Students and mentors see these things differently, as a rule. I've been at this for five seasons now, and I can honestly say that I rarely see a robot approach that wasn't considered and discussed during game analysis and prototyping. I had a terrific experience at the Vancouver VRC tournament yesterday, where one robot went 6-0 in Qualifications, followed by at least 11 robots with 4-2 records.
I had a chance to closely study VRC 721, which was the 6-0 bot and the winner of the tournaments "Excellence" award. The special thing about their robot is that there was nothing special about it. It was a good, solid, state-of-the-art Vex robot with clean engineering and smart features, but any team could have built it, and it looked more-or-less like 20 other robots at the competition. What set them apart was driving and execution. The details on their robot were clean and well-done, and their driving was flawless all day. They had a perfect strategy, and drove it just right. Seeing their robot and copying would not have resulted in a champion, but emulating their teamwork and strategy might.
Most of the time when a team thinks they have some cool engineering feature or unbeatable robot, they are wrong. When they have a good robot and a perfectly complementary strategy, they are far more likely to win. I sometimes think the best predictor of a successful robot is the number of practice matches the drive team got to run on a full field before the tournament.
For what it's worth, I think 721 won using a great early-season strategy. We think the winners at the world championship will look different and be following a different strategy. Given 721's record, I suspect they'll build that robot and be driving the new way. So will our teams.
