Re: Scouting, is it important?
 

I did this for myself, actually. I was surprised how much I learned from watching the old videos, despite not only knowing the games and outcomes, but having driven in half a dozen competitions for each one. You see things differently (not just in the literal sense). If not to determine what you looking for, at least to determine how to look for and interpret it.
We've never really been in a position to train multiple excited scouts at once, but it seems like it'd be helpful from that perspective as well. Not just for familiarity, but to spark discussions on exactly what you're looking at and what you're getting from it. ...hmm, maybe we should utilize this more widely.
So you do mean mathematically. That's what I thought, good idea.


EDIT: hey, 100! :)

Re: Scouting, is it important?
 

In all honesty, our scouting would have been much stronger had we placed some focus on the other aspects you mentioned. We didn't, because two people cannot keep tabs on very much :o However, my scouting partner and I would jot comments: the most common were 'good defense' and 'long kick'. Those are much more subjective of course, and harder to calibrate with a larger and less experienced scouting team: although I suppose defense always tends to be a harder quality to quantify [quality to quantify? Did I just say that?]. A way we made up for not being able to keep records on which teams scored from which zone, was just to ask our team mates what they were good at. Teams tend to be more articulate when asked what they do best, rather than just 'can you score well'.

Now that I think about it, this is almost pit scouting insofar as it relies on other teams' estimates of their own abilities. I.e. asking what they are best at. I am much more inclined to trust the drive team's judgement on, say, if they can feed balls to an offensive bot or are better at close-range scoring, rather than a member of pit crew [because scouting captains are never available to talk to at their pit...] saying that 'yes we can score during autonomous'.

As far as defense: I would say it worked pretty well. But I realize now that it might not have been the best strategy at another regional. The Autodesk Oregon Regional was a week 1 event, and I have never seen so many dead and/or unhappy robots. [All Thursday I was embarrassedly telling my rookie scouting partner that 'it really does get more exciting!']. This meant that rather often our team had to take the place of missing robots, sometimes even on D. [This maybe more than anything, because of our strategy of placing weaker bots on D]. And often this meant changing zones. If a strong opponent scorer was in mid-zone trying to score, it worked better to have our D working there, and potentially being able to hand ball-control over to our offensive bot, rather than running to and fro in the far zone.

Am I ever glad we aren't a first-week event next year. I think the fact of having to deal with frequently less-than-functioning team-mates influenced our scouting strategy. Really, it was quite an event if a team scored two or three points during Quals. :P
Hope that helps, feel free to ask more :)

and Question! How heavily do most of you-all scout earlier regionals? Just watching for game trends, or actually scouting teams who'll be at your regional later? We've never had the opportunity [AOR has been week 1 for years] but this year we are a week 4 event, and I am getting enthused for watching early competitions. I'm especially interested in Seattle, which I understand is the week before AOR, because of the possibility of actually scouting teams who will be at AOR.

Re: Scouting, is it important?
 

Scouting earlier events is a tricky thing to get right. On the one hand, it can help you see how the game is played out, and what types of robots might be good this particular year. On the other hand, the webcasts are not great for scouting, and the play style will change between events.

For example: Waterloo has (or had) a reputation as a defensive regional, where you could expect damage. Other regionals are more focused on offense, like Midwest. And if your scouts choose a "weak" regional and you're playing in a "strong" regional, the game will totally change. ("Hey, guys, you know how we said that scoring 3 balls would be enough to win 75% of our matches at that event? Uhhh... We're playing against 469 at this event, and they score 6 balls, again, and again, and again, and again...")

Going with using early events as a test of your scouting is a really good idea, but take any results with a grain of salt, as things change between, say, Granite State and Los Angeles.

Re: Scouting, is it important?
 

That's an interesting trade-off to make. I like it; glad it's working for you. I'll run it by my drive team next time I'm back from college. (I'd say scouts too, but she just graduated. :eek:)

Defense - Quantifying: yes it is, that's why I'm such a fan of 1114's average defensive score numbers. They take some calibration for off-season (new drivers, etc), but if they're high you can expect them to stay there. When possible, I put a drive team member in to scout them so we sort of standardize the feedback at a useful level.

Pit Scouting: I've found this as well; people respond much accurately if you give them (preferably two) choices than just a yes-no question. I guess that does make the less information trade-off easier. Luckily, I've usually found that at least one driver/coach stays in the pit for most teams.

Defense - Style: Good point. That's about what we did at our Week 1, I guess. (when we were moving correctly...) Later it became more zone-D based on man-D (robot-D): we'd determine our zones by robot is/moves there at the start of teleop. Once we're there though, it's more about alliance coach communication, given the tradeoffs of switching zones with so many good (moving) robots. Thanks, I hadn't really thought of it that way. I think I'll flip our system around to make this more straight forward.

Early Regional Scouting: I don't know much about the western regionals, but in the east I do it with a grain of salt. Particularly, if someone does ridiculously well (i.e. wins, is 469, etc) that's something to remember. For those that don't do so well, that's to be expected, so I try not to read too much into it. We changed a lot between Week 1 and Week 4. Then again, for my Week 4 scouting, I did look at when team's regionals were. If they'd just been to a Week 3 and had also done a Week 1, I was more inclined to think they'd preform Week 3-like. A lot of teams I've found stabilize after 2 events (sometimes 3).
Another helpful aspect of early regional scouting, though, is to gain more insight into how the game's played overall. Sometimes this really is about game-breakers (Beatty, 469), but sometimes it's an overall trend or an easy change we can adapt to. I think in '08 we added/changed our ball-knocker-downers after seeing MOE (?) use theirs. To be honest, though, I get a lot of that from CD directly as opposed to match footage. Eric's right, webcasts are tricky.