Are there things that can be improved about the game, FIRST infrastructure, Kit-Of-Parts, documentation, etc.? Sure, there will always be improvements to be made. Even when the system is "really good" it can always be tweaked a little to make it even better.* But in contrast to last year, I think that FIRST has made a quantum leap forward in terms of quality, consistency, and support. We don't have the 40%+ error rates in the packing of the KOP. The number of real rule changes (vs. rule interpretations) after kick-off is down dramatically. I see major improvements and additions to the KOP (transmissions, robot frames, CMUcam, more software, etc.).
There have been some mistakes this year. Some of them have been made by FIRST, but many of them have been beyond FIRST's control. For example, the FP motor issue was a vendor supply problem. FIRST (and parts of Fischer-Price) thought they had the same motor/gearbox as last year. It was only after the teams really started to dig into the performance of the motors that the problem was discovered. FIRST responded, got in touch with Fischer-Price, and they both worked out a solution as quickly as they could to correct the problem as best as possible. Kudos to both Fischer-Price and FIRST for jumping on this (and the solution now leaves us with an even larger set of design possibilities than just replacing all the 6v motors with 12v motors!). I consider this a major improvement over some past years when an issue was discovered late in the build season and teams were basically told to "deal with it."
Where FIRST has slipped up, I know that they have been taking notes to make sure that these are addressed after the season and do not become repeated problems next year. They will probablly address most of them. Some will slip through the cracks, and they will also discover new mistakes to make next year. That is to be expected (with ANY organization). But overall, I think that the number and severity of mistakes is decreasing and the level of support is on an improving trend.
-dave
* (anyone get the analogy to the classic engineering problem of knowing when "better" becomes the enemy of "good enough"?
)