[Tristan Lall]

I think there's value in letting a limited number of teams attend, despite their poor gameplay performance. Some of them are the winners of major judged awards—and I don't think these slots are particularly contentious. As for the ones who lack other redeeming characteristics and frankly drag down the level of competition (some qualify via the waitlist, while others are the weakest links in regional-winning/wildcard finalist alliances), as long as there aren't too many of them, I don't really mind.

One reason is the opportunity for spreading the inspiration around. I think others have amply argued that case. I think it's clearly plausible enough that FIRST should try to quantify its value and take that into account for the foreseeable future.

There's another reason, which is perhaps a bit more controversial, because it goes to the heart of what we expect a championship to be. At every other FIRST event, save perhaps the Michigan championship, there are several of these less successful teams. Through what might charitably be called entropy (less charitably: bumbling failure), they introduce uncertainty into the outcome. Uncertainty in appropriate measure is what distinguishes strategic, replayable games like poker or Magic from dreary algorithmic recitations like Sudoku or chess.¹ I support the proposition that FRC is made better by the good teams having to overcome the obstacle of dealing with the bad ones.

It's like NASCAR vs. Formula 1: F1 is clearly the superior form of racing from a technical and strategic perspective, but over the last 20 or so years, has lost much of the uncertainty that comes from outlandish tactics and mechanical failures. By contrast, people watch NASCAR because of the crashes and the passing—things which are only made possible by the fact that the cars are bloated and archaic, and the tendency of the subset of drivers who are less skilled to make stupid mistakes that rearrange the running order by suddenly eliminating whole swaths of contenders. That sort of uncertainty makes it a little more rational to root for the underdog, and it also gives symbolic meaning to the competition—the winner is triumphing over both the competition and fate itself. A competition structure that better captures that balance is a major reason why the 24 Hours of Le Mans is better than either one.

In FRC, the moderate likelihood that entropy will strike down any given opponent makes the competition more fun (in aggregate) than if the best robot is always going to be the clear winner. It gives hope to the ones who fall short of greatness, and keeps the great ones from getting complacent about their odds of victory.

Make no mistake, I enjoy seeing robots dominate on technical and strategic merits (e.g. 47 in 2000, 71 in 2001 and 2002, 111 in 2003, etc.). I just don't think that their ability to rightfully succeed is diminished by the presence of a few lower-calibre robots.² The odds strongly favour the idea that at least two of the FRC champions will have earned that result based on a clear history of technological and strategic achievement, and that all of the champions will have demonstrated those characteristics throughout the elimination rounds.

In the end, the Championship is mostly about showing off the best robots and best strategies. But you don't need to include the almost-as-good robots to prove that these are the best. Also, by allowing teams that exemplify other FIRST values to attend, along with their bad robots, FIRST provides fuel to power its engine of entropy. Conveniently, those teams with bad robots can still have a valuable experience by being in the presence of so many good robots, and by participating in the other activities that the FRC Championship offers them (from the conference to the socializing). This comes at the expense of the teams who have robots that are almost good enough, but now that FRC is as big as it is, there will always be teams left out—admitting them only shifts that painful burden to another team that didn't quite make it.

¹ You don't have to be particularly mentally dexterous to have a clear idea of how chess can be won—and this high degree of certainty combined with the reality of inadequate computational resources sucks most of the fun out of it. In fact, the real challenge is in solving it efficiently, not in actually playing the game.
² Provided that there are enough matches to adequately rank the divisions. Too few matches, and the competition suffers from outlier effects—which are distinct from entropy effects and are less desirable because they represent uncertainty due to a structural limitation of the event, rather than uncertainty due to surprises encountered during gameplay.