hey guys, I need some help

Competition Scouting
1

I’m a member of a third year team, and this is my second year; however I’m not sure if last year counts much for me… they had me stuck at a desk sending out letters… Anyway, we have 10 team members, about four of us have real experience. In previous years, we have not set up any kind of strategy, and scouting is a foreign term to us. This year we’re attempting to be more organized, and they have put me in charge of both strategy and scouting. :ahh: That’s really nice and all… and I’m glad they want my help (I’m really their only option), but the problem is I know precious little about how to actually do this. I understand that scouting entails collecting information about the other teams’ robots… and strategy, well dur.

So I guess what I’m really asking is… how do I get started? I’ve really no clue at all. I understand how the game works and all of that… but I don’t know, I’m just feeling a little lost. We’re the only team in the state, so honestly, I don’t know anyone who can help me out on this. Any tips or information you can give me would be greatly appreciated. I just feel like a very lost little girl. Thanks for reading my sob story… --yelly

2

we have a scouting database you can use at your discretion.

it should be up on the site at the end of the day…

all you do is input the required info. and let it go.

but to give you a basic overrun of scouting, we’ll break it up into 3 sections:

robot info. - size, weight, general class of bot (stacker, pusher, etc), can go under the bar, drive system, etc etc

match info. - human player strategy, auto. program, QP’s scored, general team strategy, approximate speed, etc etc

some teams even take down another set of information on how easy the team was to work with and how friendly they are.

i hope this helps, and if you have any questions you can im me, pm me, or e-mail me. :slight_smile:

*jeremy

3

Our team has found carrying a laptop with a simple Excel sheet helpful for scouting…and for strategy, it is nice to have a small dry-erase board (or something like it) for laying out the field and discussing strategies with your match alliance partner - hope it helps and good luck

4

Basically the point of scouting is to create a database of information to help quickly plan each round based on who your opponents are. I guess most teams have an electronic database of some sort, such as a spreadsheet. My team has I giant three ring binder, so it is possible to go low-tech. The information you want, about each teams capability and strategy, can be acquired by viewing websites ahead of time, going around asking questions on the first day of competition, or watching the matches themselves. Making up a questionnaire to ask each team on thursday of your competitions is probably the best idea if you’re short on people to do scouting. Hope that helps and isn’t too much you already knew.

5

Last yr our team went with the binder… one thing I have to say is, if you go that way, get real organized… our problem was we couldn’t figure out which sheet went with which bot… or where each sheet was. If you can go electronic, I really suggest it. I think our team will probably be taking someone’s pc to do all of the databasing…
One of the best tips I can give you is go out and talk to the teams, and be really nice. Bring some swag with you (we’ve found food works well) and just ask them about the different specs. One of the things you should try and think about is what they would be like as an ally and as an opponent. Think about different strategies you could use to win with that bot, but also how to counter those.
That’s all I’ve got right now.
~Am - team 783

6

I have written a vb program to do the databasing, but it could stand to have some more stuff added… if you have anyone on your team that knows vb they probably can edit it. As of right now it has options for stacker under bar king of the hill or stack attackers/defenders. Also it has a comment box to add anything you like about the robot, and quite possibly the coolest thing is the ability to add 3 pictures to every entry. Our team is going to use paper sheets, and then enter it all in a database later, hopefully with picture.

if you would like this program just pm me and i’ll send it to you.

7

erm, i dunno about you lot but we did pretty well without a complex database and “plays”. nitty gritty details about other robots are fine but i think they probably give you a bit of an information overload. So in that respect there’s no real need, i feel, for databases and reams of information about each team.

Driver skill and luck seem to me more important than how many reserve human players a team has - thats just the way things work out.

Useful information that might be worth collecting is what position other teams are on the leader board - giving you a general idea of their performance and what is the main function of their robot. We played it pretty much by ear last year, concentrating on the next game, not the one in 4 games time.

Tatics-wise, i’d just say keep it simple, the more complex a plan, the more things can go wrong and will go wrong. A too detailled approach over-constrains the driver.

8

think about auto… if you have a ramp controller you need to know what they are gonna do… if you get trapped on one side because they beat you to the ramp then you are shot… scouting will be important this year

9

but you can do that game-by-game. Theres no reason to have a database of everyone’s auto code. Theres ample opportunity to talk to your alliance partner beforehand to sort any conflicts out

10

I agree that there’s no need to know how many backup human players they have, or even what sort of drive system they have I’d add. I am not so very technically inclined, but still ran our scouting system. Focus more on what they do than how they do it. I’d argue with the idea that you should only work a game or two ahead, though. Some teams will have more than one autonomous program, and if they’re up against you, they probably won’t be volunteering too much information right before the match. The only way to know something like that is from watching their games. In the end you’ll be limited by the number of scouts you have and how many other responsibilities they have, it’s hard to watch everything, but the more information you have, the better your strategy can be.

11

I’d argue against asking teams what they can do in favor of watching practices and matches.

The state of a team’s robot is likely to be in flux. You might get info that is several hours out of date, since modules on a robot are blowing out or being added.

Also, who you talk to on the team is important for the reliability of information. Some team members will know the real deal; others know only the team propaganda.

Quantitative info is better than qualitative info.

How fast are you? Fast/medium/slow

is not as good as

What is your top speed? ___ fps

And this last question is better determined by a stop watch in the stands than asking even the lead engineer. He is more likely to tell you a design number than a real number.

Also remember on asking teams for info, they may be in the middle of putting their stuff together for their practice/inspection/match and you are the 56th scout to ask them for info in the last 15 minutes.

Although you can certainly do some scouting during unpacking, we tend to scout at about 7:00-8:00 pm on the practice day. It’s so much more informative and interesting to actually look at the robot than to ask someone about it.

12

WOA WOA WOA hold on scouting masters, no need to scare them.

We use a filemaker database to do ours but we also use an extremely low tech method…paper and pen.

its useful to carry arouind
we have about three lines for each team and we write down auto modes, powre, speed, drivemotors, special features, basically stuff your drivers would like to know. Start simple, it could get hectic fast but there will always be teams willing to help (217 for example) dont be afraid to ask

13

My team went with the binder this year and last year (not sure about the years before then…) and it works just fine. My only advice besides keeping things in order is bring a hole puncher as many teams now have hand outs about their teams that might be helpful. The binder is good just because if you’re scouting a team and someone else wants to help you can just take the page out and let them do their own thing… (ex: scouting from the stands you can have 2 people each with two teams instead of one person on a laptop trying to follow 4 teams). Also, some questions I’ve learned are VERY important to ask are whether teams have suction cups, etc. to keep them on the bridge, if they have wings or arms, and if they have been to any other competitions this year (and if so how did they do). Hope that helps, and sorry if I accidently repeated someone elses advice. ; ) Good luck with everything!

-Kelly, Team 287-

14

OK, here is another way…

First, I totally agree that you have to talk with teams and get a feel for where they are, but I don’t think that is enough.

If you have a chance of ending up in the top 8 (or 10), you need to have a draft list about 25 or 30 deep to rank who you will pick. I personally believe it is better to quantitatively rank teams, and here is what we did for our two regionals.

We made a quick little spreadsheet (all our student have ibooks) and we rated the 'bots during the qualifiers. We scored three main areas: Driver Ability, Offense, and Defense. Each of these was 1/3 of the total score. In Offense and Defense we had some sub-categories like for defense, ability to get under the bar and being a powerful pusher (versus getting pushed around). Where ever we had sub-categories, the rankings are averaged before the three areas are totaled.

We rank on a simple 1-5 scale where 5 is the best. At VCU, we ranked the last two qualifier rounds, entered the data, and then sorted the spreadsheet based on the total score. Using this draft list, we ended up with the winning alliance. Here is the strange part, I don’t know if we would have picked the teams we did if we were just going off personal subjective opinions.

The other thing that was fun about this is that we gave 5 guys on our team the job of rating the matches - one to watch each team and one to run the numbers back to the two of us who were putting the data in. It was a hectic three hours Sat AM, but the guys who helped get the data enjoyed it, and we were very happy with the results.

Our only adjustment to that original scheme for the Chesapeake regional was to add a fourth category for a subjective opinion (1-5) of the team & bot. Then each of the four categories contributes 1/4.

Background note: One of the reasons I suggested to the team that we rate this way is from my experience as a volleyball coach. Sometimes I see a player make a few outstanding plays, and that player sticks in my head. But our players keep stats for all their matches, and frequently I find that the players who were contributing more consistently weren’t the ones I noticed, but rather someone who was always doing what they needed to. It is important to make sure I play the players who consistently contribute.

15

I found that asking teams what their robot can and can’t do before a match is quite helpful. However, it is also a good idea to just watch the team perform in a previous match. Sure a robot might sound real good because it can stack, can move fast, is maneuverable and so on, but all that means nothing without reliability and a good drive team.