”. And then once I knew the solution, I needed to tediously try every direction from every location in case of further hidden exits, which wasn’t especially fun.
Fixing the first of these is relatively easy. Have people test it and gauge their reactions. Add more clues if you need to, add alternate solutions, and so on. There’s a whole lot that’s been written on this and I like to think I’ve gotten a good amount of practice at it.
Fixing the second of these is a skill I’m still developing. Testers tend to be great at determining if something is frustrating them, but they don’t always know why. Figuring out the why is the author’s job.
For a specific example, some people ran into frustration in Stormrider because they didn’t examine one of the various tools you acquire throughout the game. The easy solution is to just say “well that’s user error”. But this indicates I’ve been implicitly expecting the player to always EXAMINE everything they pick up. Someone suggested in another thread, why don’t I just make it automatic? Every successful TAKE followed by an automatic EXAMINE? Obvious in hindsight, yet not something I’d ever thought of before.
(Of course, if you make too much automatic, then you start losing the “aha!” moments. But I don’t think anyone gets “aha!” moments from realizing they need to examine things when they pick them up.)
So, are there any “just bad” puzzles? I would say yes—puzzles that too egregiously violate one of these guidelines. Breaking the first gets you moon logic puzzles. Breaking the second gets you randomized mazes that need to be solved by brute force. Breaking both of them at once is a recipe for player frustration.
9 Likes
(Since there have already been examples involving graphical games, I’ll do that too.)
There are the kind of games like Blue Ice where I suspect the design document could be neatly summed up with “'cause I got high, because I got high, because I got high”. But at least you realize pretty quickly what you’re in for, and you’re either on board with it or you’re not.
What I find really annoying is when a game gives me a problem that I could easily imagine in real life, but doesn’t allow any real life solutions.
In Phantasmagoria 2, you realize that your wallet has somehow gotten under the sofa. (The protagonist blames his pet rat.) In real life, I would find a stick or a shoehorn or something to try and push it. I might try to lift or move the sofa, if it’s not too heavy. But the game expects you to put your pet rat under the sofa, then lure it out from there with a granola bar.
In The Riddle of the Sphinx, you come across an empty camp. You suspect there may be something hidden in a pot of stew. It’s cold, but I guess if you suspect foul play you may still not want to put your hand in it. But surely there is a fork or a spoon or something you could use to fish around for it? But no, the game expects you to light the stove to boil away enough of the stew to reveal the object. Something that I would expect to take hours in real life, but here it happens in an instant.
I rage quit that game shortly afterwards, while in the middle of trying to find the combination for a lock. I had found some numbers on a note, so obviously that had to be it, right? No, those were references to Bible passages that contained numbers. So those had to be the combination, right? No, you need to put those numbers into clues gathered from cassette tapes that you found, and… You know what, game? I really don’t care any more!
But @zarf claimed it had good puzzles when he reviewed it so maybe I just got off to a bad start. 
5 Likes
Thinking about it, I came to a soft conclusion about puzzle design that I’m not even willing to call it a conclusion, given how proverbially bad I am at designing those. Although I think I take in much consideration the players’ feelings when coding.
An example: in Anchorhead, that nothing will move me from believing is the best IF ever, at least in the first version had you undressing, closing a door and then realizing you didn’t close the house’s door thus having to cross all the mansion again before going to bed on day one. That would have been successfully done with a line of text: GO TO BED → “You close all doors -just in case- and undress for a long night of sleep”. I’m certain it was intended not just as a reminder of reality but as an added layer of suspicion on behalf of the PC.
I hated it, and it was not even a puzzle.
Players have fun with the aforementioned “aha!” moment. Some authors have a twisted sense of fun when THEY have the “aha!” moment.
Due to my experience with Anchorhead (and with the feedback from my own players) I do a lot of work to avoid dressing-before-sleeping moments (consider this a metaphor). Others (especially in the retro community, I believe) seem just to want to replicate the tortures they endured on their players in a very juvenile fashion.
A friend of mine, in the Eighties, forced the players to type something like HAUL STONE, the parser not acknowledging any of the regular THROW verbs because of something very similar. On his behalf, we were fifteen.
Hiding objects is not designing a puzzle, at all, imo. It’s just a way to elongate the game a bit and/or make it more realistic based on the premises (I still stand by the difference between an open container and a pile of leaves—and hate when I have to take something from a container I hold before using it or having to look inside a glass to find it hides a coin). When it becomes a puzzle it’s because the authors wanted their “aha” moment in spite of the player.
I’m still digesting the steak.
Ps: mechanisms, when understood, should be automated in 2023 like sending a letter to a friend via email. Check Hadean Lands. Towers of Hanoi were never fun. The first one, probably, because it was new. But after that…
The same applies to mazes and all that was in Zork or Colossal Cave.
Burp.
6 Likes
By jove, I think you’ve got it! Now we need another thread for examples of good puzzles!
Edit: I will add that a great puzzle should also be thematic to the game. That it should naturally occur within the game context. That I should not pepper my game with chess puzzles unless it is within the game’s theme.
Edit: So, uh, does that mean my mate-in-26 (premove) chess puzzle a “grind”, and therefore violates the second component? 
5 Likes
While I agree that the books are really, really annoying to find, I don’t think any of them are necessary to beat the game. There are three that I know of:
Mein Kampf, which gives you a non-violent way to get rid of the guy guarding the security station so that you can disable the alarm, but even when I do that I always get captured on the way out anyway. And that leads to another annoying puzzle with Indy and Henry tied to chairs, having to repeat the same command dozens of times to move them to the other side of the room.
An airplane maual that gives you an option to completely bypass the game’s most frustrating maze. (I am, of course, talking about the Zeppelin.)
A book of maps that I just never found particularly useful.
There’s another pixel-hunt “puzzle” near the start of the game, but all that gives you is a fake Grail diary that you can use to completely skip the detour to Berlin. But doing so makes the next section of the game harder, so the game is essentially punishing you for playing it smarter than in the movie.
Not my favorite Lucasfilm/LucasArts adventure game by any stretch of imagination, but I can appreciate what they were trying to do. Giving you options beyond what happened in the movie is a good thing. It’s just that most of the actual puzzles aren’t that interesting to begin with.
It’s a game that has been a lot on my mind lately though, because I’ve been working on getting ScummVM to finally emulate the Macintosh version’s distinctive GUI properly. I’ve heard it explained that it was made at a time when Mac users were an important target audience, and it was also important that the Mac version didn’t look too much like a DOS game.
(If you know what to look for, you can see the bits in this screenshot where ScummVM still doesn’t quite match a real Macintosh. There’s room for anyone to make improvements. It’s already not a solo effort, and that that could be your name in the commit log! 
)
5 Likes
I don’t remember any of them any more, so I don’t know!
4 Likes
That is probably the case but the player doesn’t know this— the gaming norms of the time have taught them if they miss an item they may make the game unwinnable. I liked the alternative solutions concept they had in mind, it was just frustratingly implemented.
This reminds me of a puzzle from a contemporary point-and-click adventure game I was playing last month. The player has a fish bowl full of water, and a broken diving suit where it’s obvious that the solution is to get rid of the water and use the fish bowl with the suit. But the character won’t do anything sensible like this and the only hint is that the water needs to go. The actual solution is to put a piece of hot coal into the glass bowl to instantly boil off the water, and in doing so, sterilise it, and simultaneously give you an object for throwing at something else (despite there being other inventory items that appear throwable). The solution is abject nonsense that only starts to make “sense” if you started with wanting to cool down the coal and incidentally got a sterilised water bowl in the process.
3 Likes
Well, it all depends on your target audience, right? If I came across this as the final puzzle of Dr Ludwig in IFComp I’d find it both incredibly difficult and incredibly tedious and would hate it. But I imagine the audience for chess puzzles specifically, doesn’t find solving them tedious. They specifically seek them out and know what they’re getting into.
Similarly I have fun solving the New York Times crossword but I wouldn’t like it in the middle of an IF piece.
9 Likes
Prince Quisborne has both a many-move chess puzzle and a crossword in the middle of the game. Pretty funny these two examples came up together!
7 Likes
To be fair, it’s not a chess puzzle (as one understands the term) nor is it a crossword puzzle in the Times sense… 
7 Likes
Yes, on my replay I made the mistake of thinking it was a chess puzzle and get pretty stuck! And I agree on the crossword.
The chess puzzle was actually my favorite part of the replay, so I guess it just goes to show that any puzzle you can think of as ‘good’ or ‘bad’ can be the opposite.
4 Likes
Ghost has an optional ending that’s unsolvable to anyone but the person who wrote it. The only hint is that you have to type in a certain word that the author likes the sound of, and just that word as a command. What could it possibly be? Well, I didn’t feel like inputting the entirety of the English language piece by piece, so I just text dumped the game to find it. (I invite you to come up with a few guesses and then verify your answer: the word is TANGO.)
My Girlfriend’s An Evil Bitch also had a mechanic I really didn’t like. The opening of the game has you solving a bunch of puzzles that aren’t super hard but are more just tedious (chasing stuff around your house, fixing your car, looking at post-its to find computer passwords and then typing those in) and then going to a general store for supplies. Some of these are survival books that give you hints for later on, and others are things that you need in the woods to help you get rescued. You don’t have enough money to buy everything. If you buy the books, you’ll be stuck with instructions for stuff you don’t have; but the rest of the game is so frustratingly underclued that if you just buy the tools instead, you won’t know how or when to use them. Without savescumming or undoing, the only thing you can do is buy the books, memorize or write down the information, then replay the entire game up to that point and buy the tools instead.
4 Likes
That has to be a Spider and Web reference…
6 Likes
You reminded me of how it’s probably best to focus on good and bad practices… and not necessarily on good or bad puzzles themselves. Any rule can be broken, if it’s deliberately done so with consideration of how it breaks the rule.
This is not to say that puzzles cannot be objectively bad, but we definitely should give the author the benefit of the doubt that they weren’t being completely ignorant and try to identify why it failed… not so much that it was a failure.
For example, the “dead horse” link I shared with the Gabriel Knight 3 moustache puzzle is interesting because the author of the article states:
In order to construct the costume, Gabriel Knight must manufacture a fake moustache. Utilizing the style of logic adventure game creators share with morons, Knight must do this even though Moseley does not have a moustache.
I’m thinking, because it’s a photo ID (and you look nothing like the man in the picture), it’s actually advantageous to distract from the facial features with a moustache. It’s quite clever, in that respect. However, was this idea conveyed in any useful manner to the player? My guess is not (though, I have not played GK3). I would have allowed the player to pretend to be Mosley without the moustache and then have the police called… then in the game over screen, hint that maybe there was something else to do to either your disguise or the photo ID somehow.
Anyway, food for thought.
4 Likes
Ideally the “rules” are there for a reason - usually it’s “over time, people have collectively decided this is good practice and this is bad practice” and the aphorism is generally “you must learn the rules before you break them” but it might be a little better to express “it’s best to know why a rule was established before you break it.”
Consider Hard Puzzle which is a game which purposefully exploits bad parser puzzle design principles to complicate what should be a simple task. This game probably could be subtitled Here’s What Not to Do and becomes illustrative of this in a humorous way. Bad design is the gimmick.
These separate quasi-examine actions were common in early text adventures when part of the goal was to make the game last a while to justify a commercial purchase. Authors nowadays tend to disregard these because nobody wants to examine everything three times to make sure they haven’t missed anything, however these may be extremely appropriate in a detective adventure where evidence might be concealed and not obvious from a cursory search. As said, it needs to be well-clued, and non-standard or not obvious actions hopefully tutorialized. Most people won’t object to a fourth-wall break message. If you have an encyclopedia object, the description can say
It’s a huge leather-bound book with all kinds of information in it. (You can SEARCH BOOK FOR [TOPIC] to look things up.)
(Regarding Anchorhead)
Yes. That game straddles the classic era, so it has all kind of “gotcha!” moments that would make them unwinnable. Most players are used to leaving doors open - either because it helps with mapping and progress, or because the importance of a closed door may not be evident. RIVEN actually exploited that - one route the player needed to proceed was concealed by the open door they just traversed. Nowhere in the game is the importance of closing doors ever addressed. Sure, it’s thorough adventuring to check everywhere and move everything from all directions, but it’s a cheap puzzle that makes you think the author is laughing at you for not realizing when it’s never been a consequence prior to that.
In Anchorhead it makes sense that not closing and locking your front door is logically a bad idea, and all it would need is atmosphere messages like A breeze from the open front door chills you. At least that’s not the worst thing that could invade your house. to remind players to lock it. Or, make locking the door from the inside implicit when closing the door - as it is for many people in real-life habitually where they don’t think about it. Sometimes authors do want that element of routine where forgetting to do something causes consequences and that can be effective in a horror game - you forget to lock the door once and you wake up to find a sinister message written so you curse yourself for forgetting - but the author needs to set the situation and parameters initially to make that work. Knowing you forgot to lock the door and suffering due to it is good plot; a player having their house ransacked but unaware the reason was because they forgot to lock the door is bad design.
7 Likes
See now, I actually thought this was a stroke of genius. Michael Gentry is a master of atmosphere, and some of that comes in the form of passive messages (e.g., about closing your umbrella when you step inside), but there’s an added layer of immersion that comes from making the player physically do something just because their character would do it, not because it’s strictly necessary to advance the game.
Too much of this is annoying (which is probably why the aforementioned umbrella behavior is automatic rather than manual), but I found the “physically going downstairs to close the door” task enormously effective at immersing me into the PC’s mindset. It’s such a vivid (and slightly creepy) real-world task that you would only do if you were really thinking like the PC, not if you were thinking like an efficient player trying to get through all the puzzles.
9 Likes
This could be very effective - if the PC gets ready for bed and right before they close their eyes just have the subconscious nag, “Did you remember to close the front door?” so it’s like an eye-rolling task. Remind them a couple times for free, then the third time they get that message when they wake up: “Did you remember to close the front door?”
If it’s a puzzle the author participates and reassures the player what they are doing has been thought of. If the player is confused or feel they’ve been cheated by the game or not given any warning prior to failure it’s a trap not a puzzle.
7 Likes
To me, the Anchorhead one counts as “fun to execute the solution”. Maybe “fun” isn’t quite the right word here—but it’s atmospheric and engaging.
3 Likes