The Keyring Problem

[projectyl]

This isn't my recap of the 2011 Hunt. I don't know that I'll have the energy to do a really good full-length one like I did for 2009. For what it's worth, it was definitely an A++ Would Do This Team's Hunts Again sort of Hunt, even more so than S.P.I.E.S. was.

My one major gripe with this Hunt was with two of Zelda's three metas. I may write a post

Comments

[projectyl]

I should mention that I'm as guilty of this as it gets, especially in some of my earlier stuff. AcidFast from Sekkrets comes to mind a pretty nasty keyring puzzle that I don't think anyone ever forward-solved.

[canadianpuzzler]

I have many things to say about puzzle hunts in general and a number of things to say about this hunt in particular, and I hope to actually post a wrap-up entry soon. But one thing that I've said several times already is this

The Taipei Keyring problem

[thatwesguy]

First: I agree with what you're saying

Re: The Taipei Keyring problem

[tigupine]

I can only comment on Advanced Maths. I had the same concern about that IEEE paper when I was proofreading / checking for correctness, but then I realized that base -1+i works the same way as any other base.

I wouldn't have been able to recognize or even research most of those encodings, but as an author more accustomed to writing for linear events in which the puzzles must be solvable by every team, I know I sometimes forget that crazy-hard science specialty puzzles are admissible in the Hunt as long as there are enough specialists out there to solve them. I think the puzzle reached its intended audience, because I know that a (small) number of teams forward-solved it.

Re: The Taipei Keyring problem

[thatwesguy]

This is good to know! Thanks for the reply.

I think my issue with Advanced Maths really comes down to the idea that by the time I had confronted the puzzle, I had already lost some faith. Therefore I lacked the resolve to follow up the math appropriately.

I hope that that makes sense; what I'm trying to say is that I had a bit of fear that if I spent a long time trying to solve it mathematically, I'd be punished by a solution that would feel arbitrary. Knowing in retrospect that the math was actual math (with the possible exception of "negafibonacci numbers," which feel entirely arbitrary to me, lacking as I do a framework in which that idea seems consistent and intuitable) makes me wish I'd worked at it more straightforwardly instead of giving in to the nagging fear that in the end it was going to be based on 70's movies titles or something.

My bad!

Re: The Taipei Keyring problem

[tigupine]

I totally know what you mean about losing faith. I've had that happen during puzzle events before.

I've resolved to learn more at some point about how negafibonacci numbers are actually used in CS, but for now I agree that they're a total pain in the ass. It didn't help that at the time I proofread the puzzle, at least, the Wikipedia page about them gave an incorrect algorithm for how to generate them (and the way I ended up dealing with them in my program was so janky that it didn't belong on Wikipedia). I spent a bit of time wondering whether this should be considered our problem. Given more time, I would have tried to fix it.

[jangler_npl]

To stretch the metaphor even further, I feel that "Rocky Horror" had a variant of the keyring problem where your given a ring with a bunch of little keys and one big one labeled "master key", but the master key doesn't work. I mean, if you're reading out binary digits in an order specified by a previous instruction, there is absolutely no reason to make the representation anything other than big-endian. It'd be like having a diagonal spell LEVER and the answer be REVEL without the need for reversal being clued at all.

That said, there were very few puzzles that I saw the solution to and said "how was I supposed to think of that?" rather than "I should have thought of that.".

[dr_whom]

...I think I'm not sure what you mean here about Rocky Horror? As far as I can tell the binary digits read off in conventionally big-endian order: the first one is 1101, which is thirteen (not eleven), and extracts the 13th letter of the corresponding caption, as described; the second one is 1110 and is fourteen, and so on. Maybe I'm misunderstanding your point?

[jangler_npl]

Um...hm...er...never mind...serves me right for not double-checking with the answer page. What happened was I spent a while looking at the collected data (the climbing moves, the binary codes, etc.) and not getting anywhere, and then a few hours later overhearing that the puzzle was solved remotely by someone who just switched the endian-ness of the binary digits. Clearly I misheard, or had an extremely vivid hallucination, or something. And to think I was doing relatively well on not embarrassing myself this year...

hm

[jangler_npl]

If you look at how many teams forward-solved vs. back-solved any puzzle, you should be able to note which puzzles were either too random with their logic leap or too grind-y "keyring" and un-fun.

On a general note (not specific to Rocky Horror), I dislike random extractions... when my energy has been spent fully conquering a puzzle, and then I have to randomly choose to convert to binary or morse by arbitrarily picking this thing as -/1 and that thing as ./0, that's what I find the least fun.

[noahspuzzlelj]

Er, what was the problem with Zelda broadly? There was certainly no pattern that we were *consciously* breaking. Mario the meta answers were things that matched their round (a mushroom, a flower, a star). In Megaman, as we told you, all meta answers were weapons. Those are two totally different patterns. The pattern for Zelda metas is that they're all 9 letters long and spelled out letter-by-letter. The pattern for Civ is that they fit into the pure supermeta (which isn't really a pattern at all). The pattern for KD is that they're all objects of the appropriate scale (provided that you interpret Oberon as the moon not the character

[projectyl]

The intro skit (which may not have been given to blindsolvers) told us that "Mario will need certain items to break into the location where Peach is being held. The puzzles are the key to finding those items." Which, looking back, could be interpreted as applying only to World 1, but I definitely spent the Hunt under the impression that every normal meta was supposed to give a game-appropriate object and every metameta was supposed to give a star piece location, which caused me to discard both CREATURES and POLLINATE as possibilities.

[noahspuzzlelj]

Ah, ok, so it was a miscommunication about metas vs. metametas. I think you're referring to the intro which says "If you solve enough puzzles in a world, their answers considered together should lead you to one of the objects that Mario needs. If you find the objects from all the worlds, Mario should be able to collect them and use them to rescue the Princess."

This was intended to refer to the *world metas* (whose answers were AIRSHIP, WILY COYOTE, NERF SWORD, ORANGE, and OBERON which are indeed all objects). I can see how we the miscommunication happened here, but the problem was us not being clear in the handout, not with breaking a pattern.

[noahspuzzlelj]

There was various issues with the "World 1" to "Mario World" transition. I wanted the original intro to be clearer about the "world" "level" distinction (that is it's World 1, which has 4 levels in it), but I see that the actual final version of the handout was misleading on that point. Which is to say, I agree with you that we messed up on the handout.