In sixth grade, I watched a show on PBS about MIT’s 2.007 robot competition, and I knew that was what I was going to try do in college.

I wanted the challenge of solving some problem, building a machine to do it, and compete using both my wits and my engineering skill. And in 1997, I got to take the class and live out a childhood dream.

I love building things with my hands. I don’t do it as often as I’d like these days, but I find that that’s when I feel like my mind and body are fully engaged. Part of the challenge is figuring out the rules – what to build, and then executing in real time during the competition.

The year I took the class, the design of the competition… was terrible.

The goal was to get balls into your bin. They fell from a “waterfall” mechanism that triggered a few moments after the competition. There was a ramp from the starting position near the waterfall and your bin down to a central arena where the balls would land.

The problem with the rules is that there was one very clear, very obvious solution. An arm that would extend and redirect the balls directly into your bin was the clear answer, and the winner would simply be the person who built the fastest, most robust arm that couldn’t be deflected out of position.

That was it. The best strategy by far required zero human interaction – in fact, human interaction would make it *worse*. If you wanted to build the most robust arm, that’d take all your material and space. You’d build the strongest, fastest arm you could, and hope that you could beat everyone else to the punch. If someone built a stronger, faster arm, you were screwed, but here’s the thing – if you built *literally anything else* you were even more screwed.

I didn’t build an arm.

I decided that even though the rules had a clear solution, this isn’t what I wanted out of the competition. I wanted to build an RC robot that relied on my control, that gave me options and flexibility, where I could construct a strategy on the fly and do interesting things. So I built a little bulldozer thing with skids and wheels that could Hoover up balls into a front cage, dump them into a hopper in the back, and then the hopper could extend up from the central arena and dump balls into the bin. I could theoretically use the extending hopper to also block anyone else from dumping balls into the bin. It was really maneuverable and quite fast. I practiced any strategy I could think of over and over again.

But I knew that if I went up against even a marginally competent extending arm, I’d lose.

I don’t have strong memories of the competition or its results. I think the winner in the end was an extending arm. As for me, despite hours and hours of practice, the thing that undid me in the end was a small screw protruded from the bottom of the machine, and in what felt like one-in-a-zillion odds, when I turned off the ramp into the central arena one round, that screw caught on the ramp and “beached” the machine.

I was proud of that machine, though. Aside from the protruding screw, a mistake that was easily fixable with iteration, it was a fairly robust, well-built, flexible machine. It did what I’d hoped it would do, and I got to play the competition the way I wanted to.

And I think that there are really strong parallels to my professional career.

When I look at mobile gaming, there are lots of people who “build the arm”. They look at the current winning strategy, and think they can build the best arm.

I *hate* building arms. I’ve never done it. I’ve never believed that the next hit looks exactly like an iteration of the last hit. I’ve never wanted to work on solved problems. My focus has always been on making things that are flexible, and “broad” re: strategic options, and have attempted to wield that breadth in interesting ways.

And it’s weird – until this afternoon, I never realized *how* strong the parallel was between my experience with 2.007 and the rest of my career.

But I think that “strategic weirdness” is why I love working with a broad range of multidisciplinary people. It’s why I love working on new things. It’s why I hate the kind of “iterate until optimal” business model of highly derivative games.

Back at a job long ago, we made a bingo game, under some amount of pressure from the company that’d acquired our team. The obvious thing they wanted us to do was ape the leading bingo game, and they’d market it aggressively and hope to do something. I kept kicking the can down the road, until one day we figured out how to do something interesting with powerups in bingo, where teams could collectively bring powerups to the game, and trigger them in fun, synergistic ways.

It was a system that created a huge potential for interesting gameplay, because it wasn’t just about *your* powerups and what they’d do for your cards, it was what you could do for the *other* players on your team. It made bingo a wild cooperative team game, and the results were delightful.

The game didn’t succeed in the end. I believe it still could have, given the “possibility space” of the system we’d created. The company basically said, “Welp, dislodging players from (whatever game was winning the bingo category at the time) is too expensive.” And while I think we’d have gotten to a point where we’d have a better game and those costs would have come down, they decided to pull the plug (there’s a lot I’d say here about judgment and necessary courage, but…)

But to me, the point is that fighting the winning bingo game with a clone of the winning bingo game was a sure fire recipe for disaster. It’d have been trying to “build a better arm” when the arm you’re fighting has 100x your budget, 100x your size & weight, and is *already deployed*.

We didn’t have an option to build a better arm. We could only build something new. And I think there are some situations and companies that are foundationally only about “building arms”. They’re not for me, and never will be.

Work is many things, and depending on circumstance, you can’t always think this way. What I have is a luxury. But work, to me, is about the joy of creation, and building things with other people. The most joy comes not from building arms, but from building new weird shit with strategic breadth and the possibility of wielding that breadth in shocking and interesting ways.

And it’s weird that I’d already known that back in 1997 when building a robot for a class.