[Adam McLeod]

As arguments against using a scoring aid go, I'm fully comfortable with the "coaches should be watching the whole field and not spending time punching figures into a program" line of reasoning. I know a number of excellent coaches who all subscribe to this philosophy, and it makes the most sense to me, for what that's worth.

What I'm not comfortable with is when we start to speak categorically. My feeling is that over the course of the season, a good coach watching the entire situation in every match will win more matches than a good coach tabulating score and making decisions based on that in every match.

Following from this, I think that the best coach is a composite, provided perfect information prior to each match on which coaching method is going to end up being optimal. The assumption I make in this assessment is, of course, that there ARE matches where a loss can be turned into a win by switching from pure field watching to a watching/tabulating combination, so I suppose I have to support that.

Where I think knowing the exact score and optimal next move is important is in the last 20 or 30 seconds of the match, when alliances are going to start to think about their call backs for end-of-match elevation. Implicit in this strategic consideration is an issue with the placement and timing of your last ringer/spoiler. I doubt that even a very good coach can consistently know the exact score well enough to determine the risk/reward tradeoffs with regard to time remaining. For example: do we need to leave our best hanger out in the middle of the field for 5 seconds longer than the safest call back time, given the endgame capabilities of our alliance and our opponents and the score? What is the downside risk associated with this decision? Is there a less risky way to assure victory in the match (e.g. do we need to get a spoiler on the side, or just another ringer right in front of us before getting back)?

There is no reason why a very good coach without an aid can not make a winning decision in these types of situations every time they come up over the season, given the limited number of matches. I highly doubt that this coach can perform on the fly the calculations necessary to come to an optimal solution as many times out of 100 as can a computer program with the right degree of sophistication. This theoretically reduces the aid-less coach's winning percentage over the long run.

On the other side of the coin, I think we can all agree that there are match situations where looking down at a screen can cause a coach to miss things vital to success, so I'm not going to spend any time on that.

Another thing to consider is that one of my assumptions is a "good" coach. As the quality of a coach diminishes, I think we eventually reach a point where that person would be better off using a program to determine the next best move to tell the drivers while completely ignoring ''big picture'' situation. My gut feeling is that if we could graph it out, that point would come much earlier along the coach quality decline curve than we would think. To use an example from this thread, Karthik: I know you're confident that your coach will be able to consitently make the best or at least a satisficing decision to win. I wonder: without Derek (assuming you can't fill in), how confident would you feel?

(Forgive me a small digression, but sometimes I wish that coaches would switch teams more often so that we could more accurately determine the impact of above or below average coaching on bottom line win totals. I'm thinking about baseball, where many analysts have determined that a "good" manager probably adds around 3 to 6 wins per year by strategy alone - not counting player development and management - which seems like an awfully low number when we consider the popular image of the manager as genius. In football it seems like the right head coach can turn a team around completely, and I suspect that this is true, because football, being an emotional game, requires more ''inspiration'' from the coach, and because there are so many more strategic options than in baseball. I suspect that coaching in FIRST would fall somewhere between those two sports, but evaluation suffers from the same pitfalls as in sports, in which coaches and managers with good players look good, and the ones with bad players look like idiots. This pitfall is clear when we think about the best, big name coaches in FIRST: they're all from teams who build great robots.)

All this being said, my conclusion is the same as many others, that a good coach is likely better served by watching the field than by spending any time inputting data, with the qualifier that a mediocre coach might want to investigate his or her computer based options. As for me (scouting and strategy from the stands), I would consider tracking these real time statistics so that I can evaluate our match strategy and report it to the drivers and coach afterwards, but I'm reasonably certain that I won't be using one either. After all, you can probably tell I'm into objectivity, but if we can't rely on experience and the human mind from time to time, then we're in trouble, aren't we?