Thank you so much to everyone who took a crack at my Combination Lock Puzzle! I had a lot of fun receiving e-mails and Facebook messages and tracking everyone’s progress! There was a great feeling of excitement to see who would be the five prize-winners, a race to the finish! I wasn’t expecting that sort of atmosphere, and it was a lovely surprise.
And thank you to everyone who gave me feedback on their puzzle-solving process. I really learned some valuable things about puzzle-making and solving, and I feel confident that my next puzzles will be even better =) Now I’d like to share the puzzle solutions for those who are tired of beating your head against the wall, and I’d also like to share some of the feedback I got and what it has taught me. So be sure you’re ready to read all the answers before clicking the spoiler button below. (You can still go back and try to solve the puzzle for fun if you like, though I’m not giving any more prizes).
Onward to the answers and analysis!
First of all, my thought process on designing these puzzles was very simple. I just needed a fun way for people to come up with three numbers for the combination lock. I didn’t even care what the numbers were. I started brainstorming ideas and just went with the flow.
Design: The puzzle for the first Right number ended up a lot different than my original idea. There is a puzzle in the Professor Layton games where you are given a simple line maze to solve. When you solve the maze, you see that the path you drew actually makes a 2-digit number. I started designing something similar to that in Photoshop, but it seemed like too much work to actually design and execute the whole maze. I was anxious to get these puzzles up (and I’m lazy), so when I started to see a different puzzle forming in the shapes I was making, I decided to go with that instead. I had made a border and a center line for my maze, and I could see kind of a negative-image shape forming, so I went ahead and tweaked it until it became the number 08. I was a little worried this one would be too obvious and not fun.
Feedback: Several reactions to this one. Some people said their first reaction was just confusion, since it looks pretty abstract. From there, most people got 12 as their first answer, the big block representing a 10 (or a 1 in the tens place), and the two small blocks representing 2. So many people got this answer that my eventual hint was that the answer was NOT 12. A few people thought 12 seemed too obvious from the start and kept looking for another solution until they saw the 08. Others looked again with fresh eyes or from far away and saw the 08. Jen commented that once she saw the 08 it couldn’t be unseen, and that made her feel confident that it was the correct solution. One other comment said that the variety of possible answers (12, 08, or 03) was a little frustrating since it was hard to check answers (you need all three answers to be correct, and the second puzzle had even more possible answers to choose from).
Analysis: This was by far the most successful puzzle. It stumped a few people at first, but almost everyone I talked to eventually found the answer and felt pretty confident that it was right, and said it was fun to solve. I was amused at the 12 answer since I didn’t even see that when I was making the puzzle. I saw the 03, but I made sure that the line width between the two numbers was accurate for 0 and 8, not for 0 and 3. I’m curious what this puzzle would look like with a different color scheme, if the whole negative-image thing comes across better or worse with certain color combinations.
- It’s possible to design a puzzle that gives solvers an a-ha moment without you, the designer, even realizing it.
- I’m bad at judging the difficulty and fun of a puzzle I design.
- A puzzle gets frustrating when it’s hard to check your answer and the answer isn’t 100% clear.
Design: I recently mentioned a flash game I played called Alice is Dead. There is a similar puzzle in that game which inspired this one. In the game, five apparently unrelated words are written top to bottom. A note nearby gives you the hint to take the second letter of each word to spell out a new word, the answer to the puzzle. I decided to use the number TWELVE since it was kind of short, and I chose the third letter of each word instead of the second. From there, I actually started trying to write out a grammatically correct sentence with my words, but that wasn’t working out. I had written OFTEN and NEVER, and that seemed like a good jumping-off point to write words that all had to do with time/frequency. I really had no purpose for doing this, I just thought related words would look better and seem more intersting than a list of random, unrelated words (and it was kind of a puzzle for me to find time-related words that fit the space, that’s why the very weird and too long “Half of the time” is in there). For the alignment, I thought left-aligned would make it too easy. I considered center alignment for a while, but eventually I thought right-align looked the coolest and would throw people off (since I guess I thought this puzzle was too easy??). At one point I tried getting the words to make a big 3 shape as a clue, but it wasn’t working out so I ditched that idea and decided a clue wasn’t necessary.
Feedback: It immediately became apparent that this puzzle was not so great! Everybody had trouble with this one, and it’s no wonder why (more on that in the analysis section). There were many different approaches to solving this one, such as: counting all the letters, taking half of the letters, rearranging the words in order of frequency, counting syllables, representing the words as numbers, trying to anagram certain strings of letters, somehow Magic Square-ing things, and probably many more methods that I don’t even want to hear about because they’ll make me cringe! When it was time for hints, I said that the “magic number” would help. This hint was disastrous, as it turns out that the phrase “magic number” doesn’t mean very much to most people! I’ve always connected it to the School House Rock song, “Three is a Magic Number,” but other people thought of 42 (Hitchhiker’s Guide), 8 (infinity symbol and Magic Eight Ball), 39 (the highest number on the combination lock), or no number in particular. So basically, instead of giving a hint, I gave a bunch of red herrings! Incredibly, despite all this, several people managed to solve the puzzle without ever seeing the word TWELVE spelled out. Several people decided that since the puzzle was about time, that the numbers 24 and 12 might be significant, or even “magic” numbers. Luckily, one of them was.
Analysis: Right after I posted this puzzle, I thought to myself “There is way too much information in this puzzle.” But then I had Nick look at it and he was able to solve it in a few minutes, so I felt reassured. But he had me sitting right there next to him to give, or not give, certain amounts of feedback on his guesses and paths of thinking, so that wasn’t an accurate test of the puzzle at all. The frustrations over this puzzle started coming in pretty quickly, and it became clear that I had designed it poorly. Letters and words have so many potential meanings to begin with (definitions, syllables, numeric associations, the ability to be rearranged into new words, the ability to be counted, etc.) that you can’t just throw six of them in a row and expect people to come out with just one right answer. I think simply changing the alignment would have improved this one so much. In the Alice is Dead version, the words are roughly center aligned with a bit of variation, but not so much that it’s hard to read (like mine). When I played that version, the word didn’t jump out at me at all, so I’m not sure why I thought it would jump out at anyone else in my version. I think part of me just wanted another factor to help make my version different from the Alice one, which is not a good reason to alter a puzzle. I still like that I made the words all have similar meanings. I think that aspect added to the feel of the puzzle, but it was one of WAY too many red herrings. The most incredible thing was that people used the arbitrary meanings of the words to get the answer 12. If the answer had been 17 or something, or if my words had all been pizza toppings instead of measures of frequency, there would have been no hope! That was a happy accident.
- When you make a puzzle, be sure to take some time to think of all the possible answers someone might find before they find the correct one. If there are too many, it’s time to redesign.
- Be very careful with words and letters, they’re very powerful and hold a lot of meaning.
- The more data your puzzle has, the more opportunities your players have to get de-railed.
- Be careful when adding arbitrary information or meanings to your puzzle for aesthetic purposes. Find a way to let the player know what is and isn’t important to the solution.
- Avoid over-obfuscating. A player won’t feel that bad if they solve a puzzle that was too easy, but they might get frustrated and give up if a puzzle is too hard (especially when it is hard because it an unsatisfying way).
- Try not to use pop culture as a hint or puzzle element, unless the thing you’re referencing is the top Google result on a search for that hint.
- I’m bad at judging the difficulty and fun of a puzzle I design.
Design: The concept was simple, I just wanted to use Morse code somehow to make a number. When I started looking at the codes for numbers, a few of them (like 3 and 7) were mirror images of each other. I thought that would make a nice design for a puzzle, and decided on the number 37. My intention was that the visual design would make it look less like Morse code at first glance, similar to the way the first puzzle didn’t look like 08 at first.
Feedback: The easiest puzzle of the three, by far. Pretty much everyone recognized right away that it was Morse code, and just had to look up a chart and translate. As Matt put it, there wasn’t really anything to figure out here.
Analysis: I was a little disappointed that this one was so easy. It was difficult to judge the design. I could still definitely tell it looked like Morse code, but I could also totally see the 08 in the first puzzle. If I had to do this one over again, I might actually make it 33 instead of 37. Having similar shapes grouped together might make the image look more like a math problem or something (kind of like the squares seeming to represent 12 in the first puzzle), and might give people an a-ha moment when they realize it’s Morse code. Then again, it might just give too many possible answers. In this particular case, however, even though I was disappointed, I think it was for the best that this one was so easy. The other two puzzles had people guessing a lot of different answers, and since it was so hard to check answers, at least one of them needed to be a constant.
- Sometimes it’s okay to have an easy puzzle.
- Sometimes you might think you’ve made a puzzle, but you really haven’t given the player anything to solve. Try to figure out what the core question of your puzzle is and test it. In this case, the core question of my puzzle was “What do these shapes mean?” That turned out to be a really easy question.
- If the puzzle design doesn’t seem ideal, keep experimenting and trying to think of new variations, one of them might be better.
Conclusions: So I ended up with one pretty good puzzle, one stupid hard/bad puzzle, and one too-easy puzzle. The easy and good ones helped balance out the bad one a little bit, but not enough. One big oversight with the game overall was in not considering how hard it would be to test your answers. Since you had to have all three answers right to open the lock, it made things really hard when two of the puzzles had more than one potential answer. Granted, this made it harder for anyone to just brute-force the answers, but it was not ideal. Maybe in the future, I would add sort of a mastermind feedback where it told you whether you had any of the right numbers and in any of the right spots.
I also didn’t really consider how people might have expected the puzzles to be related to each other, or related to the padlock part of the puzzle. This didn’t come up much in the feedback, but it got me thinking about the idea in general.
I would say that the biggest lesson I learned was to TEST YOUR PUZZLES, but I think I kind of considered this little contest to be my test =) All the people who won are basically the people I would have asked to test the puzzles anyway! Maybe when I have more readers someday, those people will be my inner testing circle, and I can refine all the puzzles before posting them for everyone else to try.
All of this analysis might seem a little overboard for three puzzles I just made up, but I’m really interested in what makes a good puzzle or a bad puzzle, and I’d like to make more and better puzzles in the future. Thanks again to everyone who played and gave feedback! The general consensus seems to be that this was fun and people would be interested in seeing puzzles like these more regularly on Clavis Cryptica, so expect more in the future! Also, expect updates as I work on my big goal of making a small-but-complete puzzle/mystery game in HTML5! Thanks!