Calling all scouting data!

I keep getting this message: You need permission to access this published document.

Both

Hatch Panels and balls at levels 2 and 3, also hab 3 climb

Started with everyone scouting, at DCMP I reviewed and scouted every match on my own at the end of the day with help from a few team members for a list of teams to focus on.

My head

Idk, 5 or so hours of looking at the data and matches and then somewhat arbitrarily deciding what teams I wanted to pick.

To say the least I need a better system.

It’s probably the email I’m using, I’ll have to log into a different email to fix the problem, I can fix it in less then twenty minutes

i highly recommend using the app robot scouter on amazon tablets you have to back door download it but its extremely easy to use its customization and uploads to a google spreadsheet with averages and individual team data. i suggest making a email to link all the tablets but it has been so much easier to do everything of course there is room for it to improve but it has been so simple.

1 Like
  1. electronic
  2. the main data points I remember that we made use of were pretty usual things such as cargo/hatch count, rocket 1/2/3 counts, cargo ship counts, climb counts, and per-match defense ratings
  3. once or twice, don’t remember details, probably little functional things in our app
  4. we wrote an Android application to use for scouting, each scouter would record data then send a CSV file to a master device with Bluetooth. We then analyzed the file with Tableau on someone’s laptop. Late in the year, I added a basic analysis screen so that we could take a quick look at scouting data on the fly without deep analysis
    4a. it worked great. I would recommend using a scouting app, and if you don’t want to program it, I think Robot Scouter on the play store looks great. However, I would recommend programming it for the reason that it allowed us to have complete control of the behavior, and it was a great learning experience for the programmers involved.
  5. we analyzed it in Tableau and looked for different criteria, which were a combination of general things we look for and specific things we wanted at the competition.

for some general advice, you should try to get the input of as many people as you can on the team; this way, you can find out things that they noticed that you may have not.

here’s a good paper: Your scouts hate scouting and your data is bad: Here’s why - Competition / Scouting - Chief Delphi

1 Like

Scouting Systems Survey relevant thread about how teams scout

Sorry, I’m a bit late to the party.

1.) How did your team scout? (Paper, electronic, etc)?

Our team scouts electronically with cheap Samsung and Lenovo tablets using the Robot Scouter android app.

2.) What was the main data you looked for/found most important? (Cargo, climb, etc)

We calculated the number of points robots scored during matches by tallying their in-game actions. In qualification matches, this was pretty useful for comparing the “point scoring potential” between two alliances. For example, if one team on the opposing alliance is a much better scorer than the others, we might choose to send one of our 'bots over to interfere with their scoring.

Keeping track of HAB climbs was also really helpful for our team. Our robot’s climber wasn’t the most reliable this year so we liked being able to compare climb success rates to determine which one of us would go for Level 3 and get the ranking point.

Rocket stats were another thing we looked at a lot. If our alliance partners want to and the scouting data indicates that it’s possible, we’d try and go for a rocket RP during the match.

3.) Did you change your scouting at all throughout the season?

Not anything major that I can remember. If we were able to pick more often, we probably would’ve liked to alter our scouting sheet to get more specific data on the defensive capabilities of a robot. We made a lot of changes to our scouting data briefings that were sent to our Driveteam though.

4.) What program did you use to take in and look at your data?
This was our first year using Tableau and we really like it! We export our data directly from Robot Scouter and run it through a simple command-line tool that converts it into a format Tableau can understand. The app also exports a JSON file of raw data that you can write your own application to process as well.

Here’s a link to the sheet we made for alliance selections and an example driveteam briefing. Sorry that they’re not as high quality as what other teams have made – we still have a lot to learn.

4a. Opinions on this system?
Overall everything worked pretty well for us this past season. We don’t really have the resources to program a separate data collection/analysis application every year so we like the simplicity and versatility Robot Scouter gives us. We haven’t done any substantial data analysis in the past so we have no way of comparing our current system to others but based off of this year we’ll definitely stick with Tableau in the future if it’s in the KOP again.

5.) How did you use this data to make a pick list for alliance selections?
We first create an alliance selection workbook in Tableau (see above). It’s set up so we can easily find robots that do specific things well (best scorers, hatch panel placers, cargo ship fillers, climbers, etc.). We also look at qualitative notes that scouts have made about the different robots.

Lemme know if you have any questions and I’d be happy to try and answer 'em! We’ve written a detailed introduction to our scouting process here if you want to learn a lot more about our method.

Robot Scouter was also developed by one of our former members so please feel free to ask me if you have any questions about that!

1 Like

1. Our team originally used a paper scouting system at our first event but later used a tablet scouting system (electronic).

2. The data that we found most important was whether a robot of interest could effectively place hatches, especially on the rocket.

3. We changed our scouting system during the season.
3a. We went from a paper scouting system to a tablet scouting system. At our second event, we transferred the tablet data by connecting the tablets with wires to a master computer. At States, we experimented with Bluetooth transfer but it didn’t go very smoothly and we ended up having to use wires again to transfer data. However, by Worlds, we used Bluetooth to great effect as we were able to essentially collect data at will. We did this by transferring the files from the tablet to a master computer. We had apps on both the tablets and the master computer to help with this process. We used ASTRO Bluetooth Module and ASTRO File Manager on the tablets and on the master computer, we used an app called BlueFTP.
3b. We changed it due to paper recon being very exhausting for the people who enter the data into laptops. By using electronic scouting, we were able to completely bypass this step and have data instantly.

4. We used Excel to view data.
4a. Excel was great for me since I’m the most familiar with that over something like Google Sheets or Tableau.

5. We sorted this data by descending order (average hatches placed per match) and with input from our drive coach, made a picklist for robots that we could work well with.

  1. My team used a program, called epicollect 5, for data entry on mobile phones


It’s a really simple and easy to use app, super easy survey customisation and works with IOS and Android. It also allows offline entry for later upload. I would highly recommend it

  1. The data we focused on was mostly basic. We did total hatches and cargo separately, maximum height of placement, habitat at start and finish.

  2. Epicollect 5 does some basic tabulation, and you can export it to excel easily. You can do that if you want, but if you want data to constantly be put in an excel spreadsheet or whatever program you are using, the system has pretty simple apis.
    Here’s the basic table:
    https://five.epicollect.net/project/koalafied-2019-qualifiers/data

Here’s the csv format: (look at the web address and you can get your program to do it by entering the name of your survey in the right spot, assuming it’s public)
https://five.epicollect.net/api/export/entries/koalafied-2019-qualifiers?format=csv&headers=true

From the csv format page excel was able to automatically access it and do whatever we wanted with it. Here’s a screen shot of the final table


There’s a few things that I’ve messed with a bit, but the bulk is apparent. We didn’t just rely on this, and looked at this with a combination of communication, and more qualitative data as well.

I’d highly recommend using this system, and I’m happy to help if anyone has questions about it. From my experience, epicollect 5 has the same if not better functionality than custom apps, and is way easier to use.

2 Likes

We mostly used paper and found that it was very inefficient. We tracked basically everything worth points, and things like if they were defended or not. We added the “was defended” box after provincials to help differenciate the data. We also redesigned the sheet after provincials to make it easier to use. We interpreted the data in Excel, which worked really well for us, since we could see the actual numbers. For picklists, we mostly looked at things like high level stuff and their Overall scoring, which was in a sheet called “organized Scouting”

This topic was automatically closed 365 days after the last reply. New replies are no longer allowed.