Lessons from the Skulldug! ARG

My partner and I launched our board game Skulldug! earlier this year. It was our first tabletop project to hit commercial release, and it’s done about as well as could be expected. We received a smattering of generally positive reviews, we have a decent score on BoardGameGeek.com, and over a thousand households worldwide now have a copy of our game in them. There are things I’d change if I could, of course, but all in all, I’d call that a success.

Like most indie tabletop games, Skulldug! was funded on Kickstarter. We ran the campaign back in May 2015, and at the time we estimated that the game would be released by January. We actually managed to stick to that date (although a week’s delay in a Chicago warehouse nearly fouled that up), and we were posting monthly updates to keep our backers engaged throughout the development process. Still, eight months is a long time to wait for a game, and I wanted to give our backers something to thank them for their patience. Something special and different, something that would build hype for the game in the weeks before its release while rewarding the backers for their continued support.

I decided to make them an ARG.

Lesson 1: You probably shouldn’t make an ARG

If you’re anything like me, every time a new alternate reality game (ARG) appears, confounding the internet and teasing all kinds of dark and thrilling secrets with inscrutable puzzles, you think about how fun it would be to design one yourself. Don’t get me wrong, it is a lot of fun. I had a great time designing puzzles, creating props, and leaving hints for my players to discover. Ultimately, though, I can’t call the ARG a success — it took a lot of time and effort that could have been better spent elsewhere, only a small number of players engaged with it, and ultimately it didn’t manage to build any hype for Skulldug!.

However, if you really are like me, none of that will dissuade you from trying to create your own ARG. In that case, I’d like to share a few lessons that I took away from running the Skulldug! ARG to help ensure that you see more success than we did.


Skulldug! is a competitive pulp exploration game — think Indiana Jones or Spelunky — and we wanted our ARG to reflect some of the drama inherent in the genre. I decided to frame it as a treasure hunt, with the players racing to track down a missing explorer who had stolen an ancient artifact from the villainous Baron. We announced the start of the ARG with a simple footnote at the bottom of one of our monthly Kickstarter updates. A fan had reached out to us for help looking for their missing friend, we said, but the only clue they had to go on was a list of URLs for videos hosted on Vimeo.com.

These videos were password protected, which presented the first goal of the ARG. To unlock the videos, the players had to decipher a series of coded messages sent in hand-typed letters to some of our higher-tier backers. Those messages each contained the name of an actor and a timestamp, which they could use to locate quotes from famous pulp adventure movies.

Once the players unlocked the videos by using the quotes as passwords, they discovered a series of choppy clips of the missing explorer staggering through a forest to his final resting place, each of which had a strange, high-pitched tone playing on top of it. By combining the pitch of each video’s tone, punctuation marks hidden within the image frames, and the first letter of the clip’s title, players were able to create a set of GPS coordinates. These coordinates marked the location of the explorer’s grave and, more importantly, the treasure he left there.

Simple, right?

By ARG standards, it kind of was — there were only three real puzzles (the letter cipher, the movie quotes, and the GPS coordinates), none of which required the kind of hyper-specialized technical skills that larger ARGs sometimes ask of their audience. This was intentional, since I wanted them to be solvable by the small community of board gamers that had supported our Kickstarter, whereas many larger ARGs target an audience of thousands of players or more.

This was my first mistake.

Lesson 1 (for real): Reach a large audience

ARGs aren’t for everyone. If you’re creating one to promote another project, it’s probable that the members of your existing audience are not all going to be comfortable or interested in participating in your ARG. We only had 10 or so players over the course of our ARG’s 6 week run, about 1% of our total backers. Worse, we had many players who would drop in briefly, comment once or twice, and then disappear. Since these players’ participation was spread across the entire run of the ARG, there were frequently only one or two players active at any one time. This lack of constant participation meant that progress was always slow, which made existing players lose interest and discouraged newcomers from taking part.

If you’re going to run an ARG, you need to make sure that you’re able to reach a critical mass of players that can create enough activity to sustain their own interest. Remember that only a small portion of your audience is likely to be interested in your ARG, and so you may need to open up participation to players outside of your existing fanbase. ARG-enthusiast communities such as Unfiction.com may be good places to recruit new players, but be aware that dedicated ARG players are likely to solve your puzzles much faster than your regular audience.

Lesson 2: Don’t limit participation, especially at the start

In addition to having a small overall pool of potential participants, I made the mistake of cutting those numbers even further by restricting the initial participation to a tiny subset of our total audience. Although the first puzzle (the locked video links) was technically available for everyone to solve, only the 60 or so backers who received handtyped letters had the codes required to unlock them. Worse, those players couldn’t even complete the puzzle on their own — there were five different letters, and the players had to share all of them in order to complete the key that decoded each letter’s message.

On paper, this was a great idea. The players who received the letters would feel special and compelled to participate (who isn’t excited about being sent a personal secret message?), and having to share the codes to proceed would foster a sense of collaboration and encourage the players to communicate throughout the remainder of the ARG.

In practice, though, only a couple players were interested in the ARG enough to share their letters of their own accord, and I was forced to reach out to other players directly in order to get the whole set shared. Meanwhile, all the players who didn’t have letters of their own were left with nothing to do, which caused them to lose interest at a higher rate than they might have otherwise.

While the idea of limiting information to certain players is a good one, I would recommend not using it until later in your game, when you’ve had time to build up a certain level of participation and interest from your playerbase. Additionally, you should make sure to provide that information to players you know are already active in the game, rather than giving it to random people in the hopes that they’ll become motivated to participate.

Lesson 3: Plan your difficulty curve

Overall, I’m pretty happy with the difficulty of the puzzles in the Skulldug! ARG. They were relatively tricky and required a varied skillset, but were not as hyper-specific as those in most larger ARGs. I knew that much of my audience would not have participated in an ARG before, and so I was essentially trying to create an entry-level version of those larger, better known games. The goal was to create puzzles that would tax a newcomer while not being completely insurmountable when a group of players put their minds to it.

Although none of the puzzles were significantly harder than that target, none of them were significantly easier, either. The difficulty of the puzzles was fairly constant throughout, and there was no real difficulty curve to speak of. Instead, players found themselves immediately slammed against a brick wall as they struggled to decode the cryptic letters (which, as mentioned above, was technically impossible until all the codes were shared more than 2 weeks later). Rather than providing an approachable introduction to ARGs, this made our players feel like ARGs are too hard for them and discouraged further participation.

Although generally you want ARG puzzles to be quite hard, since they are designed to take a whole community of people days or more to solve, I recommend starting your ARG with several simpler puzzles that a single person could solve in a sitting or two. While these won’t present a challenge for the community as a whole, they provide a comfortable on-ramp for new players who will hopefully then be more motivated to continue on and participate in the harder puzzles with the more experienced players.

Lesson 4: Avoid ambiguous puzzle solutions

It's vitally important that all of your puzzles have a built-in mechanism to confirm their own solutions. The locked video files are a good example of this. While it took our players a long time (and some help from our in-universe Twitter account) to discover how to locate and format their passwords, they were able to confirm with 100% accuracy when they had it correct — either the video unlocked or it didn’t.

Ambiguous solutions are dangerous things. If your players are solving a puzzle that will give them critical information for use elsewhere, then getting that information wrong can send them on a wild goose chase that may frustrate them and lose you players. Worse, it could encourage your players to perform unintended and potentially dangerous behaviors in the real world.

We ran into this with the final puzzle of the Skulldug! ARG, which tasked the players to reconstruct the final GPS coordinates from clues embedded in the locked video files. With no built-in method to confirm whether or not the players had come to the right solution, it was entirely possible that an overeager player might head out into the wilderness and get themselves lost based on a misinterpreted puzzle solution.

Our players actually did settle on an inaccurate set of coordinates at one point. They seemed to match all the clues, and the location looked reasonable enough when checked on Google Maps. We had to directly intervene to correct them, pointing out several places where they had misinterpreted the clues to set them back on the right path. In the end, we were able to use the Twitter account to confirm when the final solution had been reached. Still, since checking the Twitter was not part of the required flow for progressing, this was a patch job at best.

We could have avoided all that trouble if we had just made the puzzle have an unambiguous solution. For instance, instead of hiding the GPS coordinates themselves in the video, we could have hidden pieces of a URL that would display the coordinates (or even a complete map) when visited. These kind of clear-cut solutions allow your players to more easily gauge their own progress and help avoid the potential safety issues that can occur when players go off the reservation (sometimes literally).

In Conclusion

Although I opened this post by saying that you should not make your own ARG, the truth is that the Skulldug! ARG was a lot of fun to create. What game designer doesn’t want to send their players on a code-cracking, truth-seeking, real-world treasure hunt? However, because of the mistakes I outlined above, the process of actually running the game was extremely stressful. I fretted constantly that people weren’t making progress fast enough and that the few players I did have would lose interest, leaving my work forever unsolved. If nothing else, I really didn’t want to drive 3 hours into the Washington wilderness to retrieve the treasure I’d hidden out there.

I hope that the lessons I’ve outlined above will help your own ARG efforts be more successful than mine were. If you’re interested in reading more about the art of ARG design, I highly recommend Adam Foster’s excellent article on running the (much larger) Portal 2 ARG. And if you have any questions, I'm always up to chat design on Twitter!

Until next time, fellow puzzle fans!

Jon Gill

Jon Gill is a Seattle-based ex-pat with a love for elegant game design and dark humor. By day he codes augmented reality games; by night he designs board games as half of Ruddy Games.