[IF Comp 2013] Threediopolis
Threediopolis, Andrew Schultz, Z-code
version 2013/10/10
By this point Andrew Schultz has pretty much staked out his territory: he makes what I'd call Stereotypical Programmer IF - concerned with maths, logic, word puzzles, tech jobs and baseball nerdery, distinctly less concerned with writer stuff. My tastes are towards the other end of the spectrum: the writer stuff is really important to me, and I don't get a whole lot out of puzzles unless they're integrated into a robust story. So my assumption by this point is that, while he seems like a dedicated guy and likely to continue growing as an author, and there's definitely an audience for the things that he enjoys making, he's probably not going to produce something that I straight-up love.
And oh boy, the initial impressions of the game confirm this in spades: it's going to be a big, unabashed puzzlebox with the thinnest of narrative justifications. The titular city is basically a set of Cartesian coordinates. You're an AFGNCAAP who has been given an enigmatic task by some guy; the task consists of a list of coordinates, each with a crossword-clue-like word or phrase beside it.
My initial guess is that the clues need to be spelled out by your movement, making words like DUDE with down-up-down-east or whatnot, and probably ending at the location on the final letter. I'm not clear whether the additional one-letter commands count, but it seems like they must do, or it'd be horribly misleading to have put so many of them in. So that leaves me with the letters BCDEFGHIKLNOPQSTUVWXZ. Or it might be some subset of that - perhaps everything but the out-of-world commands, which would be... maybe DEIKLNPSUWXZ, but perhaps not exactly that. Or it might just be directions, in which case I've got DENSUW, and it also mentions IN and OUT. P can't be part of it, though, because that returns you to the centre regardless. And presumably once you have the set of letters, you go into cryptic-crossword mode and try to figure out a word you can make that fits the clues?
"Ed Dunn's Needs" only uses directions, so let's assume that it's that. (I'm not awfully secure about this, because of all the extra one-letter verbs, but I can't see any other way of filtering.) I'm not entirely confident that cryptic-crossword is the right approach here, because clues seem to repeat themselves a good deal, but let's give it a try. I'm not sure about the status of IN or OUT, either, so let's ignore them for now.
Except that this response suggests yet another list:
Being a messenger, you don't have much more to do than go in the six basic directions (transport tubes make going up and down easy.) Or enter places. Maybe knock. Nothing fancy. Type C to see everything.
...which would be CEDKNSUW and maybe IN and OUT again. Nothing is very clear; I think I can kind of see the vague outline of how the puzzle works, but there are too many unknowns for me to work with here, and none of the crossword clues are so clear that I know what my target is.
Ach. Fine. Looking at the hint sheet. It's written in a cryptic notes-to-myself style, but it seems to suggest that my intuitions were more or less right - the letter set is DENSUW, and the crossword-clue looking things are just that. Other than that, it doesn't explain what the trick I was missing is. It seems as though there was something I was meant to notice while walking to the listed coordinates, but there's nothing that really jumps out above the noise of randomised flavour text.
At this point I think I'm done - no, wait. Maybe we just try entering words from the centre, since the game keeps taking us back there. And that's the bit that I was missing, really: each path has to begin in the centre and end at the listed coordinates, stepping through the word you need to spell out. You can get a bunch of the points just from trying all the DENSUW words you can think of; even the ones that don't correspond to clues sometimes have special messages. Once you've exhausted this scattershot approach, you're probably still left with a lot of unsolved clues. For these, you have the end-point of each location, and a good idea about the number of commands to get to it, and this is presumably the real meat of the game; but by that point I was running out of time and patience. In retrospect, it really didn't help that the entries in the first row of clues were pretty much all some variation of 'pal', which led me to look for synonyms of 'friend' rather than names of people.
So, hm. This is a nice enough pure-puzzle idea, but (for me, at least) working out how the central mechanic worked took long enough that I was pretty close to being done with it by the time I was able to start solving things. Or maybe I was unlucky that I didn't randomly wander into any of the clues early on - possibly because I was exploring in straight lines. But the time I took on getting this first step meant that it never really got to be fun. For a game that doesn't aspire to anything more than being a fun puzzlebox, No Fun means a maximum score of 5. But 5's my usual score for "totally not my thing, but does an excellent job of what it sets out to do," and I don't think Threediopolis perfectly attains that.
A lack of clarity about how a puzzle mechanic is intended to work has plagued a lot of Schultz's games. Some proportion of it, I'm sure, is because writing for clarity just isn't his forte, but some of it is because of a really deep desire to make another puzzle out of working out what the puzzle's parameters are. That's not a bad ambition - the 'aha, that's how it all works' moment he's after is one of the rarer and more lovely pleasures of IF - but he's a lot better at designing the first-order puzzles than he is at offering clues for the second-order ones.
Specifically, it seems as though the game expects you to solve one or two of the clues by accident before you can figure it out: this never happened for me. But once you get it, you can solve a lot of the words just by spamming words made with the letters available; of the remaining words, some are distinctly unfair (like the name Wendee, which isn't the common spelling, or indeed a spelling that I've ever encountered.) I am no expert on either setting or solving cryptic crosswords, so I can't say a whole lot more than that about how it measures up.
Score: 4 or 5
version 2013/10/10
By this point Andrew Schultz has pretty much staked out his territory: he makes what I'd call Stereotypical Programmer IF - concerned with maths, logic, word puzzles, tech jobs and baseball nerdery, distinctly less concerned with writer stuff. My tastes are towards the other end of the spectrum: the writer stuff is really important to me, and I don't get a whole lot out of puzzles unless they're integrated into a robust story. So my assumption by this point is that, while he seems like a dedicated guy and likely to continue growing as an author, and there's definitely an audience for the things that he enjoys making, he's probably not going to produce something that I straight-up love.
And oh boy, the initial impressions of the game confirm this in spades: it's going to be a big, unabashed puzzlebox with the thinnest of narrative justifications. The titular city is basically a set of Cartesian coordinates. You're an AFGNCAAP who has been given an enigmatic task by some guy; the task consists of a list of coordinates, each with a crossword-clue-like word or phrase beside it.
My initial guess is that the clues need to be spelled out by your movement, making words like DUDE with down-up-down-east or whatnot, and probably ending at the location on the final letter. I'm not clear whether the additional one-letter commands count, but it seems like they must do, or it'd be horribly misleading to have put so many of them in. So that leaves me with the letters BCDEFGHIKLNOPQSTUVWXZ. Or it might be some subset of that - perhaps everything but the out-of-world commands, which would be... maybe DEIKLNPSUWXZ, but perhaps not exactly that. Or it might just be directions, in which case I've got DENSUW, and it also mentions IN and OUT. P can't be part of it, though, because that returns you to the centre regardless. And presumably once you have the set of letters, you go into cryptic-crossword mode and try to figure out a word you can make that fits the clues?
"Ed Dunn's Needs" only uses directions, so let's assume that it's that. (I'm not awfully secure about this, because of all the extra one-letter verbs, but I can't see any other way of filtering.) I'm not entirely confident that cryptic-crossword is the right approach here, because clues seem to repeat themselves a good deal, but let's give it a try. I'm not sure about the status of IN or OUT, either, so let's ignore them for now.
Except that this response suggests yet another list:
Being a messenger, you don't have much more to do than go in the six basic directions (transport tubes make going up and down easy.) Or enter places. Maybe knock. Nothing fancy. Type C to see everything.
...which would be CEDKNSUW and maybe IN and OUT again. Nothing is very clear; I think I can kind of see the vague outline of how the puzzle works, but there are too many unknowns for me to work with here, and none of the crossword clues are so clear that I know what my target is.
Ach. Fine. Looking at the hint sheet. It's written in a cryptic notes-to-myself style, but it seems to suggest that my intuitions were more or less right - the letter set is DENSUW, and the crossword-clue looking things are just that. Other than that, it doesn't explain what the trick I was missing is. It seems as though there was something I was meant to notice while walking to the listed coordinates, but there's nothing that really jumps out above the noise of randomised flavour text.
At this point I think I'm done - no, wait. Maybe we just try entering words from the centre, since the game keeps taking us back there. And that's the bit that I was missing, really: each path has to begin in the centre and end at the listed coordinates, stepping through the word you need to spell out. You can get a bunch of the points just from trying all the DENSUW words you can think of; even the ones that don't correspond to clues sometimes have special messages. Once you've exhausted this scattershot approach, you're probably still left with a lot of unsolved clues. For these, you have the end-point of each location, and a good idea about the number of commands to get to it, and this is presumably the real meat of the game; but by that point I was running out of time and patience. In retrospect, it really didn't help that the entries in the first row of clues were pretty much all some variation of 'pal', which led me to look for synonyms of 'friend' rather than names of people.
So, hm. This is a nice enough pure-puzzle idea, but (for me, at least) working out how the central mechanic worked took long enough that I was pretty close to being done with it by the time I was able to start solving things. Or maybe I was unlucky that I didn't randomly wander into any of the clues early on - possibly because I was exploring in straight lines. But the time I took on getting this first step meant that it never really got to be fun. For a game that doesn't aspire to anything more than being a fun puzzlebox, No Fun means a maximum score of 5. But 5's my usual score for "totally not my thing, but does an excellent job of what it sets out to do," and I don't think Threediopolis perfectly attains that.
A lack of clarity about how a puzzle mechanic is intended to work has plagued a lot of Schultz's games. Some proportion of it, I'm sure, is because writing for clarity just isn't his forte, but some of it is because of a really deep desire to make another puzzle out of working out what the puzzle's parameters are. That's not a bad ambition - the 'aha, that's how it all works' moment he's after is one of the rarer and more lovely pleasures of IF - but he's a lot better at designing the first-order puzzles than he is at offering clues for the second-order ones.
Specifically, it seems as though the game expects you to solve one or two of the clues by accident before you can figure it out: this never happened for me. But once you get it, you can solve a lot of the words just by spamming words made with the letters available; of the remaining words, some are distinctly unfair (like the name Wendee, which isn't the common spelling, or indeed a spelling that I've ever encountered.) I am no expert on either setting or solving cryptic crosswords, so I can't say a whole lot more than that about how it measures up.
Score: 4 or 5