Collecting Info + Managing Scouts

Competition Scouting
1

Hey guys! It’s a little later into the build season and teams are making progress (hopefully!) on their robots and so competition strategies are developing.

What do you think is vital information that scouts should be looking for that some people may not consider? Obviously most teams will cater to their robot’s needs, but there is some info that most teams consider important, regardless of design.

How is your team scouting this year? My team is doing simple paper scouting mixed with laptops for graphing and compilation in Excel or Access.

Do you rotate scouts to keep them from getting too worked out, tired or restless? I dont mind sitting and working on a computer all competition, some kids get restless and tired out, understandably.

2

my team does the same thing as you. we also have an app in progress. it is still being worked on. it is called our alliance. we use both paper and our app to do this. the app is for android. this year i feel like human players will be a big role in this so i will focus on that. drive train as well. for example i am hesitant to teams with tank because they don’t have the same mobility as a team with mecanum and swirve. (pardon my spelling). another thing is how sloppy there drivers are. if they are sloppy drivers they can knock over totes more likely then those who aren’t.
good luck this year with scouting

3

Following the extremely beneficial addition of a dedicated scouting mentor last year, we have converted 99% of our scouting process to electronic collection, compiling, and analysis.

We use 6 Windows Surface tablets for collecting solely objective data in a program created using Access. This data includes:

  • Autonomous capabilities (where they start, what they pick up, where they put it, etc.)
  • Teleop capabilities (where they pick up, what they pick up, where they stack, how they stack, etc.)
  • Other data that I can’t remember despite being a lead scout :smiley:

The tablets are also used for pit scouting prior to the start of matches. We’re looking at drive bases extensively this year, for reasons that should be somewhat obvious, but we still collect many other pieces of information about every robot at the competition.

As for rotating out scouters, we have a larger than average team this year, so we’re planning to have objective scouting shifts running about an hour each, depending on the team’s attendance at each competition.

In order to collect subjective data, something that’s used extensively during pick list meetings, we use a different scouting program (still created in Access) that’s designed to specifically record the scouter’s opinion of each robot in each match. Subjective data is exclusively tracked from laptops, and includes categories such as agility, willingness to work with teammates, and a general comments section for anything that doesn’t fit into the pre-made categories (which is usually a lot).

All of the data collected from the tablets and laptops is synced via USB and compiled into one Access program, created specifically for analysis. In order to streamline the process of data analysis, we split our scouting team into several smaller subteams, each with their own specific task: the level I’s collect objective data, the level II’s collect subjective data, the analyzers sort through all the data during competitions to plan for future matches, and the strategists work with the analyzers to plan for future matches, as well as creating a pick list. We’re hoping to utilize Tableau (more than we did last year) to analyze the data and recognize possible trends.

We realized a large barrier we had last year was effective communication between the scouters and our drive team. In order to combat this, we plan on installing a monitor in our pit that will show a condensed and easy-to-read version of the data collected throughout the day. Additionally, our lead strategy mentor has his own Surface which will also constantly be updated with the data collected, meaning he’ll be able to view info about upcoming matches from wherever he’s gone off to in the arena. Hopefully, the pit TV will allow the drive team easy access to any data regarding future matches they wish to see.

Since you’re from 3525, OP, you’ll be able to see our CyberScouter system in action at Waterbury! Any questions, please feel free to ask.

4

If you have enough people available, definitely do shifts. In my experience, only a precious few teammates are able to scout at full capacity for an entire day. Scouting takes a lot of focus, and as soon as scouters get tired it gets harder and harder for them to keep track of the right data. When scouters are tired and frustrated, it severley limits their ability to observe effectively. Plus, they’ll totally hate you for making them work all day. :yikes:

If you have too small a scouting team to rotate in shifts (like our team last year), here are some tips:

Bring snacks. I personally carry around a backpack full of candy, snacks, and drinks at competition just for the scouters. Keeping scouters hydrated and fed is very important for keeping them engaged.

Provide feedback, don’t just collect and file away the data. If you can, have someone browse over every paper submission before filling to congratulate good insights or offer suggestions. This makes the data you collect more complete and lets the scouters know that their work is being looked at and is appreciated, thereby making them more engaged and productive.

See if there are any opportunities to lighten the scouters’ workload. For example, our team makes a preliminary picklist the night before eliminations, as we have pretty sufficient data by that point. We took advantage of this last year by reducing our scouting team the morning of elims to only about 3 students who wrote comments on major events/ certain teams. This meant we could run shifts with the smaller team and still get enough data to update our match strategies and preliminary picklist.

5

My team does paper scouting sheets which the data is then entered into a computer for further purposes. We do six scouters, one field viewer, and one data monkey at any given time during the competition. There are two data monkeys that switch in and out, usually two field viewers to switch between, and then the six scouters usually have an hour or two of scouting then an hour break, rotating in everyone who isn’t working in the pits to help at some point.

6

As a former scouting lead myself I would say that the general view of what to do is pretty close to what I did. We rotated shifts between the objective scouters whenever someone needed a break, and then we had two subjective scouters that did their best to stick in it all day. Having a few individuals who are able scout the entire day gives you a few people who really know the ability of all the teams and can make the best decisions.

7

Thanks for such an extensive reply! I didn’t want to quote your entire post, but I wanted to note the CyberScouter system specifically. I got to see last year’s program/tablets at FIRST Day at CCSU, and I’m a fan. The layout is convenient and easy to learn.

Currently my team has about 50 kids, and from what I’ve seen 195 has always had a good amount of students so we’ll probably end up with slightly longer shifts, like 4 or 6, unfortunately. Can’t wait to see you guys at our event!

Again, this is awesome advice. While I love my job, being in charge of a group of tired, stressed and restless teenagers was a nightmare sometimes, especially because I was expected to reassure them. I was definitely not a ray of sunshine 24/7 myself, which is something I really want to work on this year. I’m always looking to lighten their workload, but making a conscious effort to povide positive feedback and bringing snacks (FUEL) seem like good points for me to focus on this year, thanks!

Any advice for any Scout Lead that works through the entire competition? Sanity is important!

8

Over the years, I’ve had some really great and really terrible experiences with scouting. One thing that was standard across all the teams I’ve worked with is that consistency is absolutely necessary to gain accurate insight from data. Regardless of collection or compiling methods, there needs to be some training hours put in for the scouts.

To demonstrate the concept, I’ve had a scout team sit in our computer lab, and all watch the same robot in the same match and fill out a standard scouting sheet. When we compare the sheets, a lot of them are very different despite watching the exact same video. Autonomous placement can be off, the scores are wrong, and the subjective feedback can be very different from person to person. Running lots and lots of practice matches with your sheets is important to minimizing this problem.

Another thing that can help with accuracy/consistency is having “double teams”, or a scouting shift of 12 students instead of 6. Obviously this isn’t feasible for every team, but having two sets of eyes on a robot can make a huge difference in terms of the quality of data that gets collected.

I like to organize my scouting teams into shifts of three: Red Shift, Green Shift, and Blue Shift. Each shift is about an hour long, organized by match number rather than time (15 matches per shift vs. 2 hour shifts). The key to making shifts work is to have students that will show up when they’re supposed to. Competitions are hectic and exciting, and shifts are great when all students are accountable. Make sure you have mechanisms in place to get students back to the stands for their shift, or the whole system will get thrown out of whack.

Generally, the head scouting student is in charge of logistics, managing the match schedule, and ensuring that the needed information gets to the drive team on time. I try to assign them a pit runner, whose only job is to go back and forth with information, or find people from other teams that the drive team would like to meet with. Having a walkie talkie set between the pits and the stands has worked incredibly well in the past as well.

I’m a huge fan of technology, but it makes me incredibly nervous for scouting. No Wi-Fi, unreliable sources of power, and glitches in programs can make for some scary days. Personally, I stick to paper systems. Sheets are split into morning and afternoon, with the morning sheets permanently transferring to the pits after lunch begins. Data is compiled by ONE person after sheets are full for analysis in excel.

On Friday night, I like to create a preliminary pick list. Going into Saturday morning with a tentative list can be a game changer, because you’re not rushing to create the whole thing during the day. This way, you can identify which teams you’d like to talk to or observe more closely during the first matches.

9

Your post is full of great advice, but I’m interested in hearing more of why you do shifts by match number instead of time? I’ve always done it by time because, like you mentioned, ensuring students are there for their shifts is critical, and it is much easier to find the time than find which match is being played (especially at the Championship).

10

You mean a cell phone call, right? Walkie-talkies have been outright banned at events for as long as I’ve been in FRC.

11

Those of you guys who use paper to collect data for the matches, do you have one sheet per team per match? In the past we’ve done one sheet per match but when it comes to the different matches papers are flying trying to find the right teams for that match. Any ideas on what we can do to fix this?

12

Last year our team had to use one half sheet per team per match. To keep track of all the paper, we filed the sheets into a big binder after each match. Each team had their own sheet protector in the binder that we filled with their match data and pit scouting sheets. We keep the binder sorted by team number, so finding a team’s data was very quick and easy. Whenever we had a match coming up, we would just take the whole binder to the pits for a while.

13

I’ve seen one sheet per driver’s station for about 6 matches. 6 sheets covers all the stations for 6 matches or so, turn in for a new sheet, repeat.

It depends on the layout of the scouting input areas, though. Try to have each sheet last for multiple matches.

14

I feel so dumb now, I didn’t realize I made a typo in my comment. We usually do one sheet per TEAM. not match. Is there any specific way you guys choose which teams are on each sheet?

15

We’d organize the sheets by driver’s station. Print ahead of time with blanks for match/team, and once the official schedule is out, fill in each sheet by running down a column. After entering the data, we’d sort by team.

The team I was on at that time has since moved to other, more advanced methods of collecting the data.

16

It’s been a while since I last scouted, but a few points that I always like to note.

1.) Pit Scouting: It is oftentimes a simple task that teams will make unnecessarily complicated. For the most part, you should be able to simply observe anything worth noting during pit scouting. Take a photo, write down notes about the drivetrain (dedicate 1 or 2 experienced people to rate them), note any dimensions that may be relevant in the game (Can it fit under the pyramid, or drive wide on the bridge?). Be friendly with the people on the team you’re scouting (If this is not the hardest part, your system may be too complicated).

2.) Motivation/Shifts: How you run your shifts really depends on how many people you have and how long you can keep the best scouts in your rotation. Motivating scouts is much trickier, but usually positive reinforcement is the way to go. Water, Cookies, Stickers, SimBucks, being allowed to attend the scouting meeting are all ways that have been used to motivate the scouts by teams that I knew scouted well.

As far as dealing with the external and personal stress (which OP mentioned in one of his posts), you get used to it as you run more events as scouting lead. Identify what stresses you out (Because in my experience the source is usually the same few things), then take measures to minimize it.

17

I’ve noticed that a lot of scouting advice (not only on this thread) is geared towards alliance selection for the playoff (formerly elimination) matches. As a mentor for a team that has done a horrendous job of scouting our first three years (the first two years our efforts were barely up to half-hearted, and last year, we didn’t actually use any of the data collected so that anyone would notice): How much do you feed scouting info into the qualification (formerly seeding) matches? If you do use scouting info for qualification/seeding, when and how do you deliver the appropriate up-to-date scouting info on the six teams involved (three opponents, two alliance partners, and yourself) to your driver team, where the info would be of use for planning match strategy?

Back to the general scouting process, I’m figuring that this year with reduced interaction between the alliances, the scouting info on the other alliance’s teams can be limited to four(composite) questions, in approximate order of importance:

  1. What do they bring to coopertition? Zero, One, or three totes? If one or three, can they “complete” a stack of four, or just originate it?
  2. How quickly do they get the RCs off the step? (seconds to first, seconds to fourth, average number per match)
  3. How quickly do they get the totes off the step? (seconds to first, seconds to twelfth, average number per match)
  4. How capable is each team’s HP as a “spear chucker” (throwing debris over the wall onto the other side of field)?

The info on the alliance would include these, as well as the following, again in approximate priority order:

  • How high (and fast) can each team consistently place an RC?

  • Does each team have a history of knocking over stacks, committing fouls, or getting in the way of alliance partners?

  • How fast can each team clear the landfill?

  • How efficiently can each team work an HP station?

  • How high can each team stack totes?

  • Can each team get a noodle in/on the RCs? How fast?

  • (How fast) can each team deal with inverted totes?

  • (How fast) can each team deal with totes on a long/short side?

  • How capable/fast is each team at processing litter?

Please advise if I missed something important or you believe that I got priorities seriously out of order.

18

Gee Two, your reasoning is solid. It depends on how you scout if you use the information in the seeding matches. The best way to use the information during the prelims is to have some way of compiling the data electronically. That can be using a scouting app or your own computer based system, or it could be a paper system that you feed the data into a computer after every match. Because its electronics you can quickly analyze past performance and that can help your drive team.

Obviously you won’t have any current data for the first match and maybe not even by the second, but by the third match you should have a least a little understanding of what the other teams can do and that can be very helpful in planning for your upcoming match.

Then when you have the data compiled you send a runner, preferably your scouting lead student who knows the most about what is going on, down to the pit with the information and let your drive team know.

19

Thanks for the insight as to logistics - particularly that the compilation doesn’t need to take place in the pits, which is obvious once it is stated. We can compile the numbers in the stands, where the scouts are.

I’m wondering if we can make the scouting form look a bit like a standardized test, with lines or bubbles to fill, so we can take a picture of it and OCR that into a spreadsheet; it sounds easier and more recoverable than a scouting app, especially as our venue is known for poor cell signal.