Scouting data available

I have seen various teams share links to their scouting sheets. I am looking for any sheets that contain detailed scouting information on individual teams for the 2019 season. Keeping it simple, I will just say that I wasn’t very successful at getting my students to really understand the purpose of scouting and how the purpose determines the data captured.

Our season ended Saturday. Looking towards next year, I would like some detailed data that I can use to teach them how to scout and run data analysis on the data. Looking at TBA, that is still at the match level while I am looking to show them data that a good team would capture during the qualification matches.

Here is a link to a thread about FIRES. Just sign up and you will have access to all of the scouting that has been done by multiple teams at multiple events.

Mr. Mike

There you’ll find the stats from the past two South Florida Regionals, the past two Orlando Regionals, and last year’s Hopper Division at the Houston Championship.

Check back in a couple weeks and you’ll be able to follow along live with whatever division we’re assigned to in Houston this year.

1 Like

For some FMA events 1676 makes their data public.

^excellent resources.
Here is a link to some of our (imperfect) scouting data for Orange County and Los Angeles regional events:

We add data to the sheet using AppSheet.
You can make a copy and sort by team # on the LAR and OCR sheets, or use the summary sheets to select a single team’s average performance.

If your students spoke to any team at the competition to discuss match strategy, they should understand the importance of scouting. Just about everyone lies (or exaggerates) about their abilities when speaking with alliance partners. Having your own data to confirm/deny is always a good idea.

Thanks for the help. My issues with the scouting students are more basic. They put together a scouting sheet and capture data typically using a scouting sheet slightly modified from the previous year competition. But they don’t seem to understand the “why”. From my perspective, in a nutshell, scouting data is used to pick alliance partners. This means complimentary teams, not clones of our team. Thus, why do we need the data, how we will analyze the data to answer this which then drives the data needed which drives the data collected. Thus start from the end and work backwards. In our case, it is collect data (based on the previous year) and hope it might be useful with no real plan.

I am hoping that if I have detailed scouting data, I can show them how to analyze it and compare to the results. In other words, what did the top teams do to be the top teams.

I am more focused on teaching the students a general analytics methodology that can be applied in a variety of situations.

Thank you all for the help. I greatly appreciate it.

Scouting is not just for playoff alliance picks. We run qual match sheets that show stats on each team in our next qual match. It helps us determine alliance strengths and weaknesses. It allows our alliance the opportunity to form a match strategy to utilize our strengths, or take advantage of the oppositions weaknesses.

The next trick is to have each team follow the strategy.:crazy_face:

Here is a link to 2370’s scouting system. I’m working on cleaning up the code to put it on github, but we use it for next match scouting as well as our “pick list” meeting we do the night before alliance selection to go over how we feel about each team and make a ranked list of who we would pick / who to try and get to pick us.

This is for the UNH New England district event.

Each col is sortable, and what they mean is in the alt text if you hover over the heading. the + on the side expands it out to the full details of what that team did in each match.

I’d love to answer any questions about it as well

1 Like

There is a ton of things you can get from scouting.

You can compare your stats if you weren’t picked to stats of a 2nd pick.
What do we need to do to get picked on an alliance?

You can compare your stats if you were a 2nd pick to stats of 1st pick.
What do we need to do to be a 1st pick?

You can compare your stats if you were a 1st pick to stats of captains.
What do we need to do to be a captain?

You can compare your stats if you were a captain to stats of 1st/2nd ranking bot.
What do we need to do to be the best?

You can also analyze data from captains/1st picks to see if there were mechanisms they have in common.
What makes the best bots the best?

You can use the data to determine strategy for your next match.
How can we win our next match if the data says we should lose?
How can we win our match if one of our partners has a problem?
How would the opponents try to beat our alliance since we are stronger?

You can use the data to pick playoff alliance partners.
Which bot would best compliment our team/alliance?

I did a quick scouting of a team that asked why they weren’t picked for elims as the 14th ranked team in this thread. HELP! Why weren’t we picked?
I had a mix of qualitative and quantitative data, both are needed for scouting. One thing I didn’t record was whether they scored on both sides of the field in a match, something that wouldn’t be seen in the stats but could be of great importance to a team looking for a 1st or 2nd pick.

|Qual 7 C|3 scored cargo||1 scored hatch|||
| — | — | — | — | — | — |
|Qual 12 C|2 scored cargo|4 missed cargo||||
|Qual 20 C|3 scored cargo|||1 dropped hatch|Ineffective defense|
|Qual 25 H|3 scored cargo||1 scored hatch||Scattered cargo, blocked loading station|
|Qual 28 H|3 scored cargo||2 scored hatch|||
|Qual 37 H|||1 scored hatch|2 dropped hatch||
|Qual 45 H|||2 scored hatch|1 dropped hatch||
|Qual 49 H|4 scored cargo||1 scored hatch||1 foul, end on lvl 2|
|Qual 58 C|4 scored cargo||1 scored hatch||1 foul, crossing cargo line in sandstorm, end on lvl 2|
|Qual 62 H|3 scored cargo||1 scored hatch||Scattered cargo, blocked loading station|
|Qual 73 H|1 scored cargo||3 scored hatch|||
|Qual 77 H|3 scored cargo|3 missed cargo|1 scored hatch|1 dropped hatch|1 foul, end on lvl 2|

I looked at your matches on twitch and this is what I could come up with. I mostly didn’t care what level you started on or if you finished on lvl 1.

The three negative things that stand out to me is your missed cargo shots, your scattering cargo in depot which blocks the loading station, and the three fouls. You really didn’t play defense that denied scoring (qual 20), the other two times you tried you got a foul for being outside the frame.

Now for the positive. You could if you tried score 3 cargo consistently. There were also three matches where you scored a hatch panel and trapped a pre-loaded cargo. Doing more of that would make your bot really stand out, as well as a more consistent lvl 2 climb. You would be a good first pick. Your fouls when you played defense prevented you from being a 2nd pick.

You should tell your alliance partners pre-match that you can fill half the cargo ship and will take one side. You should work on putting a hatch panel on a pre-loaded cargo bay in sandstorm, fill three cargo on one side and then try to fill the lower level of a rocket. You have the ability to fill half the cargo ship and at least 1 hatch panel on the rocket. Try to be quicker. Climbing to lvl 2 every match will also get you noticed. Climbing to lvl 3 could make you a captain.

I do realize that the scouting data can be used for more than alliance selection but the various uses discussed above are using the data to work with another team(s) and complement the team(s). While technically not alliance section, it is similar. :grinning:

I was trying to get the scouting students to understand the “why” you scout but my seniors were not interested in learning that. That is not what they have done for the four years. (This is my second year as a mentor but I was clueless last year. :grinning: It was my first tournament last year where I finally understand alliance section.) The students focused on putting together a scouting sheet, pretty much copied from 2018 game, but when I tried to get them to understand that the purpose and use of the data determines what to collect, I didn’t get very far. :smiley: Without re-hashing the long thread about the role of mentors, I looked at my role as advising. If the students didn’t want to listen, I wasn’t going to push the situation.

Since the seniors are graduating, I am trying to take the opportunity to train a new set of scouters. Thus, I want to use the data to explain “why you scout” which determines how you measure this (for example, cycle time) which then determines the data to capture. In our situation, we capture a bunch of data and hope that the data captured, might be useful. That is not the methodology I want my students to understand.

You might be interested in this thread

To get to the “why” look no further than blue banners to start. Then look at the characteristics of those teams.

Get the students interested.

The common problem with students is they come from an instant gratification childhood not their fault its just the world they lived in. So asking a student to scout is work. Now ask your self how good will that data be if a scout who does not like the drudgery of scouting will be? Garbage in garbage out is the answer,

One key fact of life is people do well what they want to do well… scouting is not a draw unless the process of scouting shows results and is highly valued as a first class part of the team. In essense show you can win games with scouting alone. Show you can assemble very strong alliances. Show that you can know at all times who is on the field with you. Show at all times a story of who you are facing and their history in your next competition.

Start with the above then do data analysis on your observations. IMO most scouting apps pull aggregated data that is not specific enough. Also IMO most collected data is tainted by scouts that collect bad data. So only collect data+notes that helps you win or beat teams.

Scouting itself needs to prove itself to the team, it neeeds to be seen as a key part “the team can count on” the driving team is always highly complementary of our scouts becuse they win games with the intelligence they provide them…most of that is not data rather what makes those teams tick on the field. Thats much more than data points alone.

Once again we made elims twice in 2019 this makes 9/10 since our RAS year. Our scouting data/notes won a game against the 1 team this season by using a technique ths scouts found in a game we were vastly outmatched and should not have won by sending a defender to do X (That 1 team finished 2 because of that loss). We started scouting in year two this way basing the scouting on actionable advice. The team believes in scouting and we have plenty of interested scouts because it works and is a key working part of the team and the system works. At the end point I had 26 scouts active out of 40 or so team members by choice. They found it interesting.

I used to be pretty good at horseracing in fact used to bet before work , many of the same ways people analyse horses apply to FRC robot teams. Its not all data drivin. Lots more goes into determining if a horse has a chance in a race event , same for FRC teams each season.sample.pdf (404.4 KB)

The attachment is part of our preliminary excel sheet (without the notes) it gives you an idea of some of the past performance metrics we track going into a competition. The names have been witheld to protect the innocent and the dark greens won the actual event as expected…no surprises

The simple answer for WHY teams scout is:

You scout to improve the performance of your team and your alliance.

I had already bookmarked this thread. :grinning:

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