One Day Scouting

[NonStopScouting]

Hi everyone!
So my question is: How does your team do one day scouting? What I mean by one day scouting is either a single day competition, or the last day of a multiday competition.

For multiday competitions there is time after the competition day to analyze the data and create a picklist, but creating one for a single day competition, or updating your picklist on the last day of competition can be much harder. How do you deal with this difficulty?

[GKrotkov]

Dawgma has a few possible answers for this - one of which we're trying out at Ramp Riot.

1) Superscouting/Subjective Scouting. Back in 2014, team 1676 introduced me to the concept of "superscouts." Superscouts are generally experienced strategists that simply watch matches subjectively and give direct input to the drive team*. With fewer matches to watch (like at a one-day event), subjective scouting can become vital, even to picklisting, assuming that the subjective scouts know what they're doing. At a longer event, however, subjective scouting becomes more valuable with regard to match strategy than anything else. That is, more matches decrease the value of subjective scouting with regards to picklisting.

2) Visual scouting sheets. Instead of having scouts put tallies in boxes or circle letters, at Ramp Riot we're going to have them draw what the robot did and record key figures using symbols. The numbers get input into Excel as usual, but the match scouting sheets help us to show where teams go and visualize the upcoming match for the drive team.

Now that I think about it, it's a bit like a nonelectronic version of TechFire's scouting system this past year.

3) Watchlist. If you're on the last day of a multiday competition and want to update your picklist, putting together a "watchlist" of teams that you want to pay particular attention to can allow your scouting team to shift resources around to scout those teams in more depth. For example, you might have superscouts pay attention to only those robots, and give some match scouts a break if you're confident about your assessment of a team.

I hope these are somewhat helpful.

* it should be noted that this should rarely, if ever, take the place of regular match scouting. The only time I've done that is when there were only two available scouts. (offseason event, I was there to help another team)

[BrennanB]

When making pick lists for one day events, start making the list right after the first match. Shift teams up and down based on how they preform with the input of 2-3 lead people to create your pick list. Nothing worse then getting to the end of matches and having 10-15 mins to fully build a pick list.

[BenjiSG]

LigerBots Scouting uses the first days of competition to collect objective data on the performance of all other teams (although a notes section is also kept available for specific observations).

We then compile this data into a pick list based on input from lead scouts and analysis of our data, and use the final day of competition prior to selection to take subjective notes on teams we want more data on, such as how their performance may have changed, and whether their data and their presence on the field match up.

Our pick list is modified if necessary based on these notes after qualifications are done, and then passed off to our drive team to make sure it matches with what they have been seeing throughout competition.

[chandrew]

I watch every match and make/modify my pick list based off of how robots perform and the specific things they do in matches that could be beneficial in various alliance scenarios.

[bkilen]

5172 for the most part switches over to subjective scouting on the last day of a competition. Make sure you have a watchlist of robots who could potentially work well in an alliance with your team. Our experienced scouts would go over the list before the day and talk about why each robot would be good for us. These experienced scouters take notes on those robots throughout the day. Almost no match data is recorded as we didn't have a way to compile it and use it. All the scouters still watched matches and they take note of anything interesting that would be valuable knowledge for the drive team or alliance selection.

Don't only look at the best robots. Look out for the robots that would be a good third robot. The third robot can really make or break an alliance. On multiday matches the best robots are usually known by the last day and your team probably has a picklist of teams you want to pick first. The last day, at least for me, is spent watching the mid-tier robots.

[BrendanB]

A one day event means it is an offseason. Being in districts we generally know the teams we are competing with that day and what they are capable. Things I tell our students to watch for is teams who are competing better or worse than their typical performance.

With it being the offseason teams are using different drivers at some events so its wise to find out if teams are planning to switch their drivers for the elimination rounds.

My advice would be come up with a top 14 list give or take for a 24 team elimination tournament at a bare minimum. It doesn't need to be as perfect as your competition season picklist but at least you are more prepared than you were that morning. Unless we have a specific strategy in mind for the elimination rounds I trust our team rep to make a decent 2nd round selection again using our previous knowledge of them mixed with what we've seen throughout the day.

It is rare though to attend an offseason with normal rules for the alliance selection. For example, Battlecry does random selection in the second round leaving us to just focus on who we would select first. River Rage makes the first round picks random so you have to be flexible depending who your first pick was.

With events making up their own rules its why I generally encourage our students to focus more on exercising their abilities to watch matches and make basic notes on performance to sort through the field of teams. Its a lot less stressful and if we don't have a list we don't worry about it. The purpose is more for the students to learn and become better prepared for the following season. In general the best policy is to have fun and learn something new.

As for the final day of an official competition, you want to know what traits you want to see in your picks. The first round selection is a little easier to sort through who you want. Typically the top tier at an event sticks out and their performances can be sorted by average balls scored, results from matches together, and generally just watching them to see where they peak or struggle during their matches. For our second picks we usually make columns on a whiteboard from the previous day listing the top teams in the fields we want to see our third alliance member playing. Then we watch the final matches sorting them based on how they are playing while keeping an eye for teams not on our current lists as they change how they play on the final day. After a round or two of matches we have our near final sort and aim for our list to be 26 teams deep.

Then we copy our whiteboard list on to two pieces of paper: one stays with the scouting team in the stands and the other goes on the field with our rep. Our main group in the stands puts the number we believes is best up on the board for our rep to see. We like this system as it keeps us flexible and if we select someone in the first round or are selected ourselves we can conference with our partner so both teams are heard. Worst case our rep has a list of teams with them so they can follow along with what our group in the stands is thinking or they can follow the list and make a good decision that was already vetted by the scouts.

I prefer this system as it allows us plenty of time to sort through our picklist while still keeping an eye on the field. It also means we can receive input from our entire scouting team without being hurried or panicked so we can make clearer decisions.

[Captain_Kirch]

The last time I ran a one day shift the quals got cut short by 3 hours! (To be fair, we finished packing up at 8, i cant imagine how late it wouldve run with the original plan)

Instead of making a picklist over lunch and doing the watchlist thing. I had to be on the field for selections before lunch. We just kinda winged it,(superscouted as you called it) used the same anaylsis techniques as the build season and ¯\_(ツ)_/¯. We forced a rubber match in the finals before Sir Bart broke and got its shooter jammed. I'm pretty happy with that!

I'm a big fan of the watchlist idea. Make your picklist before that last day of competition. The numbers give you an idea of who is who, but if you have to choose between .1 more shots per match, and a team with clean, fast, and confident driving, you pick the latter. We picked at both our regionals. We were confident in the driving of all our partners, and we had observed their abilities time and time again during quals.

Data doesn't show you how well a team drives, and THAT is the most important thing in playoffs. Can your partners handle defense? Can your d-bot deliver pressure and avoid penalties?

Put down your clipboard and watch a match every once in awhile

[Kevin Leonard]

So one day competitions come with a few additional challenges when trying to still approach picking objectively, and when still trying to get students experience like it's a real competition.

Without having a lot of human resources, it can be hard to do full data entry like an excel based scouting system might require.

At Ruckus, 5254 used a 1 sheet/team system, and when it came time to throw together a picklist, we were able to rearrange the team papers and look at the important stats as we were making our list. This only worked because there were just 36 teams at Ruckus, and we already knew what most of those teams were capable of during the season.

20 has a long running joke for Day 2 of competitions about our 'watchlist', where we put teams that could move up or down our picklist a lot based on their day 2 performance. At one event we had nearly half the event on our watch list, because so many teams had a ton of potential to improve, but also a ton of teams had a ton of potential to fall.

[logank013]

So, we have typically do completely different things on each day of scouting.

Day 1: We usually do all the number stuff. We collect a ton of numbers about the data we wanted to collect. We usually make a rough draft list after day one at the hotel (or lab if we don't stay at a hotel).

Day 2: We usually have a couple kids that are really into scouting like me take notes about the main robots we are watching. The other scouters on Day 2 usually report back to me their opinions and then I report to our drive coach. We usually talk about it and might make some slight adjustments.

One Day Competitions Scouting: We usually combine a mixture of both. At Boiler Bot Battle for this year, I gave kids on our team note cards and they wrote down the team number, how many high goals they got, and how many low goals they got. I kept track of who scaled and took notes. Then after each match, I entered the data into our database and kept what I call a "live database". Then when we got 5 matches before alliance selection, I made a draft list with other scouters who are into scouting and then reviewed it over with my drive coach.

If you want more details, feel free to PM me. Good luck.

[legoguy1000]

Hey everyone, I am developing a Web Application for teams to do their scouting. Please go to https://www.frcscout.resnick-tech.com and take a look.

[dirtbikerxz]

Originally Posted by legoguy1000

Hey everyone, I am developing a Web Application for teams to do their scouting. Please go to https://www.frcscout.resnick-tech.com and take a look.

I'm just getting a "Authentication required to view resource" message. And when i sign in with my google account, it just redirects me to the main page and gives me the "Authentication required to view resource" message again.

[legoguy1000]

Originally Posted by dirtbikerxz

I'm just getting a "Authentication required to view resource" message. And when i sign in with my google account, it just redirects me to the main page and gives me the "Authentication required to view resource" message again.

Sorry about that. Please email frcscout-contact@resnick-tech.com with any further issues. If it happens again, please take a screen shot and email it to me.