Detailed analysis of a dnd puzzle room
Jun. 10th, 2019 04:34 pm![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
jackSome spoilers for both groups of players (telling you stuff the other group found out, and confirming which of your guesses were right.) If you're interested in what went on behind the scenes, please ask me and I can talk through a non-spoiler version (or if you read this by accident, let me know, it's not a big deal as long as I know).
I talked about the puzzle rooms before, but I'm going to go through in more detail, to compare the amount of type of prep I did to that of other GMs reading.
What I wrote down
I don't have an exact copy, but it was really brief. For room one, something like: 100 wooden boxes scattered around the room. A little plant in a pot that can answer questions about them. It alternates between true and false in answers to questions. It otherwise tries to avoid declarative sentences. One has the Macguffin in (brief notes). Three others have some other treasure. Each wrong one gives a shock, or a 1/10 chance of a snake.
For room two, five alters, five bowls, five knives, all engraved. A message on the wall, saying "Cut yourself, let a drop of blood into the bowl on the altar. If you choose right, you will gain the favour of the gods. If you choose wrong, you will bear the brunt." Then I had a chart of the effect each one had, with "good" meaning a shower of golden sparks, and using all three "good" together conjures the macguffin statuette. I won't reproduce the effects here, but they were mostly pretty simple, like "lose 1d6 hp", a few duplicates, one or two random-magical-effects table.
What I had in mind but didn't write down
When I'm expecting to run something myself, I've learned what I need to write down, and what will probably come out roughly the same if I write it down or not, once my mind has the general idea of the room.
In this case, I had an idea that the boxes varied in size, some crate-like, some about 10cm across. They were piled all over, stacked on each other. There might be some pattern to the layout just to make it easier to refer to them (say, six separate collections of boxes).
I had an idea the plant would be able to answer questions. I had a couple of ideas about it's personality. I thought it was probably derived from some faerie plant, but not otherwise affiliated with faerie.
I thought there might be a few other rewards as well in other boxes.
I had a few images of how the whole thing would look.
I had the general idea that the players might have to figure out the lies and do a divide-the-boxes-into-two-groups-repeatedly 20-questions binary-chop style. I assumed they might find some other ways to to investigate the inside of the boxes (they had a few abilities that might find things out), and I tried to construct things
For room 2, I had a clear idea of how it would look. I knew that you'd only get a magical effect if you cut yourself with a knife, dropped a drop of blood into a bowl, on one of the altars, using one of each type of object, else nothing would happen. I'd thought through some of the same questions as above. I expected the party would need to solve it like a puzzle, trying multiple combinations, once they got the basic idea, but that they might find another approach.
For both, there's some things I didn't think consciously about, but I implicitly needed to set the rules of the challenge, of finding a macguffin. In both cases, that was partially ambiguous, but it did just happen automatically -- probably because in the labyrinth, whenever I describe the players finding anything other than a tunnel, they usually are primed to want to explore it.
How it went
Several things changed as the players encountered the room, but as I thought, filling in the details worked fine. Describing the box room I wanted to fill out the description of the room a bit, so I put shelves of books round, which seemed appropriate, but then I needed to fill in information in them, and because the players were interested, I changed/added some extra rewards in the boxes which were books, so the players got some of the book excitement it looked like the room was promising them.
The random encounter table told me the PCs would meet one of the NPCs, a minotaur who works to maintain the labyrinth's smooth running. So I had the custodian be killed and the minotaur working on regrowing it. The minotaur took on the role of answering the questions, which helped in some ways (she could offer editorial asides) and hurt in others (it wasn't immediately clear when she was following the rules of the room and when she wasn't).
Several other things went as I expected: The players got the general idea of finding the macguffin, they got the idea of getting zapped and trying to open some boxes but not too many. The players came up with several sensible ways of exploring the contents of the boxes without opening them, and they did, by turning-into-a-woodlouse, and by picking the boxes up. In those cases, I decided it worked on some of the boxes but not all, and rolled to see if they found any of the rewards. They tried a few things which maybe deserved to have worked, but I couldn't let work without obviating the entire challenge.
I made up the "shock" when you opened a wrong box, as 1d6 damage, and that felt about right. I'd specified a similar amount in the other room, and it seemed to work well. Enough to put people off, but not enough to be a danger unless they kept persevering after they were clearly getting too low on health.
I'd forgotten that one of the PCs was resistant to electricity, so after a while they made use of that, which counted as "a helpful approach I hadn't considered" :)
The altar room worked basically how I expected: the PCs got the idea, they tried several combinations, they considered a few false alleys, they got a bit low on health and decided to return later to be on the safe side, but had a good idea what the remaining possibilities to try were.
The box room was more awkward. The PCs constantly had good ideas how to approach the problem, but even after I managed to work in several hints, it hadn't been clear enough that some of the answers were unreliable and they could figure out which with the right approach. They were on the point of burning the boxes down (which would have worked, but done them quite a lot of damage as the zaps would still have happened). But I used the warlock familiar to deliver a big answer in exchange for some devil worship from one of the other players, which basically told them what to do, and was fairly satisfying in-world. So that mostly turned out well, but it did get bogged down for a while which I would ideally have avoided.
What I should have done
I should have explicitly thought was detect magic would show.
I should have had a more specific plan for "pick them up and shake them" as the most obvious course of action.
When I had to have a sensible plan not work, I should have done more to make it feel like a success somehow even if it didn't technically help (like having a bad skill fumble sometimes be a pratfall and sometimes be a dramatic last minute save, and most PCs should be wish-fulfilment-y quite often).
I should have made sure that if the players were completely stuck, there was a sensible way forward: included a way for them to get hints, or made sure there was an answer which could brute force a solution even if it cost them other resources (health, money, time); or set it up so leaving and coming back another time felt more reasonable.
I should have thought what happens if GP 1 had looked at the blood sacrifice room and said "looks evil, not touching it", whether I should have had some more clear way to show them it was ok to experiment (although that wasn't a problem in practice).
Write-up
I don't usually write up in this level of detail because obviously, it takes a lot longer than it did to write the planning! But it's interesting to do sometimes, so I can consciously notice what I was doing explicitly, and what I was implicitly doing in the background.
When I was less experienced, I would have needed to write down a lot of the specific things that here I just kept in my mind. Or if I was writing an adventure for someone else to run, I would have to spell out the extra detail I only verbalised now, because a lot of the good things about the encounter were the general idea of how it should play out, not the specifics.
Although even then, a lot of what went well was because I knew the players and know how to slot appropriate connections in as I went along. A published adventure can't easily do that. The best you could do is list the treasure as what was definitely there, and then offer the GM to fill in whatever felt appropriate, with some specific examples for them to use if they didn't have a better idea. Things like, here's some prompts (for things connected with the room, for things with an interesting backstory, for things that are of particular interest to the players somehow)
In terms of "puzzles" I think the most important thing is to give the players some experience of "trying different approaches and seeing what works", but make sure that the encounter being satisfying doesn't rely on them being able to figure it out logically. And single-answer puzzles are usually bad for this, unless you build an adventure about them with going asking different NPCs what they think the answer might be, etc. In this case, because I wanted a like-Labyrinth-film feel, I went as close to "single answer puzzle" as I could.
I think I learned how to do it better in future. Partly by using puzzles sparingly, but mostly, how to do it well when it comes up. Now I think of it, some of it is sleight of hand: it's giving the *experience* of solving a puzzle, but actually, having a fallback so it matters less whether you SUCCEED or not, as long as you give it a fair try.
Of course, that's for *most* DnD games where the PCs are supposed to be reasonably successful. Sometimes you want to play a gritty deadly tomb of horrors dark souls game, in which case you can go for "they may never solve it and that's ok, it's up to them if they want to push their luck or not" approach.
I talked about the puzzle rooms before, but I'm going to go through in more detail, to compare the amount of type of prep I did to that of other GMs reading.
What I wrote down
I don't have an exact copy, but it was really brief. For room one, something like: 100 wooden boxes scattered around the room. A little plant in a pot that can answer questions about them. It alternates between true and false in answers to questions. It otherwise tries to avoid declarative sentences. One has the Macguffin in (brief notes). Three others have some other treasure. Each wrong one gives a shock, or a 1/10 chance of a snake.
For room two, five alters, five bowls, five knives, all engraved. A message on the wall, saying "Cut yourself, let a drop of blood into the bowl on the altar. If you choose right, you will gain the favour of the gods. If you choose wrong, you will bear the brunt." Then I had a chart of the effect each one had, with "good" meaning a shower of golden sparks, and using all three "good" together conjures the macguffin statuette. I won't reproduce the effects here, but they were mostly pretty simple, like "lose 1d6 hp", a few duplicates, one or two random-magical-effects table.
What I had in mind but didn't write down
When I'm expecting to run something myself, I've learned what I need to write down, and what will probably come out roughly the same if I write it down or not, once my mind has the general idea of the room.
In this case, I had an idea that the boxes varied in size, some crate-like, some about 10cm across. They were piled all over, stacked on each other. There might be some pattern to the layout just to make it easier to refer to them (say, six separate collections of boxes).
I had an idea the plant would be able to answer questions. I had a couple of ideas about it's personality. I thought it was probably derived from some faerie plant, but not otherwise affiliated with faerie.
I thought there might be a few other rewards as well in other boxes.
I had a few images of how the whole thing would look.
I had the general idea that the players might have to figure out the lies and do a divide-the-boxes-into-two-groups-repeatedly 20-questions binary-chop style. I assumed they might find some other ways to to investigate the inside of the boxes (they had a few abilities that might find things out), and I tried to construct things
For room 2, I had a clear idea of how it would look. I knew that you'd only get a magical effect if you cut yourself with a knife, dropped a drop of blood into a bowl, on one of the altars, using one of each type of object, else nothing would happen. I'd thought through some of the same questions as above. I expected the party would need to solve it like a puzzle, trying multiple combinations, once they got the basic idea, but that they might find another approach.
For both, there's some things I didn't think consciously about, but I implicitly needed to set the rules of the challenge, of finding a macguffin. In both cases, that was partially ambiguous, but it did just happen automatically -- probably because in the labyrinth, whenever I describe the players finding anything other than a tunnel, they usually are primed to want to explore it.
How it went
Several things changed as the players encountered the room, but as I thought, filling in the details worked fine. Describing the box room I wanted to fill out the description of the room a bit, so I put shelves of books round, which seemed appropriate, but then I needed to fill in information in them, and because the players were interested, I changed/added some extra rewards in the boxes which were books, so the players got some of the book excitement it looked like the room was promising them.
The random encounter table told me the PCs would meet one of the NPCs, a minotaur who works to maintain the labyrinth's smooth running. So I had the custodian be killed and the minotaur working on regrowing it. The minotaur took on the role of answering the questions, which helped in some ways (she could offer editorial asides) and hurt in others (it wasn't immediately clear when she was following the rules of the room and when she wasn't).
Several other things went as I expected: The players got the general idea of finding the macguffin, they got the idea of getting zapped and trying to open some boxes but not too many. The players came up with several sensible ways of exploring the contents of the boxes without opening them, and they did, by turning-into-a-woodlouse, and by picking the boxes up. In those cases, I decided it worked on some of the boxes but not all, and rolled to see if they found any of the rewards. They tried a few things which maybe deserved to have worked, but I couldn't let work without obviating the entire challenge.
I made up the "shock" when you opened a wrong box, as 1d6 damage, and that felt about right. I'd specified a similar amount in the other room, and it seemed to work well. Enough to put people off, but not enough to be a danger unless they kept persevering after they were clearly getting too low on health.
I'd forgotten that one of the PCs was resistant to electricity, so after a while they made use of that, which counted as "a helpful approach I hadn't considered" :)
The altar room worked basically how I expected: the PCs got the idea, they tried several combinations, they considered a few false alleys, they got a bit low on health and decided to return later to be on the safe side, but had a good idea what the remaining possibilities to try were.
The box room was more awkward. The PCs constantly had good ideas how to approach the problem, but even after I managed to work in several hints, it hadn't been clear enough that some of the answers were unreliable and they could figure out which with the right approach. They were on the point of burning the boxes down (which would have worked, but done them quite a lot of damage as the zaps would still have happened). But I used the warlock familiar to deliver a big answer in exchange for some devil worship from one of the other players, which basically told them what to do, and was fairly satisfying in-world. So that mostly turned out well, but it did get bogged down for a while which I would ideally have avoided.
What I should have done
I should have explicitly thought was detect magic would show.
I should have had a more specific plan for "pick them up and shake them" as the most obvious course of action.
When I had to have a sensible plan not work, I should have done more to make it feel like a success somehow even if it didn't technically help (like having a bad skill fumble sometimes be a pratfall and sometimes be a dramatic last minute save, and most PCs should be wish-fulfilment-y quite often).
I should have made sure that if the players were completely stuck, there was a sensible way forward: included a way for them to get hints, or made sure there was an answer which could brute force a solution even if it cost them other resources (health, money, time); or set it up so leaving and coming back another time felt more reasonable.
I should have thought what happens if GP 1 had looked at the blood sacrifice room and said "looks evil, not touching it", whether I should have had some more clear way to show them it was ok to experiment (although that wasn't a problem in practice).
Write-up
I don't usually write up in this level of detail because obviously, it takes a lot longer than it did to write the planning! But it's interesting to do sometimes, so I can consciously notice what I was doing explicitly, and what I was implicitly doing in the background.
When I was less experienced, I would have needed to write down a lot of the specific things that here I just kept in my mind. Or if I was writing an adventure for someone else to run, I would have to spell out the extra detail I only verbalised now, because a lot of the good things about the encounter were the general idea of how it should play out, not the specifics.
Although even then, a lot of what went well was because I knew the players and know how to slot appropriate connections in as I went along. A published adventure can't easily do that. The best you could do is list the treasure as what was definitely there, and then offer the GM to fill in whatever felt appropriate, with some specific examples for them to use if they didn't have a better idea. Things like, here's some prompts (for things connected with the room, for things with an interesting backstory, for things that are of particular interest to the players somehow)
In terms of "puzzles" I think the most important thing is to give the players some experience of "trying different approaches and seeing what works", but make sure that the encounter being satisfying doesn't rely on them being able to figure it out logically. And single-answer puzzles are usually bad for this, unless you build an adventure about them with going asking different NPCs what they think the answer might be, etc. In this case, because I wanted a like-Labyrinth-film feel, I went as close to "single answer puzzle" as I could.
I think I learned how to do it better in future. Partly by using puzzles sparingly, but mostly, how to do it well when it comes up. Now I think of it, some of it is sleight of hand: it's giving the *experience* of solving a puzzle, but actually, having a fallback so it matters less whether you SUCCEED or not, as long as you give it a fair try.
Of course, that's for *most* DnD games where the PCs are supposed to be reasonably successful. Sometimes you want to play a gritty deadly tomb of horrors dark souls game, in which case you can go for "they may never solve it and that's ok, it's up to them if they want to push their luck or not" approach.
This was great!
Date: 2019-06-10 10:00 pm (UTC)I agree about it being hard to create good puzzles that don't involve 'pausing the game'; I think you can bend this if you're with a group that love puzzles, but much better if you can integrate it, as you did here. A recent one I did involved deciphering a 'gnomish mechanism' which opened a castle gate, the twist being that the party were in combat at the time, having stormed the gatehouse and sealed themselves in with the guards (who they at face value could overwhelm) while many more guards were trying to break down the doors, giving time-pressure. There were seven wheels/levers and each round a PC could study the mechanism (Int check to find out something), try something (which might give information, but if moving something jammed increased the probability of it breaking) or, indeed, fight the guards, heal people, cast spells and all the usual combat activities.
Re: This was great!
Date: 2019-06-11 02:17 pm (UTC)Definitely, if you can have something interactive at the same time as a combat encounter, that's usually really great, it implicitly makes everyone involved, and everyone invested in finding a solution, and has a natural good-and-bad-consequences the players are already familiar with from other encounters. I don't usually manage it, in fact, but I think it's a good example of what to work towards!