Stranger Manor: A Social Escape Room Game

Stranger Manor was a Stranger Things-themed escape room game which I made for Haunted Manor 2016. Haunted Manor is an exclusive, ticketed Halloween event in Los Angeles that a host of creative individuals puts on yearly; this time, the property was transformed into the Byers’ house (the interior and bars) and the Upside Down (the grounds and dance floor). 


We capped the guest list at 200 to prevent crowding and allow for flow throughout the evening. Since going to a party is different than choosing to attend an escape room, and selects for different people, my job was to create a game that wouldn’t alienate casual players and which didn’t rely on knowing the tropes and history of puzzle hunts.

The design goals of Stranger Manor were several:

  • create a clues-based puzzle game in which guests could wander around the house and grounds;
  • not disrupt the general flow of party activities, including acquiring drinks, waiting for bathrooms, dancing, etc.;
  • allow interested players to solve as much or as little as they’d like and shift their priorities as new events and people came up;
  • get people moving around the lot and exploring the amazing build; have them share secrets and ideas.

The stated aim of the game was to “find the clues, solve the puzzles, and escape the Upside Down”. I went with a simple tag because it’s less of an investment to potential players. You’re already here in the Upside Down; there are already clues surrounding you; you might as well give it a go.[1] To escape, players had to solve the puzzles to find 5 digits, go to the “gate of Hawkins Lab” at the top of the property, and try the combination with the 5 digit lock.

Guests coming up the stairs were greeted with this general outline of rules.

I placed clues throughout the property, both inside and outside the house, at intervals I thought would be unobtrusive to people who didn’t want to play. The “tutorial level” clue was placed within eye line of the door and next to signage directing guests to extra bathrooms, so people would see it immediately and get used to what a “clue” was.


There were 5 proper clues and 2 hints for particular puzzles (the tutorial puzzle, to get people started, and the social puzzle).

This clue was linked to an Arduino that was programmed to spell out a message with a string of addressable LEDs that were modified to look like Christmas bulbs.
In action! The message here spells out “EIGHT HOURS”.

The clues were keyed to a cipher I made which I placed on the bar at the beginning of the game, which ensured that even people who arrived first and were going to be competitive about it couldn’t actually start solving the cipher until 9:30 when the game went live.

Yes, that’s Linear B. Yes, Linear B is syllabic. Yes, I’m evil.

Most clues asked people to identify an object on which a symbol or number was written, and do a bit of basic math to determine the answer to that digit of the lock. (“Make your guests do math while partying,” I thought to myself. “That’s so evil it might come right back to fun.”)

He tried to help.

My favorite puzzle was the social puzzle: I hung a clue that when deciphered read “red – green + blue” for digit 4; as you can see above, a triangle was drawn around “red”, an oval around “green”, and a rectangle around “blue”. Every guest received a drinking glass from the doorman when he checked their names off the list. At the bottom of every glass was a colored number inside of a shape.


Because each person could only receive one stein, my idea was that people would have to ask other people about what was on the bottom of their glass, thus offering an excuse for conversation if you saw someone interesting, and drawing casually interested players in.

Once players had all 5 numbers, they could come to the top of the hill, enter the combination, and if they were successful they would come face to face with the final puzzle. There was a sheet marked “survivors” for anyone who had solved the game, and prizes for the first five people.


A fairly optimistic number.

Prize winners were to come find me, dressed as Nancy; the first winner would be taken up to the stage and would then, using their impressive brain, harness their intelligence to think the Demogorgon to death. Because, as we all know, the overlap between “people who like solving puzzle games” and “people who enjoy public improv” has a ton of overlap. (See the part where I’m evil.)

So that was the plan. How did it actually go?

Well, essentially, “not exactly as planned” and “incredibly”.

First, I wildly, wildly underestimated the number of people who were going to try this game. The conservative estimate we have is that 150 people were playing at peak. That’s the conservative estimate. I theoretically planned for that in the sense that I printed 200 ciphers, just in case; but I did not expect what 150 players would look like. The dance floor cleared. I spent a solid half hour being asked “where can I get a cipher?” and smiling enigmatically when people asked me if something was a clue. People were ignoring friends they hadn’t seen in years because they “had to go look for clues”. It was unbelievable.

I also miscalculated how people would play: I assumed players would be solitary or a couple, or perhaps a duo or trio of friends. Instead, people played in packs. I think we only went through 40 ciphers because groups were huge; 5-8 people seemed about the norm.

Second, the lock to the gate broke after the first players won. This would not have been terrible if there had not been two teams waiting behind them to try their hand. The game ground to a halt while I frantically tried to reset the combination and get it to close. I ultimately had to have the waiting teams whisper their guesses to me and then I stood in for the lock.

This meant that when a clue malfunctioned later, I wasn’t in a position to notice. When someone brought it to my attention, I couldn’t immediately fix it. Eventually a friend offered to step in to serve as lock while I fixed the clue and got a much-needed drink.

My sense is that more people would have won if the clue had been fixed earlier, but some people did win. And much more importantly, nearly everyone had fun. I gave up being the lock about half an hour after the dance party started in earnest and was promptly swarmed by people telling me how much they’d enjoyed themselves. I’d meet new people and introduce myself, and I’d invariably get something like “oh, you made the game! I got a couple of the clues!” with great excitement and pride.

Everything in the backyard was bathed in the strange blue light of the Upside Down.

Stranger Manor by the numbers:

  • Projected budget: $150
  • Printing: $70
  • Prizes: $60 (some were donated)
  • Enclosure box: $40
  • Number of players (casual and hardcore): ~150-170
  • Hours worked (design, build, playtesting): 50
  • Burns from the heat gun: 2
  • Scars: 1

So on time, on budget (…mostly), a somewhat buggy game which a ton of people played and still had a massive amount of fun: I think I created the escape room version of Pokemon Go.

This was an amazing experience and I couldn’t have done this without the design team who transformed the lot. You guys are rockstars. Special thanks to Brendan Patrick Hennessy for early clue-vetting and Nick Fortugno for advice on player flow. Extra special thanks to Astrid Dalmady for real-time playtesting. Thanks also to Emily Short, Andrew Plotkin, Bruno Dias, Katherine Morayati, and furkle for advice and encouragement. Photo credits to me, Alex Klimavich, Amanda Powers, and Elly Glavich. Apologies if I’ve left anyone off the list who should be here.

[1] This idea is directly taken from Nick Fortugno’s 2016 Indiecade talk.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s