3 min read

Guess the Pattern

I find myself returning again and again to the question of how to find the sweet spot between chaos and stability, that point where a game is just hard enough to be fun but not so hard it seems futile. (I talked about that sweet spot in my last newsletter. If you missed it, you can find it online.)

In 2000 - 2012, I did a lot of training. As part of that work, I created a bunch of games and simulations for my classes. So last week I decided to seek inspiration in the game design work I did over a decade ago. Which games worked? Which didn't? Could I gain new insight by examining them through the lens of chaos and stability?

Most of the games I created were small, like the little five minute cooperative board game titled “Swamp” about factors that increase or decrease risk. I created another board game—Shortcut—to explore technical debt. I was delighted when SEI adapted it into a game titled Hard Choices.

Other games were big. WordCount is a full day in-person simulation where participants take a dysfunctional siloed organization and make it agile over the course of a single day. It was both my biggest hit and worst flop. I ran the simulation over 150 times, and each run brought new surprises. Like real world agile transformations, the outcome at the end was never quite what anyone expected at the beginning and sometimes went painfully sideways. (You can read a handful of WordCount stories on my blog.)

As I sifted through my archive of a decade worth of exercises, one in particular stood out: “Guess the Pattern.” Loosely based on the classic board game Mastermind, you are trying to guess a pattern of five colors. The twist is that you choose the level of feedback you receive on your guesses.

At the lowest level of feedback, the game just tells you “right” or “wrong.” Like bug reports that just say “doesn't work,” the information is useless and the odds of being right without more concrete feedback are quite low. At the medium level of feedback you receive indirect clues telling you things like 4 of the 5 colors you chose are in the pattern, but you have to really work to get the answer. At the maximum level, you get feedback on each individual dot in your guess.

The point of the game was to seed a classroom discussion about the qualities that make feedback in software development (such as test results and bug reports) more or less useful. I ran the game in probably a few dozen classes. Usually the conversation went exactly where I needed it to go to serve the needs of the curriculum.

Occasionally, however, the group would have so much fun at the medium level of feedback they were reluctant to move on to maximum. At the time I felt that was evidence that smart people love puzzles so much that they sometimes make their work more challenging than it needs to be. Or as one executive once observed to me: “The problem with hiring smart people is that they create problems so big only they can solve them.”

Now that I’m more aware of the chaos / stability continuum, I wonder: did the medium feedback level represent that elusive sweet spot?

Fortunately this was the one exercise I had created in code rather than being a physical exercise involving paper and props. How hard could it be to put it up on the web so I could make it available to others to play? Alas, it was implemented as an HTML page with 2006 era JavaScript. Time to bring the one page app into the world of modern web development.

Like many a modernization project, the scope of work turned out to be bigger than I expected. What I thought would take a few hours became maybe 40 hours of effort. Along the way I refreshed my Rails skills, learned Tailwind CSS, and delighted in new JavaScript ES6 syntax. It may not have been how I expected to spend the week, but none of that learning will go to waste. It will all come in handy when it’s time to skin the software development simulation.

If you’d like to play the game, it’s available here. Worth noting: in addition to revamping the user experience, I added instrumentation so I could analyze patterns of play. If anything interesting turns up I’ll share my analysis and insights in a future newsletter.

If you’d like to see the before and after, I made a video about it.

Next up: apply what I learned from "Guess the Pattern" to find a better balance of chaos and stability in the software development simulation.

Stay Curious,