In our case it was really a death by a thousand cuts rather than any one match. Our bot was a heavy defensive wedge coupled with a thick vertical spinner. That may sound like the meta today, but back then the edge in the meta was still solidly in the realm of heavy horizontal spinners like Tombstone, and our design sacrificed offensive capability for the ability to tank massive horizontal blows to the wedge and spinner. It worked very well, we KO'd Son of Whyachi and destroyed Gigabyte's weapon with this strategy. But taking these blows on the chin over and over and over just rattled the bot apart over the tournament. Before our final match we realized there was nothing to it, we had to tear the bot down entirely and reassemble it with the freshest parts we could find. This was a nightmare; the bots aren't that complicated internally, but because of size and weight constraints and for the sake of survivability everything is very compact, interlocked, and secured. It was optimized for fighting, not easy maintainability, and that's what killed us. We were almost burnt out by this point and getting the bot back together again very nearly finished us off. Several times we put the bot together, then tested it, then something wouldn't work, so then we have to diagnose and tear the bot apart again and put it back together again, and then something else wouldn't work, and it goes round and round and round. Our team captain actually gave into despair at one point and wandered off in a fugue; our welder told us to keep working on the bot until she could manage to snap him out of it. We missed our match deadline, which is potentially grounds for disqualification, but we were lucky in that they had room the schedule to push us back by an hour or so. We finally managed to get everything working with mere moments to spare, but remember how I mentioned that our wedge had been tanking hits nonstop till this point? Well, while our wedge was hardened steel, our frame was just (thick) aluminum. More than 20 bolts held the wedge onto the frame, but all the bolt holes had suffered so much trauma from distributing the repeated impacts that in truth the wedge was barely hanging on by this point. With all the other things we had to do, there just wasn't time to manufacture a new frame or find a way to bore and tap new holes. Instead we found some metal inserts designed to let you "re-tap" holes that had been stripped out (I forget what they're called). But our final match was a three-way against two other vertical spinners, and in the opening moments End Game caught us under the chin at full spinup and absolutely blasted our wedge into the stratosphere, obliterating part of the ceiling rig and several lights in the process. It might as well have been held on with bubblegum. We weren't even mad, it was amazing carnage and we were howling with laughter at our now-headless robot.

> What was a design choice the team believed was solid, only to prove a weakness in the Battle Box?

A weakness of vertical spinners (as opposed to vertical drums) is that they're not invertible, so getting flipped can mean game over if you're not ready for it. Our bot was designed so that if we got flipped, our wheels would still have contact so we could ram a wall in an attempt to self-right. Which worked (in both testing and in-match), but while flipped the front of the bot was riding on the weapon itself, and we were concerned about damaging it. So after we arrived at the filming we machined two little aluminum "rabbit ears" and topped them with slippery plastic, which also increased the height of the wedge's lip when inverted thereby making it easier to self-right by ramming. But this was our undoing. In our fight against Gigabyte we successfully destroyed their horizontal spinner, but in the process they rode up our wedge and snapped off exactly one of our rabbit ears. We should have won that match, but then somehow we wound up inverted and the remaining rabbit ear made us asymmetric and caused us to tip to one side, meaning that one wheel didn't have traction, so we were stuck. Heartbreaking.

> How are these bots tested outside the arena? Do the equivalent of “friendlies” exist?

Teams will do their best to test, but these bots are so expensive (relative to our funding, anyway, which is essentially nil) and so labor-intensive to manufacture and assemble that only the most dedicated teams can afford real testing against other bots. For our bot we did organize an impromptu street fight outside of a bar in Boston (there are plenty of teams in the area to scrimmage with, and obviously for the safety of the public all the primary weapons were disabled). Then after we arrived in LA and unpacked the bot we ran it around an empty Target parking lot to make sure the drive wasn't damaged in shipping. For actual testing at the venue, they provide miniature versions of the arena for you to drive your bot around with the weapon working in a safe environment.

I know exactly which bot you're talking about, it's actually one that I remember pretty well from that season! That three-way rumble was wild, although to be honest it was overshadowed by the one right after where Duck!s win was tragically stolen.