First things first: congratulations to Hyunmin Fan Club for being the first team to complete the puzzle hunt by defeating the Spectral Thieves, in just under 24 hours, before hints were released on Sunday! Additional congratulations to Reborn's Most Trusted Citizens for being the first pure Smogon team to complete the hunt; they'll receive banners on the forums commemorating this feat, as soon as I get around to arranging them. We were also relieved to see P Squad slip into the finishers' list with their final time of 9d 12h 48min 9s, just under the deadline at 9d 13h.
For those of you who missed the hunt and would like to work on the puzzles, the full set of puzzles and solutions can be found at https://spo.ink/sph18archive.
The rest of this post is primarily about the hunt, the process behind it, and how we thought things turned out in the end, from a staff perspective. Warning: spoilers incoming!
FORMAT AND PUZZLES
Preparations for the whole hunt started back in February, with a simple goal: to introduce Smogon to puzzling. Prior to this, I'd posted a couple small puzzles to subcommunities, and given the generally positive feedback from those, I thought Smogon as a whole might just be about ready for a small puzzle hunt of its own. With this goal in mind, we had a couple directions to generally aim for with the hunt:
- Several Pokemon- or Smogon-related puzzles, to allow solvers to work with information they were relatively comfortable with
- Easy early puzzles, to provide positive feedback to solvers and afford them a sense of success
- Give a good first impression of puzzles with high-standard content: no random anagramming, no random Caesars, no mind-reading steps, and no boring puzzles!
With the idea of badges set in stone, the next logical step was to figure out a reason solvers would want to collect all the badges, and what better way than to combat a Big Bad Evil Guy? So we settled on a faceless villainous character, and went ahead from there. Capitalizing on the theme of "get a lot of badges to defeat a bad guy", the meta answer fell into place soon after. On the Smogon forums, badges a user's earned through their contributions are displayed under their postbit on the forums; when a user gets too many badges, they no longer fit, and turn into a marquee, known as a "scroller". As you can imagine, THE ELDER SCROLLER popped up pretty soon afterwards, which was appropriate not only thematically but also because its length exactly matches the number of distinct badges on Smogon.
Eventually, we realised that having teams showing up and presenting them with 16 puzzles wasn't a great idea, even if said puzzles were staggered. In the light of this revelation, we decided to add a Round 1 to the hunt, which worked well with the flavor in creating some sort of context to the attack of the (as of yet still faceless) Evil People. We hoped that this additional round would enhance teams' experience and add to their sense of accomplishment, even if they didn't manage to complete the entire hunt. Ironically, though, a poorly-designed meta took this round in the entirely opposite direction (more on this later).
After hammering down our structure, though, we were stuck. Our present Round 2 metapuzzle involved looking at the TM that was given out with each badge, and indexing into the answer by that number... except this involved taking the modulo of the number when it was too big. Horribly unintuitive, bland, uninspired, and all-around unimpressive, we knew we had to come up with something better, but it took us until April (as well as the addition of some staff and the kicking out of inactives) to come up with the idea that eventually became Round 2. A few days of finalizing answers later, we were stuck with a set of answers that included such gems as BOULTER and GLACIS RAMPART (more on this later), but in the words of Macbeth, "I am in [bad puzzle design] / Stepped in so far that, should I [construct] no more, / Returning were as tedious as go o'er".
And so, with the metapuzzle finalized, we altered some of the puzzles we'd already constructed to fit new meta constraints, and from there had about two months to design the rest of the 24 puzzles. Especially in the context of a task which requires creative input, this isn't the best timeline to be faced with, but we managed to finalise and polish most of the puzzles in time, with 46/51 feedback responses agreeing that the puzzles were well-designed. Though this feedback came from a set of puzzle newcomers with arguably lower standards, I think it's notable that we appealed successfully to our key audience. (Just 39/51 agreed that they were fair, though, indicating an area with significant room for improvement.) Finally, 49 out of 51 feedback responses indicating an interest in participating in a similar event again, with the last two responding neutrally to the relevant question—recall that, as mentioned earlier, our primary goal was to introduce Smogon users to the world of puzzles and to give them a great first impression, though fun and hopefully memorable puzzles; in this respect, I'd call this event a success overall.
On a finer scale, however, I'd like to take this opportunity to go into some of the puzzles. If you've yet to work on the puzzles and intend to do so, this would be a good section to skip, since there's even more puzzle-specific spoilers in here.
Double or Nothing: starting on the right/wrong foot?
I got a significant amount of feedback that this puzzle was too difficult to be the very first puzzle in the hunt, especially when 1-2 and 1-3 were way, way easier. The reason I placed this in the hunt right at the start is because I knew it was definitely easier than 1-4 through 1-6, and looked way less intimidating; furthermore, it contained a word search, which solvers were likely to be familiar with, so I hoped it would give solvers something to immediately get into (instead of figuring out what a list of words or pictures was supposed to mean, for example).
Another reason I didn't place 1-2 first was because I didn't want the first puzzle in the hunt to be an anagram-centric puzzle, especially after I spent some effort telling teams that random anagrams wouldn't show up in puzzles. I also didn't want to place 1-3 first because I didn't want the hunt to start with a Pokemon puzzle, which may have colored solvers' expectations of the quantity of Pokemon-related puzzles in an inaccurate manner. Judging by the number of teams which tried random anagramming and went down Pokemon-related paths at every opportunity available, this may have been for naught.
I got a significant amount of feedback that this puzzle was too difficult to be the very first puzzle in the hunt, especially when 1-2 and 1-3 were way, way easier. The reason I placed this in the hunt right at the start is because I knew it was definitely easier than 1-4 through 1-6, and looked way less intimidating; furthermore, it contained a word search, which solvers were likely to be familiar with, so I hoped it would give solvers something to immediately get into (instead of figuring out what a list of words or pictures was supposed to mean, for example).
Another reason I didn't place 1-2 first was because I didn't want the first puzzle in the hunt to be an anagram-centric puzzle, especially after I spent some effort telling teams that random anagrams wouldn't show up in puzzles. I also didn't want to place 1-3 first because I didn't want the hunt to start with a Pokemon puzzle, which may have colored solvers' expectations of the quantity of Pokemon-related puzzles in an inaccurate manner. Judging by the number of teams which tried random anagramming and went down Pokemon-related paths at every opportunity available, this may have been for naught.
A Brief Oration: a study in poor meta design
I placed this first because it was the single most commented-on part of the hunt. There's a few things that went wrong with this puzzle, so I'll try to break it down in point form:
I placed this first because it was the single most commented-on part of the hunt. There's a few things that went wrong with this puzzle, so I'll try to break it down in point form:
- The big one: we somehow missed out on the fact that we had SPE, EV, and HP in the same abbreviation-based puzzle... this led a lot of teams down a wild goose chase for Pokemon-related abbreviations.
- We underestimated the intuitive-ness of "common abbreviations" like SPE for SPEED and CTR for CENTER; SPD, for example, makes the grid look like it should be read downwards rather than across due to the triple consonants.
- ALTERNATE UNIVERSE had its own acronym, so we avoided that pairing and ultimately left it on its own; similarly, ELECTRONVOLT is hard to pair with another word, so we left both as single words. This caused the unfair logic leap of "use the first word, unless there's only one word"—when giving out hints, we basically gave this step out for free, since it was quite unfair to expect teams to figure it out.
- Upon providing a right answer, the bot was coded to return the input answer and a message that it was correct, instead of providing an "official" parsing of the answer. For this reason, we couldn't sufficiently show that ELECTRONVOLT was intended to be one word, or that SPEED READING was intended to be two. For this reason, we saw a number of teams attempting to fit in (e.g.) WPM for the latter. We also saw at least one "e-" for the former (ELECTRON).
- The answer phrase itself, while a fitting pun, wasn't extremely obvious, and it was hard to fill in without all 6 of the puzzles.
CAP Contributor: a different kind of bad design
Full disclosure: the original form of CAP Contributor was far simpler. Each line consisted of a line like "Weaken to behead (10)", indicating DEBILITATE to DECAPITATE. This, of course, made it necessary to realise the CAP connection in order to solve the puzzle, but we thought that it looked a little boring.
The night before the hunt, we decided to spruce it up a bit: the original idea was to depict the first word with a cap on it, and then either a description or a picture of the second word. Unfortunately, it's difficult to get pictures that indicate DEBILITATE or HANDIEST, so we ended up with using words for the left-hand side instead, hoping to force the realisation of the connection via ambiguous clues: unfortunately, only some of the clues were sufficiently ambiguous, meaning that teams could completely solve the puzzle without ever realizing what CAP meant in the puzzle; additionally, this only caused frustration in some teams over the seemingly random and insoluble ambiguities in the puzzle. Moral of the story: review puzzles earlier than the last night!
Full disclosure: the original form of CAP Contributor was far simpler. Each line consisted of a line like "Weaken to behead (10)", indicating DEBILITATE to DECAPITATE. This, of course, made it necessary to realise the CAP connection in order to solve the puzzle, but we thought that it looked a little boring.
The night before the hunt, we decided to spruce it up a bit: the original idea was to depict the first word with a cap on it, and then either a description or a picture of the second word. Unfortunately, it's difficult to get pictures that indicate DEBILITATE or HANDIEST, so we ended up with using words for the left-hand side instead, hoping to force the realisation of the connection via ambiguous clues: unfortunately, only some of the clues were sufficiently ambiguous, meaning that teams could completely solve the puzzle without ever realizing what CAP meant in the puzzle; additionally, this only caused frustration in some teams over the seemingly random and insoluble ambiguities in the puzzle. Moral of the story: review puzzles earlier than the last night!
Spacetime: the other side of the Dunning-Kruger effect
Spacetime was intended to be a challenging puzzle, but it ended up being way more difficult than intended. In the solutoin page, I mentioned a number of distinguishing features which I hoped would be fairly apparent to anyone who had played the games before, but it turns out that playing Pokemon for 13 years tends to make one more familiar with the games than they might immediately realize. I came across teams which spent hours scanning overworld maps in the hope of vaguely matching shapes, and even then had huge trouble trying to find the indoor locations (Power Plant, Meteor Falls, Team Rocket's Castle). I hope teams didn't get *too* frustrated with the difficulty of the map-finding!
Spacetime was intended to be a challenging puzzle, but it ended up being way more difficult than intended. In the solutoin page, I mentioned a number of distinguishing features which I hoped would be fairly apparent to anyone who had played the games before, but it turns out that playing Pokemon for 13 years tends to make one more familiar with the games than they might immediately realize. I came across teams which spent hours scanning overworld maps in the hope of vaguely matching shapes, and even then had huge trouble trying to find the indoor locations (Power Plant, Meteor Falls, Team Rocket's Castle). I hope teams didn't get *too* frustrated with the difficulty of the map-finding!
Up Against The Wall: a necessary evil
This was an interesting puzzle. It occurred because we had the final answer GLACIS RAMPART, which as it turns out is one of the only reasonable ways to insert the letter S into GLACIER for the meta extraction. We knew that we couldn't possibly encode the answer in the puzzle itself, since solvers would likely self-autocorrect it to GLACIER[...], thus making the metapuzzle extraction even more obvious than it already was. The only way they could get the answer without becoming suspicious of it, we figured, was if we gave it out as an awarded answer after completing a thematic puzzle. But what puzzle would be thematic about walls?... and so, we wound up making teams win ladder games with walls. We were actually ready to accept a wide range of answers: 6 Stakataka in AG, 6 Pokemon named after walls, or really any other way teams could justify their replays.
This step was really polarizing: while some teams found it hilarious, the teams which didn't come from a Pokemon background argued that it was very unfair, and even some who did just picked out a pre-existing replay instead of fighting a new one, somewhat diminishing the point of the step. The puzzle as a whole, in fact, garnered some mixed reviews, with some teams unappreciative of the seemingly rather random first step. Still, the puzzle managed to conceal the true intent behind GLACIS RAMPART, so in that sense its job was carried out successfully.
This was an interesting puzzle. It occurred because we had the final answer GLACIS RAMPART, which as it turns out is one of the only reasonable ways to insert the letter S into GLACIER for the meta extraction. We knew that we couldn't possibly encode the answer in the puzzle itself, since solvers would likely self-autocorrect it to GLACIER[...], thus making the metapuzzle extraction even more obvious than it already was. The only way they could get the answer without becoming suspicious of it, we figured, was if we gave it out as an awarded answer after completing a thematic puzzle. But what puzzle would be thematic about walls?... and so, we wound up making teams win ladder games with walls. We were actually ready to accept a wide range of answers: 6 Stakataka in AG, 6 Pokemon named after walls, or really any other way teams could justify their replays.
This step was really polarizing: while some teams found it hilarious, the teams which didn't come from a Pokemon background argued that it was very unfair, and even some who did just picked out a pre-existing replay instead of fighting a new one, somewhat diminishing the point of the step. The puzzle as a whole, in fact, garnered some mixed reviews, with some teams unappreciative of the seemingly rather random first step. Still, the puzzle managed to conceal the true intent behind GLACIS RAMPART, so in that sense its job was carried out successfully.
Mint Condition: what happens when you work together
2-14 was, to us, an absolute banger of a puzzle, and we really believed it was one of the better puzzles in the hunt, if a little on the easy side. As I mentioned in the puzzle's Author's Notes, it had been written in 6 hours the night before the hunt was due to start—the key difference, though, as pointed out by lovemathboy, is that it was written by the two of us: lmb threw out various ideas and provided the picture assets for two puzzles, while I handled the word-based sections, and stayed up to compile the entire document. By being able to bounce ideas off each other live, we were able to quickly eliminate poor ideas before they made it to the document and stayed there through inertia.
If we do do this again in coming years, we're hoping to be able to assemble a larger team to have more puzzles built like this without compromising the number of unspoiled users who can solve the puzzles.
2-14 was, to us, an absolute banger of a puzzle, and we really believed it was one of the better puzzles in the hunt, if a little on the easy side. As I mentioned in the puzzle's Author's Notes, it had been written in 6 hours the night before the hunt was due to start—the key difference, though, as pointed out by lovemathboy, is that it was written by the two of us: lmb threw out various ideas and provided the picture assets for two puzzles, while I handled the word-based sections, and stayed up to compile the entire document. By being able to bounce ideas off each other live, we were able to quickly eliminate poor ideas before they made it to the document and stayed there through inertia.
If we do do this again in coming years, we're hoping to be able to assemble a larger team to have more puzzles built like this without compromising the number of unspoiled users who can solve the puzzles.
Crazy Bracket: when a puzzle's harder than the setter can handle
The original idea of 2-16 was to make it the absolute hardest puzzle in the hunt. With my (actually decent) idea of a March Madness-themed puzzle, there were many ways to go, and I unfortunately took it down an over-convoluted path with the hopes of making it "hard enough". Add in a few inaccurate and vague crossword clues, and you find out why this puzzle wasn't forward-solved until 113 and a half hours into the hunt: not because it was hard, but because it was unjustifiably winding and twisted. Moving forward, I've personally learned an important lesson in building puzzles within my own league, and hopefully this sort of puzzle can be avoided in future.
The original idea of 2-16 was to make it the absolute hardest puzzle in the hunt. With my (actually decent) idea of a March Madness-themed puzzle, there were many ways to go, and I unfortunately took it down an over-convoluted path with the hopes of making it "hard enough". Add in a few inaccurate and vague crossword clues, and you find out why this puzzle wasn't forward-solved until 113 and a half hours into the hunt: not because it was hard, but because it was unjustifiably winding and twisted. Moving forward, I've personally learned an important lesson in building puzzles within my own league, and hopefully this sort of puzzle can be avoided in future.
SITE AND BOT
While the puzzle-writing was going on, talkingtree was working on getting the site ready for the event, as he'd volunteered earlier. Everything was going smoothly: he'd picked out a design scheme he liked, puzzles were being added to the site on a regular basis, and he had the teams and leaderboard mechanics in place. The only issue was that he couldn't seem to quite get the site to actually go online, but we figured that was a minor detail we could figure out near the end with the help of a few friends or with some online tutorials. (Spoiler: it wasn't.)
The week leading up to the hint saw general anxiety levels rising as continued attempts to get the site to just go up dammit repeatedly failed, leading to the first postponement of the hunt by 12 hours. In this 12 hours, it became apparent that the site wasn't going to go up, so we wrote SPH Bot, intending for it to be able to check answers, store team progress, and dish out additional puzzles as the hunt progressed. However, the SQL database we'd set up to store points somehow broke too, and after hours of attempted bug-fixing, we gave up and settled on a semi-manual format. During the hunt, the bot tagged all Staff whenever a team got a correct answer, letting us know when and where to distribute new puzzles. This wasn't as bad as I had feared—thanks to the effort of all our hardworking staff, we managed the event (in my opinion!) pretty well, with most hints being answered in under 5 minutes of being asked due to staff being on call almost 24/7.
MOVING FORWARD: 2019?
We've been asked a few times if this event will take place again next year, and the general consensus among the staff is that it would be a good idea, but moreso if we had more people to help.
In light of this, we're opening up our inboxes to user submissions. If you would like to join the puzzle team for SPH 2019, please send me a PM on Discord (Level 51#0375) with a puzzle you've written. The puzzle can be anything you like, as long as it's interesting, fair, non-trivial, and outputs a single English word or phrase as the answer. (Also, prepare a quick write-up of the solution to your puzzle.) We'd love to add more people to our team for next year!
Additionally, if you'd like to help out with the site next year to heighten our odds of actually getting a site going, feel free to contact me via Discord too!
STAFF HIGHLIGHT REEL
The staff had a lot of fun during this hunt, mostly by reading really weird answer submissions. Here's some interesting answer submissions we received, as well as some fun moments we witnessed:
(Submissions not necessarily shown in chronological order.)
Double or Nothing
[11:00 PM] SPH Bot: splashattacklols guessed porn controller for doubleornothing
A Hating at the Carthorses
[12:26 PM] SPH Bot: Decrypting Deoxys guessed ghost of the china sea for ahatingatthecarthorses
Bridging the Gap
[9:29 PM] SPH Bot: Big Brain Theory (We Are Really Smart) guessed spanishflu for bridgingthegap
Sense of Belonging
[9:30 PM] SPH Bot: Big Brain Theory (We Are Really Smart) guessed electronvolt for senseofbelonging
[11:12 PM] SPH Bot: BHad BHiatches guessed sigma phi epsilon for senseofbelonging
Ambidextrous
[9:13 PM] SPH Bot: Team Ezra guessed sph bots birthday for ambidextrous
[11:29 PM] SPH Bot: Team Ezra guessed sph bot is a good guy for ambidextrous
[8:16 AM] SPH Bot: Team Ezra guessed
b a start for ambidextrous
[8:49 PM] SPH Bot: Team Ezra guessed delete yourself for real this time goddamn robot for ambidextrous
[8:49 PM] SPH Bot: Team Ezra guessed i'm not done with you yet for ambidextrous
A Brief Oration
[10:47 PM] SPH Bot: Reborn's Most Trusted Citizens guessed serebii hit squad for abrieforation
[9:11 PM] SPH Bot: atrocityhal guessed russiahackedusa for abrieforation
[11:53 PM] SPH Bot: Black Rover (TV Size) guessed spaceanaltheives for abrieforation
[8:00 AM] SPH Bot: Firebot Development Lads guessed the spanish inquisition for abrieforation
Fan Service
[7:54 PM] SPH Bot: Team Ezra guessed roadworthiness for fanservice
[8:25 PM] SPH Bot: Team Ezra guessed neurotransmitter for fanservice
Namesakes
[1:28 PM] SPH Bot: Team Ezra guessed sophie's choice for namesakes
[1:52 PM] SPH Bot: Team Ezra guessed psychoanal for namesakes
[4:41 AM] SPH Bot: Squirtle Squad guessed pablo diego josé francisco de paula juan nepomuceno maría de los remedios cipriano de la santísima trinidad ruiz y picasso for namesakes
Gender Studies
[9:58 PM] SPH Bot: Bobby Squad guessed bank jew for genderstudies
[9:58 PM] SPH Bot: Bobby Squad guessed banking jew for genderstudies
Expansion
[6:04 PM] SPH Bot: Team Ezra guessed answer recreation for expansion
[6:05 PM] SPH Bot: Team Ezra guessed creating answer for expansion
[6:07 PM] SPH Bot: Team Ezra guessed recreation answer for expansion
(they solve it at 6:14)
[6:15 PM] SPH Bot: Team Ezra guessed uncreate yourself please for expansion
word f(ind)
[4:21 AM] SPH Bot: Team Ezra guessed abduct anuses for wordfind
[10:15 PM] SPH Bot: Team Ezra guessed one is toucan for wordfind
CAP Contributor
[8:11 AM] SPH Bot: Decrypting Deoxys guessed body camo u flame for capcontributor
[8:48 AM] SPH Bot: P Squad guessed gory camouflage for capcontributor
[9:44 AM] SPH Bot: P Squad guessed army camouflage for capcontributor
[9:47 AM] SPH Bot: P Squad guessed holy camouflage for capcontributor
Spacetime
[2:00 AM] SPH Bot: Detective Pikachus guessed palkiadialga for spacetime
[1:20 AM] SPH Bot: Team Ezra guessed brothel for spacetime
Building Bridges
[7:52 AM] SPH Bot: Tex_tSegmentation guessed hashtag swag for buildingbridges
[1:39 AM] SPH Bot: Team Ezra guessed level 51 for buildingbridges
[4:56 PM] SPH Bot: Team Ezra guessed you're a foe don't misunderstand this for buildingbridges
World's Longest Port
[7:35 PM] SPH Bot: White Maria guessed shanghai for worldslongestport
[4:57 PM] SPH Bot: Big Brain Theory (We Are Really Smart) guessed port of shanghai for worldslongestport
[10:26 PM] SPH Bot: Get Off My Lawn (GOML) guessed shanghai for worldslongestport
[9:54 AM] SPH Bot: Team Ezra guessed shanghai for worldslongestport
[11:53 AM] SPH Bot: Reborn's Most Trusted Citizens guessed tamale for worldslongestport
[11:26 PM] SPH Bot: Get Off My Lawn (GOML) guessed tamale for worldslongestport
[8:55 AM] SPH Bot: Team Ezra guessed late meta for worldslongestport
[9:52 AM] SPH Bot: Team Ezra guessed tamale for worldslongestport
[10:06 AM] SPH Bot: P Squad guessed lame meta for worldslongestport
[11:59 AM] SPH Bot: Dan City guessed tamale for worldslongestport
[3:03 AM] SPH Bot: Scary Birb Squad guessed latamelatalalatamela for worldslongestport
[1:06 PM] SPH Bot: P Squad guessed talonflameloettangelarvestangelarvestangelarvestangelarvestalonflameloettalonflameloettalonflameloettangelarvestangelarvesta for worldslongestport
Mint Condition
[5:33 AM] SPH Bot: Team Reina guessed the price is right for mintcondition
[5:33 AM] SPH Bot: Team Reina guessed the price is wrong for mintcondition
Double or Nothing
[11:00 PM] SPH Bot: splashattacklols guessed porn controller for doubleornothing
A Hating at the Carthorses
[12:26 PM] SPH Bot: Decrypting Deoxys guessed ghost of the china sea for ahatingatthecarthorses
Bridging the Gap
[9:29 PM] SPH Bot: Big Brain Theory (We Are Really Smart) guessed spanishflu for bridgingthegap
Sense of Belonging
[9:30 PM] SPH Bot: Big Brain Theory (We Are Really Smart) guessed electronvolt for senseofbelonging
[11:12 PM] SPH Bot: BHad BHiatches guessed sigma phi epsilon for senseofbelonging
Ambidextrous
[9:13 PM] SPH Bot: Team Ezra guessed sph bots birthday for ambidextrous
[11:29 PM] SPH Bot: Team Ezra guessed sph bot is a good guy for ambidextrous
[8:16 AM] SPH Bot: Team Ezra guessed
b a start for ambidextrous[8:49 PM] SPH Bot: Team Ezra guessed delete yourself for real this time goddamn robot for ambidextrous
[8:49 PM] SPH Bot: Team Ezra guessed i'm not done with you yet for ambidextrous
A Brief Oration
[10:47 PM] SPH Bot: Reborn's Most Trusted Citizens guessed serebii hit squad for abrieforation
[9:11 PM] SPH Bot: atrocityhal guessed russiahackedusa for abrieforation
[11:53 PM] SPH Bot: Black Rover (TV Size) guessed spaceanaltheives for abrieforation
[8:00 AM] SPH Bot: Firebot Development Lads guessed the spanish inquisition for abrieforation
Fan Service
[7:54 PM] SPH Bot: Team Ezra guessed roadworthiness for fanservice
[8:25 PM] SPH Bot: Team Ezra guessed neurotransmitter for fanservice
Namesakes
[1:28 PM] SPH Bot: Team Ezra guessed sophie's choice for namesakes
[1:52 PM] SPH Bot: Team Ezra guessed psychoanal for namesakes
[4:41 AM] SPH Bot: Squirtle Squad guessed pablo diego josé francisco de paula juan nepomuceno maría de los remedios cipriano de la santísima trinidad ruiz y picasso for namesakes
Gender Studies
[9:58 PM] SPH Bot: Bobby Squad guessed bank jew for genderstudies
[9:58 PM] SPH Bot: Bobby Squad guessed banking jew for genderstudies
Expansion
[6:04 PM] SPH Bot: Team Ezra guessed answer recreation for expansion
[6:05 PM] SPH Bot: Team Ezra guessed creating answer for expansion
[6:07 PM] SPH Bot: Team Ezra guessed recreation answer for expansion
(they solve it at 6:14)
[6:15 PM] SPH Bot: Team Ezra guessed uncreate yourself please for expansion
word f(ind)
[4:21 AM] SPH Bot: Team Ezra guessed abduct anuses for wordfind
[10:15 PM] SPH Bot: Team Ezra guessed one is toucan for wordfind
CAP Contributor
[8:11 AM] SPH Bot: Decrypting Deoxys guessed body camo u flame for capcontributor
[8:48 AM] SPH Bot: P Squad guessed gory camouflage for capcontributor
[9:44 AM] SPH Bot: P Squad guessed army camouflage for capcontributor
[9:47 AM] SPH Bot: P Squad guessed holy camouflage for capcontributor
Spacetime
[2:00 AM] SPH Bot: Detective Pikachus guessed palkiadialga for spacetime
[1:20 AM] SPH Bot: Team Ezra guessed brothel for spacetime
Building Bridges
[7:52 AM] SPH Bot: Tex_tSegmentation guessed hashtag swag for buildingbridges
[1:39 AM] SPH Bot: Team Ezra guessed level 51 for buildingbridges
[4:56 PM] SPH Bot: Team Ezra guessed you're a foe don't misunderstand this for buildingbridges
World's Longest Port
[7:35 PM] SPH Bot: White Maria guessed shanghai for worldslongestport
[4:57 PM] SPH Bot: Big Brain Theory (We Are Really Smart) guessed port of shanghai for worldslongestport
[10:26 PM] SPH Bot: Get Off My Lawn (GOML) guessed shanghai for worldslongestport
[9:54 AM] SPH Bot: Team Ezra guessed shanghai for worldslongestport
[11:53 AM] SPH Bot: Reborn's Most Trusted Citizens guessed tamale for worldslongestport
[11:26 PM] SPH Bot: Get Off My Lawn (GOML) guessed tamale for worldslongestport
[8:55 AM] SPH Bot: Team Ezra guessed late meta for worldslongestport
[9:52 AM] SPH Bot: Team Ezra guessed tamale for worldslongestport
[10:06 AM] SPH Bot: P Squad guessed lame meta for worldslongestport
[11:59 AM] SPH Bot: Dan City guessed tamale for worldslongestport
[3:03 AM] SPH Bot: Scary Birb Squad guessed latamelatalalatamela for worldslongestport
[1:06 PM] SPH Bot: P Squad guessed talonflameloettangelarvestangelarvestangelarvestangelarvestalonflameloettalonflameloettalonflameloettangelarvestangelarvesta for worldslongestport
Mint Condition
[5:33 AM] SPH Bot: Team Reina guessed the price is right for mintcondition
[5:33 AM] SPH Bot: Team Reina guessed the price is wrong for mintcondition
seeing the first solve from Hyunmin Fan Club just 4 minutes after the hunt started
slightly ambiguous puzzle answer
slightly ambiguous puzzle answer, part 2
that's one way to interpret "fan service", i guess
and one way to interpret that line in namesakes, while we're at it
this is why there's 20 guesses!
uflage isnt a word ugh
(it technically is, if one starts sufficiently far in.)
pictured: me in math class
i'm just glad they had crossing words before they tried entering this
two different responses to frustrating a solver
this situation occurred at least five times. i've never shouted at my laptop more in a single week
The 24 puzzles featured in this hunt weren't the only ones we made. Here's two select puzzles we had originally made for the hunt, but cut for one reason or another:
- Lip Service: originally 2-1
- Extremely Slow Cookie Clicker: originally 2-14
STATS
(also contains puzzle-specific spoilers: open at your own risk)
Puzzles by solve rates (numbers represent correct solves and total guesses, including correct solves):
On the other hand, we had some very clean puzzles—Casting Call and Mint Condition, for example, have fairly low wrong answer rates, since for both puzzles there's not really much to guess with until the whole puzzle is already solved. For puzzles with fairly low correct answer rates, Suspect Test and Crazy Bracket stick out, with single-digit solves each. Suspect Test's result, in particular, was likely due to its base mechanic of a cryptic crossword, which would have been completely alien to most teams, and hard to get to grips with for a beginner; Crazy Bracket, as mentioned above, was probably overly convoluted, and most teams solved the metapuzzle without it, making putting in extra effort to solve it a redundant and unattractive option.
Puzzles by first three finishers:
Puzzles by number of responses calling it a favorite/least favorite puzzle (number before slash represents times chosen as favorite puzzle, number after slash repersents times chosen as least favorite puzzle):
On a positive note, Blind Spots, Spacetime, and Casting Call each received perfect records, and tied for the most-favorited puzzle; word f(ind) carries the same number of favorites, but appears to have been pretty polarizing, with some solvers disliking it due to its finicky nature—or perhaps it's an in-built, instinctive hatred for math?
- 1-1 (Double or Nothing): 46/273
- 1-2 (A Hating at the Carthorses): 49/73
- 1-3 (Bridging the Gap): 54/82
- 1-4 (Ambidextrous): 46/284
- 1-5 (Sense of Belonging): 41/55
- 1-6 (Blind Spots): 47/81
- 1-M (A Brief Oration): 42/112
- 2-1 (Fan Service): 28/104
- 2-2 (Namesakes): 34/94
- 2-3 (Gender Studies): 30/56
- 2-4 (Expansion): 23/48
- 2-5 (word f(ind)): 23/83
- 2-6 (CAP Contributor): 25/38
- 2-7 (Spacetime): 22/75
- 2-8 (Building Bridges): 17/25
- 2-9 (Up Against The Wall): 16/44
- 2-10 (Basic Counting): 22/100
- 2-11 (Suspect Test): 8/35
- 2-12 (World's Longest Port): 15/77
- 2-13 (Casting Call): 15/21
- 2-14 (Mint Condition): 14/19
- 2-15 (The Last Word): 14/36
- 2-16 (Crazy Bracket): 6/43
- 2-M (The Final Battle): 13/68
On the other hand, we had some very clean puzzles—Casting Call and Mint Condition, for example, have fairly low wrong answer rates, since for both puzzles there's not really much to guess with until the whole puzzle is already solved. For puzzles with fairly low correct answer rates, Suspect Test and Crazy Bracket stick out, with single-digit solves each. Suspect Test's result, in particular, was likely due to its base mechanic of a cryptic crossword, which would have been completely alien to most teams, and hard to get to grips with for a beginner; Crazy Bracket, as mentioned above, was probably overly convoluted, and most teams solved the metapuzzle without it, making putting in extra effort to solve it a redundant and unattractive option.
Puzzles by first three finishers:
- 1-1 (Double or Nothing): Hyunmin Fan Club, White Maria, Eggplant Parm
- 1-2 (A Hating at the Carthorses): Hyunmin Fan Club, Level 51 Fan Club, The Quinn Specials
- 1-3 (Bridging the Gap): Hyunmin Fan Club, Team Dospacito, White Maria
- 1-4 (Ambidextrous): Hyunmin Fan Club, Level 51 Fan Club, White Maria
- 1-5 (Sense of Belonging): Hyunmin Fan Club, Big Brain Theory, Sgt. Crepe's Lonely Weirdough Club Band
- 1-6 (Blind Spots): Hyunmin Fan Club, Level 51 Fan Club, Team Dospacito
- 1-M (A Brief Oration): White Maria, Hyunmin Fan Club, Cofagrigus Picross
- 2-1 (Fan Service): White Maria, Cofagrigus Picross, Hyunmin Fan Club
- 2-2 (Namesakes): White Maria, Hyunmin Fan Club, Sgt. Crepe's Lonely Weirdough Club Band
- 2-3 (Gender Studies): White Maria, Hyunmin Fan Club, Tex_tSegmentation
- 2-4 (Expansion): White Maria, Cofagrigus Picross, Team Reina
- 2-5 (word f(ind)): White Maria, Hyunmin Fan Club, Cofagrigus Picross
- 2-6 (CAP Contributor): White Maria, Hyunmin Fan Club, Cofagrigus Picross
- 2-7 (Spacetime): Cofagrigus Picross, Hyunmin Fan Club, Team Reina
- 2-8 (Building Bridges): White Maria, Hyunmin Fan Club, Cofagrigus Picross
- 2-9 (Up Against The Wall): Tex_tSegmentation, Reborn's Most Trusted Citizens, Level 51 Fan Club
- 2-10 (Basic Counting): Hyunmin Fan Club, White Maria, Detective Pikachus
- 2-11 (Suspect Test): Hyunmin Fan Club, White Maria, Level 51 Fan Club
- 2-12 (World's Longest Port): Hyunmin Fan Club, White Maria, Detective Pikachus
- 2-13 (Casting Call): Hyunmin Fan Club, White Maria, Detective Pikachus
- 2-14 (Mint Condition): Hyunmin Fan Club, White Maria, Detective Pikachus
- 2-15 (The Last Word): Hyunmin Fan Club, Detective Pikachus, White Maria
- 2-16 (Crazy Bracket): Detective Pikachus, Hyunmin Fan Club, Big Brain Theory (We Are Really Smart)
- 2-M (The Final Battle): Hyunmin Fan Club, Detective Pikachus, Level 51 Fan Club
Puzzles by number of responses calling it a favorite/least favorite puzzle (number before slash represents times chosen as favorite puzzle, number after slash repersents times chosen as least favorite puzzle):
- 1-1 (Double or Nothing): 3/1
- 1-2 (A Hating at the Carthorses): 1/1
- 1-3 (Bridging the Gap): 1/1
- 1-4 (Ambidextrous): 0/4
- 1-5 (Sense of Belonging): 0/0
- 1-6 (Blind Spots): 5/0
- 1-M (A Brief Oration): 1/6
- 2-1 (Fan Service): 1/3
- 2-2 (Namesakes): 4/1
- 2-3 (Gender Studies): 1/0
- 2-4 (Expansion): 2/3
- 2-5 (word f(ind)): 5/4
- 2-6 (CAP Contributor): 0/4
- 2-7 (Spacetime): 5/0
- 2-8 (Building Bridges): 2/1
- 2-9 (Up Against The Wall): 3/1
- 2-10 (Basic Counting): 1/1
- 2-11 (Suspect Test): 3/4
- 2-12 (World's Longest Port): 3/7
- 2-13 (Casting Call): 5/0
- 2-14 (Mint Condition): 1/0
- 2-15 (The Last Word): 0/0
- 2-16 (Crazy Bracket): 0/0
- 2-M (The Final Battle): 0/0
On a positive note, Blind Spots, Spacetime, and Casting Call each received perfect records, and tied for the most-favorited puzzle; word f(ind) carries the same number of favorites, but appears to have been pretty polarizing, with some solvers disliking it due to its finicky nature—or perhaps it's an in-built, instinctive hatred for math?
Teams by finish duration:
Miscellaneous:
- Hyunmin Fan Club: 0d 22h 53min 35s
- Detective Pikachus: 1d 19h 3min 26s
- Level 51 Fan Club: 2d 14h 6min 38s
- White Maria: 2d 16h 0min 26s
- Reborn's Most Trusted Citizens: 3d 8h 11min 1s
- Tex_tSegmentation: 3d 16h 50min 46s
- Team Reina: 5d 13h 28min 32s
- Big Brain Theory (We Are Really Smart): 7d 12h 49min 50s
- Scary Birb Squad: 8d 2h 7min 43s
- Get Off My Lawn (GOML): 9d 2h 49min 52s
- Team Ezra: 9d 5h 13min 32s
- Team Dospacito: 9d 10h 13min 50s
- P Squad: 9d 12h 48min 9s
Miscellaneous:
- Most guesses of any team: Team Ezra, 173 guesses
- Fewest guesses of a finishing team: Hyunmin Fan Club, 34 guesses
- Fastest solve: Hyunmin Fan Club on A Hating at the Carthorses, 4 minutes
- Only team to correctly solve Ambidextrous on their first guess: Black Rover (TV Size)
- Best response to Up Against The Wall's replay request: Team Ezra, http://replay.pokemonshowdown.com/gen7anythinggoes-780974794
Overall stats:
Top SPH Bot killers:
- Total solves recorded: 650 (283 R1, 42 R1-M, 312 R2, 13 R2-M)
- Total guesses recorded: 1926
- Teams solving at least one puzzle: 55
Top SPH Bot killers:
- Laptop mysteriously went to sleep: 3 times
- Cat unplugged laptop, causing it to die: 2 times
- Laptop changed IP address, causing bot to disconnect: 1 time
CREDITS
I'd like to express my deepest gratitude especially to the team which worked with me through the run-up to the event and through the event itself: talkingtree, lovemathboy, Scarf Nihilego, and CheeseMuffin. Thanks for all your hard work and effort, couldn't have done it without you guys. Take a bow!
Project Lead
Level 51
Tech Lead
talkingtree
Puzzle Design
Level 51, lovemathboy, Scarf Nihilego
Testsolving
CheeseMuffin, Level 51, lovemathboy, Scarf Nihilego, talkingtree
Spriting
Level 51
Additional credits to The Spriter's Resource for ripped tilesets:
- RB Tileset: Badassbill
- Crystal Tileset: Superjustinbros
- Emerald Tileset: Heartlessdragon
Special Thanks
- Former staff Nia and teal6
- Da Letter El for approving and helping to assist the event to completion
- shade for coordinating the prizes
- P Squared and The Dutch Plumberjack for their help with publicizing this event
- My brother for designing this amazing logo
- All our solvers; we couldn't have made this event a success without you!
Last edited:
