The Baroness' Jewels: a guess-who adventure

Yesterday, i proposed a structure and process for creating investigation scenarios. Today, I'll give an example.

1. I begin by deciding on a premise: an Agatha Christie-style mystery where one of the people in a large house have stolen something, because it is easy to visualize and come up with clues.

2. I decide on a five column matrix, because it gives more options than a four-column matrix but is neater than the six-column ditto. Also, I want each criteria to match exactly half of the cases meeting any other criteria, and I realized that my six-clue matrix didn't do that.

3. I come up with criteria, and clues to match.

Clues

- a blonde (edit:brown) hair found at the crime scene
- a size 9 (edit:7) footprint outside the window
- Cecil, David, Gabriella, Hector, Karen, Leonard, Oscar, Patricia have alibi.
- The wooden staircase leading between the floors squeaks and creaks terribly. Hector (with his migraine) heard no-one except the Baroness use it.
- the room was remarkably tidy, as if the burglar know where to look. Only seven people have visited the Baroness in her room.

4. I fill out my matrix.

5. I re-write the entries, from YES and (blank) to a more natural language, and introduce some variation within the categories. For example, I decide that not everyone with an alibi has the same alibi (because then you could just ask any of them which makes for a boring investigation), and that not everyone with different-size shoes or different hair color has the same because it feels artificial. I also come up with some reasons for why people have been in the Baroness room.
Since now the characters have different alibi, I also change the criteria from "No Alibi" to "Alibi".

New alibi clues:

- Cecil, David and Gabrielle were in the foyer playing Gin Rummy, when they heard the Baroness' cry for help
- Hector had migraine due to the weather, and was helped in his room by Karen and Leonard when they heard the call.
- Oscar and Patricia are having a divorce, and were arguing bitterly in their room.

Looking at the matrix, some of the clues had to go in certain columns. The two clues  involving the Baroness - having been in her room, and living on the same floor - necessarily had to include her too. Of course, I could have decided that there were 16 suspects plus the Baroness, but I don't think that works in this type of scenario. Of course the Baroness must also be among the suspects - the jewelry might have been glass, her fortune gone, and now she was desperate for insurance money. That could totally happen in Agatha Christie, so it should be a possibility here too.

Filling out the rest of my matrix, I also realized that some of the criteria felt off. I had originally though that the footprint was size 9, but in the matrix it didn't align well with the distribution of male and female names so I changed it to a more neutral 7. The same was true for the blonde hair. For adults, being blonde is more common among women and the distribution in my matrix didn't reflect that. The matrix would also result in a situation where most people were blonde. So I decided to change the clue so the hair found was brown.

Now, in reality if you selected 16 people at sort-of-random, you would often end up with some strange coincidences like all the women having really big feet or all the men being blonde. In this sense, keeping the original structure would actually have been more realistic. But what matters here isn't if something is realistic, but if it feels realistic. If I was playing with statisticians, I would have kept the idiosyncrasy because it would probably have been realistic to them. But I'm not.

The reason I point this out, is that in any mystery solving you are asking the players to look for patterns and irregularities. And in doing this, they will often use their real-world knowledge of things. Like, for example, that you could hear someone walking in a creaking stair, or that you can find out someone's shoe size by asking them, by looking in their shoes or by measuring it. Therefore, I think it is fair to limit the number of accidental patterns so that your players don't spend the evening ruling out the possibilities that the thief was wearing a wig or using wrong-sized shoes to cast suspicion elsewhere.

Result

Premise

Wealthy guests have gathered at a countryside mansion, when sudden storm cuts off the only road leading there. On Saturday night, someone breaks into the Baroness' room at floor 3 stealing  jewelry, but is interrupted by the Baroness returning to the room and flees out the window.


The PCs are detectives.

Clues

- a brown hair can be found at the crime scene
- a size 7 footprint outside the window
- the room was remarkably tidy, as if the burglar know where to look.
- Cecil, David and Gabrielle were in the foyer playing Gin Rummy, when they heard the Baroness' cry for help
- Hector had migraine due to the weather, and was helped in his room by Karen and Leonard when they heard the call.
- Oscar and Patricia are having a divorce, and were arguing bitterly in their room.
- The wooden staircase leading between the floors squeaks and creaks terribly. Hector (with his migraine) heard no-one except the Baroness use it.

Dramatis personae

Alfred. Had breakfast with the Baroness in her room on Saturday morning. Had room next to the Baroness. Shoe size 7, Brunette. (Guilty)

Baroness. Lived in the room. Shoe size 6, Blonde

Cecil. Helped the Baroness carry her luggage. Room on third floor. Played Gin Rummy in Foyer. Shoe size 7, Blonde

David. Was having an affair with the Baroness. Room further down the hall from Baroness.    Played Gin Rummy in Foyer. Size    9, Brunette

Elliot. Had breakfast with the Baroness in her room on Saturday morning. Size 7,    Blonde

Francis. Had breakfast with the Baroness in her room on Saturday morning. Size 8, Brunette

Gabrielle. Gave the Baroness a hand unpacking. Gin Rummy in Foyer. Size 7, Brunette

Hector. Joined the Baroness for tea in her room. Had migraine, w/Karen and Leonard. Size 10, Black hair

Isobel. Room on third floor, size 7 shoes, Black hair.

Juliette. Had room across from Baroness, size 6, Brunette.

Karen. Room on third floor, Helped Hector when he had migraine,    shoe size 7, Red hair

Leonard. Room next to the Baroness. Helped Hector. Size 8, Brunette.

Malcolm. Size 7, Brunette.

Nathan. Size 9, Dark hair

Oscar. Argued with Patricia. Size 7, Brunette.

Patricia. Argued with Oscar. Size 6, Blonde.

The rest can be improvised during play.

A procedure for creating investigation scenarios, or Guess-who adventures

Of all the campaigns I've played, I hold Masks of Nyarlathothep highest. And ever since I played it, I've trying to recreate its atmosphere - without success. The primary reason for this is time. Masks' basic structure is simple: there is a number of nodes, there are local mini-adventures at each node, there are also events that point to the next node, and players gain access to local and the global events through clues. Part of the brilliance of Masks is its scale. The clues are usually very straight-forward: you find a match box from a bar so you know that's where you should go, but the bar is in another country and you also have ten other clues pointing in other directions. What to do? So while the structure is simple, the fundamental feeling of investigating relies heavily on the sheer quantity of leads, locations and events. To create a similar campaign would simply take me too much time.

But another reason is that I wouldn't really know how to start. Internet, and to some extent the game books themselves, is full of advice for creating locations in general, and dungeons and encounters in particular. Other forms of adventures, barring the story-structured adventure, gets much less attention. This is reasonable, since the encounter and the dungeon are widely used adventure building blocks. But this also presents the strange situation where you are given most advice on the things you are already most familiar with. For investigation adventures, you're left with stuff like the three clue rule and that murderers need motivation. And while decent advice, it isn't really helping me. The combined takeaway is that I need three times as many clues, which triples the workload, and need to write some backstory. In other words, I was told to do more of things I already knew I had to do, instead of a different approach to the problem.

So in case your situation is anything like mine, I want to propose a design for creating investigation-centered adventures based on the kid's game "Guess who?". In this design, player-characters are presented with a number of possibilities, and tasked with selecting the correct one. Every step of the investigation leads them closer to zeroing in on the correct solution by eliminating some possibilities, until only the correct one remains.

Below is a procedure for creating such an adventure.

Guess-who adventures

1. Decide on the premise.

The guess-who investigation can only reduce the number of possibilities, never add. This presents a limit to what can be investigated, since there must be some initial separation between possible and impossible cases. Good premises are:
- something is stolen, and only a select group of people knew its location
- a murder on an isolated island
- the gates to the lost dwarven city can only be opened with the correct sequence of runes, but the wrong sequence will flood it with poison gas and seal it off for another century
- have three attempts to enter the PIN code that will open the terrorist's cell phone, and time is running short
- Professor Carlisle disappeared while visiting five ancient cities, where did he go?
- The Crown of Command must be retrieved from the Lich King's hold, but no-one knows in which of his crypts it is
- The Orcs are gathering in alarming numbers in the forest, and the King don't know where to send his troops to meet them

Deciding on the premise is a big deal, since it has consequences for what types of clues are reasonable and how to find them.

2. Set up the clue matrix.

You can make the number of possibilities as large of small as you like. Here I present the basic matrix for 16 possibilities. This makes it solvable with 4 clues, which I think is reasonable.

Criteria 5 and 6 are not needed, and only included here for the sake of completeness. You can save prep time by cutting them.

If your investigation has a lot more starting complexity, like a four-digit PIN, you need a matrix where many more possibilities are excluded than included in each step.

3. Set elimination criteria

If you look at the matrix, you see 8 YESs in each column. One of these is the correct solution, and the other fall into the same category with regards to some criterion. So what are these criteria? Well, really any characteristic that set some of the cases apart from the rest. By knowing that the correct solution fall into the category sharing this characteristic, we can exclude all cases that don't. Therefore, the criteria or characteristics serve the purpose of ruling out possibilities and thereby excluding some cases.

For a murder or theft relevant criteria could be:
Being blonde, leaving the party early, being out in the rain, being elderly, size 10 shoes, and not having an alibi for 9-10 pm.

For a PIN code or dwarven runes, relevant criteria could be:
the second digit is a 3, being a historical date (puts the first two digits at 00 through 20, assuming western calendar), being a prime number, being a birthday of some relation, the four symbols include a combination of frog, ox, moon & sphere.

For a lost city or invasion, relevant criteria could be:
west of some landmark, by a port, far from any major city, site significant in the some personal vendetta with the Lich King, it begins with an N

Four things are important when setting your criteria
1. That they can reasonably be shared between several cases
2. That they are distinct enough that they can be noted separately
3. That they can be formulated as YES/NO.
4. That they exclude a similar amount of cases.

4. Design your clues

Clues can be found in witness reports, letters, travel logs, e-mails or text messages, footprints, a hair left at the crime scene, and so on. Decide if you want the clues to be obvious or require some further reasoning. An obvious clue would be "I saw two people using the ladder that night". A indirect clue would "Smith is in a wheelchair and Hank is afraid of heights; the crime scene can only be reached by a ladder." People feel smarter when they spot an indirect clue, but they are also trickier to set up.

5. Distribute the clues, as you see fit.

Some final comments

Note that the excluded cases in a column don't have to be similar with regard to the characteristic, they just don't match the criterion. For example, your could have 4 suspects with shoe size 7, 4 with size 8, if the footprints indicate that the burglar was size 10.

The investigation will feel more realistic if the proportion isn't strictly half-half regarding each characteristic. Maybe #2 rules out 10 possibilities, while #3 only rules out 7. However, if you shift the proportions, make sure that no clue becomes totally redundant.

Also, since the case can be solved with any four clues, you only need four clues. Including five or six clues, however, give the players more freedom in their investigation. On the other hand, the stucture doesn't really permit more than six clues if they are all to be equal. If for some reason you want more than six clues, the additional clues need to overlap with your original six. A practical way to do this is to think of criteria that by definition or nature appear together. So the size 10 shoe could also lead us to conclude that the burglar was a man, or tall, or both. Or if the invaders will arrive at to the port city, this could also lead us to conclude that come by boat and that the attack is along the coastline. Strictly speaking, these additional clues are of course redundant, but they can provide the players with more options and make the investigation feel more natural.

House Rules

It's been a while since I posted and I'm struggling with the last pieces of the next adventure. But today is supposedly the last day of the Google+ OSR group, and I felt I wanted to post something to commemorate this occasion. During my time at G+, this group has filled my feed with a constant stream of inspiration and creativity, and led me to discover a wealth of outstanding blogs and creators that I might otherwise have missed.

So, in lieu of a proper post with actual content, I offer my house rules.

The starting idea behind these rules was to provide a sense of weight. This is the reason for using damage reduction instead of AC, and for the size die and ability requirements for equipment. But this reasoning also resulted in two rules that might perhaps be of some interest to others: fatigue, and pressing.

Fatigue basically provides an extended range for critical failures. To not make the book-keeping too burdensome, normal equipment and normal actions do not result in fatigue. Opportunity attacks, wielding heavy weapons - and in my game wearing heavy armor - do. In my game, it's treated as a "mild" critical miss: you become easier to hit and attacks deal more damage to you until your next turn.

Pressing, in turn, is a way of gambling a success on getting another success. After an attack hits, the player can opt to attack again instead of rolling damage. If the second attack is also successful, they score two hits. If not, they lose the first. However, for each time the attack is pressed, fatigue goes up by one. I stole this rule from detect magic where it is more unforgiving, but I scaled it back since I want to keep HP roughly in the same spectrum as ability scores (max 20). If you use standard HP progression, the original rule is probably better.

1p Adventure: House of Thousand Idols

The Rod of Seven Parts for my campaign is no rod, but relics of the Martyr King Severend. So far I've detailed the locations for the torso, the head, and the arm, with the aim of making each location as condensed and "constructed" as I can.

Today's addition is the house of thousand idols. The idea was to design a maze that is simple enough to crack for those so inclined and possible to brute-force in a non-tedious manner. My proposed solution is to use symmetric rooms with many passages and no corridors, and instead make the rooms multi-state. So the trick isn't really "do we find the way", but "do we find the right way".

To emphasize this theme, I put in slow but very deadly enemies that trigger if you don't follow the right way, and a cockatrice that can fly - and so have other movement options than the PCs - and has a petrifying attack that further limits movement.

The adventure will work best if the cockatrice is a beatable but formidable foe so that players prefer avoiding it. If they can easily kill it, the maze feel too static so in that case you might want to introduce some rival adventurers or a mother-cockatrice that arrives in 1d4 turns to maintain some urgency.