How do you scout?

[Dan Petrovic]

It is important that you get data you can rely on.

Make sure the scouts are well-trained, know what they are doing, and know what they are looking for.

If you don't like the idea of putting info into an excel spreadsheet, we made a spreadsheet on paper that covered all matches for every team and then organized them in a three-ring binder.

One of the things we look at is whether a team was defended by another team and what team that was. If a team scores well in all of their matches except one where they were defended, then that speaks a lot about the team that was defending.

Use data from the practice matches only to loosely determine strategy on Friday. Use Friday data for selecting partners and use Saturday to solidify those selections.

[acdcfan259]

Originally Posted by InfernoX14

It is important that you get data you can rely on.

Make sure the scouts are well-trained, know what they are doing, and know what they are looking for.

This extremely important. This year we had people who really didn't want to do the scouting, and we got very unreliable inconsistent data.

[Akash Rastogi]

In addition to the valuable data, make sure all your students pay attention to special driver's characteristics and how they cooperate with their qual. alliance partners. Trust me, that is a huge part of selection.

If there are any teams in your area attending an off season and actually scout, ask them if you can join in and learn how they function, its better to be prepared before next season's competitions.

[akeisic]

No matter what you system you derive, I think there's a balance you need to strike between the Excel spreadsheets and the paper notes. The Excel sheets make quick work of all the calculations, but is slow to navigate. The paper version is just the opposite.

Team Overdrive (FTC #74) had the most amazing scouting I've seen first hand. They had a sheet for every team out there with a pasted on picture of the robot. They manually recorded the happenings of each match (see other's comments for typical criteria) on each team's sheet. They then tallied the raw data in Excel and made printed their ranking sheet for their team representative to assist in alliance selection.

What they did there is pretty typical, but what set the scouting apart was they kept the paper versions filed in an portable accordion file that their captain kept ready at the field. As we progressed through the elimination rounds, we simply pulled the paper of the teams we were about to play and had all of the information at our finger tips - including autonomous routes. We were then able to strategize on the spot and did well as a result.

I was impressed so much, that I hope to emulate their system with my team this upcoming year.

[astephen68]

When you scout depending on the situation, you should have a few things in mind: How do i get the information i need that it's a quick process of checking or circling the answer, do i really need this data or not, Because in a match there is only 2:30 for you to get the answers on that team so you need to be processed. My team for scouting this year used a excel database which we input the data from every team from papers and at the end of the day discussed what teams we need to focus on the next day.

[kramarczyk]

We have generally done a web based system so that we can do multiple inputs while simultaneously having the field team access the data. It really helps when we can put up a wi-fi network (yeah, yeah, 2009) so that we don't hae to run between the stands and the pit. It also allows us to share the data easily with anyone interested.

You can check it out at http://gearheadcrossing.org/2008/

[TheOtherGuy]

How do I scout?

...carefully

I finally compiled and ran a Wii homebrew program (check out the scene), and several nifty graphics libraries have been released recently. If all goes well, we'll have some wifi within the next half year (pleeeease!), so I can combine the DS scouting with an automatically updating Wii database program that displays in the pits Now, to me at least, that would be epic!

[Devon27]

In addition to our normal scouting (posted earlier in this thread) this year we used overhead projector transparencies with the field, overpass, and starting areas printed on them and drew on the transparencies where the robot went in hybrid. We then sorted the sheets by team allowing us to see if a team had more than one auton/hybrid strategy and approximately where they went and what they did on the field.

[AndyB]

1 sheet for each team with enough data for 8 matches.

Find 7 people to scout (you might have to be one of them).

6 of them will be scouters and 1 will be a manager (could be you).

Each match, the manager sorts through the 50-some-odd sheets and finds the 6 that are competing (not as hard as it sounds), and hands them out to the 6 scouters. It's easiest to designate 3 as RED scouters and 3 as BLUE scouters for the sake of being able to find your robot.

Then after the match, the sheets are exchanged and put back into order.

On Friday night, we either Black List or Green List robots depending on their performances. Black list means that you are done with the robot and they will not be considered for the final list (barring a miracle). Green list means they will continue to be scouted on Saturday. Hopefully you will end up with 30-ish green robots. Sort these green robots into order from best to worst. You may end up with more than one catagory (I.E.: Hurdlers and Speed, Ramp and Ringers, etc).

On Saturday, we scout just like normal and about an hour before the the final match starts, the scouting leader (you) and MAYBE 1 or 2 others that have been scouting with you make final revisions to the list created the previous night.

We also make a list of teams that we would decline if we are in picking position.

I kind of adopted this tactic from 269. I modified it slightly for use with 171 and so far, it's worked wonderfully. We are contemplating using a PDA-based scouting system but with the new control system for next year, it may need to be a wired system.

The system would replace the finding, distribution, and exchanging of scouting sheets and instead, leave that to the scouter.

---------------------------------------------------------

As for creation of the scouting sheet, I try to keep things as simple as possible for the sake of my scouters. If certain fields are unimportant, not considered in the end, or are covering the same thing as other fields, then the sheet needs to be revised.

For instance, some scouting sheets I have used include fields that rate driver skill, robot attributes (speed, agility, etc). But I have found that in the end, if I work with the sheer numbers they put up (if they hurdle 6 times, is driver skill relevant?), I can get a better list. I also make sure to include a comment section if the scouter feels the need to take notes, or notify me of something regarding the robot.

---------------------------------------------------------

It is also important for the person running scouting operations to keep the coach and drivers of their team informed so they can come up with a viable strategy. If I'm not scouting, then I try to run down to the pits with the sheets for the teams in our match. If I am scouting or managing, then I give the information via cell phone.

[Akash Rastogi]

Originally Posted by AndyB

1 sheet for each team with enough data for 8 matches.

Find 7 people to scout (you might have to be one of them).

6 of them will be scouters and 1 will be a manager (could be you).

Each match, the manager sorts through the 50-some-odd sheets and finds the 6 that are competing (not as hard as it sounds), and hands them out to the 6 scouters. It's easiest to designate 3 as RED scouters and 3 as BLUE scouters for the sake of being able to find your robot.

Then after the match, the sheets are exchanged and put back into order.

On Friday night, we either Black List or Green List robots depending on their performances. Black list means that you are done with the robot and they will not be considered for the final list (barring a miracle). Green list means they will continue to be scouted on Saturday. Hopefully you will end up with 30-ish green robots. Sort these green robots into order from best to worst. You may end up with more than one catagory (I.E.: Hurdlers and Speed, Ramp and Ringers, etc).

On Saturday, we scout just like normal and about an hour before the the final match starts, the scouting leader (you) and MAYBE 1 or 2 others that have been scouting with you make final revisions to the list created the previous night.

We also make a list of teams that we would decline if we are in picking position.

I kind of adopted this tactic from 269. I modified it slightly for use with 171 and so far, it's worked wonderfully. We are contemplating using a PDA-based scouting system but with the new control system for next year, it may need to be a wired system.

The system would replace the finding, distribution, and exchanging of scouting sheets and instead, leave that to the scouter.

---------------------------------------------------------

As for creation of the scouting sheet, I try to keep things as simple as possible for the sake of my scouters. If certain fields are unimportant, not considered in the end, or are covering the same thing as other fields, then the sheet needs to be revised.

For instance, some scouting sheets I have used include fields that rate driver skill, robot attributes (speed, agility, etc). But I have found that in the end, if I work with the sheer numbers they put up (if they hurdle 6 times, is driver skill relevant?), I can get a better list. I also make sure to include a comment section if the scouter feels the need to take notes, or notify me of something regarding the robot.

---------------------------------------------------------

It is also important for the person running scouting operations to keep the coach and drivers of their team informed so they can come up with a viable strategy. If I'm not scouting, then I try to run down to the pits with the sheets for the teams in our match. If I am scouting or managing, then I give the information via cell phone.

That's pretty much a perfect summary of how we scout as well. One tip to help you out: have post-its or tabs to sort team numbers.

[Blue_Mist]

Since our team is pretty small and most people stayed at the Pit, I did almost all of the scouting for our team. Before each regional I wrote up scouting sheets that were basically one continuous table, with one copy as the Master that stayed at the Pit, and a few other copies that floated with the scout(s).

Each row was one robot and each column information on said robots. Each row was a different width depending on my speculation of what teams would be scout-worthy, based on past experience: say, 254 would get more space than a rookie. The first column would be team number and name, and then the name our team uses, say Team 192, GRT, we would refer to as Gunn instead, or Team 100 Woodside/Carlmont rather than the Wildhats. The second and third columns were Game Play, with checkboxes for (2nd Column) Auton, Laps and Control, and (3rd Column) Hurdle, Place and Remove. Each category had space after it for notes. The last column was Reliability, and that we left blank, for notes. I printed these out and let pencil and paper work their magic. Each scout I could gather would mark down and note robots and then bring their sheets back to the Master sheet, and that way the Master was always available. Feel free to PM me if you want the actual sheet. (It's a little scary; Silicon Valley and Davis were four months ago but I can put this down directly from memory...)

We never did set up an analysis system, but before each of our matches, I would mentally analyze each robot that we were working with or against and gather the other two robots on our alliance and discuss strategy with all three drive teams.

Pit Scouting was crucial here. I drew each robot in its team number and name space, and in this way was able to keep track of each robot on the field. Some teams take photographs as a parent did on our team, but I found the drawings easier to edit in case a team dramatically changes their robot. I also asked questions all three days, and while the numbers are a usually little inflated, most rapidly deflate when compared to match scores for individual robots.

[smurfgirl]

Another suggestion I just thought of for if your team is so small that you can't afford to devote more than, say 2 or 3 people to scouting:

On Thursday, have your scouters do pit scouting. Have them watch matches on Friday and Saturday to contribute to the pit scouting, but set up a partnership with a larger team that has an established system. A lot of teams are happy to share their data, so you can get all the number facts from them that you might not have quite enough people to gather, but you can supplement that with qualitative and more subjective data that you can use for analysis.

[MrForbes]

I've seen 842 have a laptop with their (fantastic) scouting data on it available in their pit for viewing....

[AndyB]

Originally Posted by smurfgirl

Another suggestion I just thought of for if your team is so small that you can't afford to devote more than, say 2 or 3 people to scouting:

On Thursday, have your scouters do pit scouting. Have them watch matches on Friday and Saturday to contribute to the pit scouting, but set up a partnership with a larger team that has an established system. A lot of teams are happy to share their data, so you can get all the number facts from them that you might not have quite enough people to gather, but you can supplement that with qualitative and more subjective data that you can use for analysis.

We did this at the Wisconsin Regional. We only 4 or 5 kids to scout so we offered to team up with team 2194. They were more than willing and I think we both enjoyed the lightened workload.

[Tom Line]

Our scouting team did the best they've ever done this year. They had paper sheets for each match with all the pertinent information - robot number, auton lines, pens, regular lines, hurdles, herds, even a map that showed the robot's course in autonomous.

Then they had a pre-made spreadsheet that they entered the data into and could sort with drop-down sort boxes. Finally, they had a couple folks going around taking pictures of all the robots. We viewed those pictures as we discussed each robot.

At MARC where we didn't have the resources to do a full scouting, we had 3 people watching what happened and feeding the info real-time to a person entering the data on the computer. That way we had it immediately sorted when team selections came right after the last match.

On thing that was helpful to me at MARC was that the formulas were set up to calculate average score in auton and normal modes over the course of the competition. So we were able to look at all-around good robots, and were able to see that some robots got by on the strength of their auton, and others the strength of their hurdling.