Michigan State: Challenge or Setback

I was the one who lost us the regional. Not the team, me, and I have the ledger to prove it. ¹As lead programmer for our robotics team, I wrote the autonomous routine that was supposed to score four rings in the opening fifteen seconds. In practice it had worked ninety times in a row. On the competition floor, in front of forty schools, our robot drove a confident, beautiful path directly off the edge of the field. I had hard-coded the starting position. ²Our practice field was a quarter-inch shorter than regulation, a difference I had never measured and never imagined mattered. My code counted wheel rotations and trusted that the world would match my assumptions. The world did not. We dropped from second seed to nineteenth, and I watched two seniors, kids who had given three years to this, realize their last season was over because of three numbers I had typed and never questioned. I did not handle it well at first. I told everyone the field was wrong, which was true and also cowardly, because the field is always going to be a little wrong. ³That was the real failure, not the line of code but the design philosophy behind it: I had built something that demanded the world be perfect instead of building something that could survive the world being normal. So I rewrote how I think. ⁴Over the next month I taught myself to use the robot's distance sensors, so it could find the wall instead of assuming where the wall was. I added a calibration step that ran every single match, measuring the field as it actually existed that day. I started writing what my mentor calls defensive code, code that expects to be lied to. More than that, I made our team adopt a rule I now apply everywhere: before you trust a number, ask what happens when it is wrong. We qualified for state that year, and the autonomous routine scored every round. ⁵But the part I am proudest of is quieter. When a freshman's code failed at our next event, I did not let him say the field was wrong. I sat down with him, pulled up the sensor logs, and we found the assumption he had buried, the same kind I once buried. He fixed it in twenty minutes. Watching him made me understand that the lesson had actually taken, because I could give it away. I still keep that ledger from the regional, the one where I wrote down every match score in pen. ⁶The nineteenth-place line is right there, and I have never crossed it out. It is the cheapest insurance I own. Every time I am about to trust a number because it has worked ninety times before, I think of a robot driving proudly off the edge of the world, and I go measure the field. The setback did not make me more careful by accident. It taught me that confidence and verification are not the same thing, and that the second one is the only one worth keeping.