So Scout6088 here from FRC Team 6088. My team started in 2016 with first Stronghold. We started scouting this last game year, 2018, with very minimal scouting information.
I know that we don’t have information on the game to start getting scouting ideas on, but my team is small, so we need almost all of our members to help scout the matches.
The thing is, no one on my team wants to help scout. No one else is on the official “scouting team” I’m the only person working on ways to scout and what to do with the information after we have scouted it.
Does anyone have any tips to make it easier, I dug through multiple programs for scouting and figured out that with my paper scout sheets it was much easier just to work with it in paper and organize it roughly on a google doc.
The other thing is pit scouting. We did pit scouting, but it didn’t work very well and I didn’t give a second glance at the pit scouting sheets. If someone could give me some tips for what to do with the pit scouting sheets, and the information on them that would be great.
So if anyone could give me some tips on how to make scouting easier on my team and on me that would be great too!
Dont overly force it. theres ways to convince them to do it (Look into Simbucks from 1114) but you dont want to make them go all event just scouting. This was a key downfall to our scouting in 2017.
Results will make people interested. If you can show that scouting is going to get you to elims/win events, theyre going to want to scout more to keep winning.
Look to join forces with another team to help lessen the load. Sometimes this can come back to bite if you rely too much, but having another group to help split the load can be a huge help
As for pit scouting, its good if you know what you are looking for from it. For most people, there really isint much use besides talking to other teams (which is important to be clear) but the data it gives is mostly not useful for teams.
As for making it easier, im gonna be a shill and say look into Tableau. You get it free every year with the Virtual KOP and its a great data visualizer tool that can help show where teams excel.
Looking at the OP here are some ideas for the issues you currently have with your scouting program:
**Issue 1: The team is very small
**This can be a real struggle because you ideally want at least one person scouting each robot in a match. If that’s not possible there are a few things you can do. First you can reach out to other teams attending the same competition as you and see if any of them would like to scout together. This can benefit both teams, and it works really well a lot of the time. Make sure to let them know how many people your team has available to scout if you do this.
Another option is to scout specific teams in each match. If you want to have data for your quals matches, you can only scout the teams in those matches. Priority goes to the teams in earlier matches. I’ve done this at many offseasons and it works very well to become prepared for match strategy.
**Issue 2: No one wants to scout
**Unfortunately this is often the case. There are a lot of other threads on chief delphi that give advice on this issue though, so feel free to look there for ideas. Personally, I found that when the team leadership stressed the importance of scouting, it was much easier to get students dedicated to scouting and collecting accurate data. You need to make sure everyone knows the importance of scouting going into a competition, and they need to know how it helped during the competition. Make sure the students don’t think scouting is just a role to keep them busy. Make them feel crucial to the team, because scouting really is that important.
**Issue 3: Pit Scouting
**This topic is somewhat debatable. I personally have found that pit scouting is rarely important. The most important data you can learn about a team is almost always knowing what they can actually do. Never ask pit scouting questions that you could answer by watching their robot in a match. With a small team, pit scouting becomes even less important because you don’t have the resources to dedicate to it. Pit scouting can occasionally be very helpful, but I would argue that it isn’t worth the resources if it compromises your match scouting.
My suggestion, in the organization of the data, do the paper scouting you are doing, but quantify certain factors in the game. My team labelled the autos for 2018 by number, with 0 being nothing, 1 being movement, 2 being 1 cube in switch, 3 being 2+ cubes in switch, 4 was 1 cube in scale, and 5 was 2+ cubes in scale. we then looked for the autos we wanted and which had a higher mode for it.
For general scouting advice, I’ll recommend this post I made a while ago about general motiviation tips for scouting. Obviously, since it seems that you may come from a smaller team, this will have to be scaled down a bit.
Depending on next year’s game, it may be possible to have one or two scouts watching an entire alliance instead of one scout per robot. This is a little more taxing, but allows for a smaller amount of students to be scouting at one time. If you do choose to go this route, make sure it fits reasonably into the game and that the scouts get plenty of practice beforehand to allow for more accurate data.
I want to provide a different take than what has been expressed so far.
First thing to remember is that the end goal of scouting is come up a reasonable guess of what each robot will be capable of doing in eliminations. There are more ways to get there other than extrapolating a full data set of statistics from the qual matches. My suggestion would be to figuring out how to get enough data to make a pick list with the man power you have before the event and concentrate on that data.
In 2018 you could put together a reasonable draft list if all you had was:
an estimate of the number of cubes a robot could place in the vault
an estimate of how many cubes a robot could place in the scale
an estimate on how likely a robot was to climb
an estimate on how capable/susceptible a robot was to defense.
It wouldn’t be a perfect list but it would be enough to make a reasonable list. For a 40 team event, that’s only 160 answers that you need to arrive at. That’s something one or two people are capable of doing.
Where pit scouting comes in is that it is a quick, accurate way to fill in the zeroes. If a team says they can’t climb, they can’t climb. If they have an omni-drive, they can’t play defense. If they say they can’t do the scale, they can’t do the scale. The key is to keep your list of questions short and to things that can be verified in the pits. Depending on the event, pit scouting can cut your scouting load in half.
If things go well, hopefully more members of the team will see the value added by scouting and want to do more to improve the process as your team moves forward.
Definitely want to stress how essential scouting is to your team. If you can get your members to respect scouting, they will be much more willing to do it if you ask them. You need to establish a respectful and trusting relationship between you and your scouts. Another good way to make sure people don’t get too bored while scouting is to make a scouting schedule. This informs your team members in advance so they understand their responsibilities, and it holds them accountable to their scouting slots.
For teams that are relatively new to scouting, I recommend starting through Google Forms. You can quickly set it up, and the interface makes it easy for people who are new to scouting to understand without much training. You should definitely move on to other programs later, but this is a good way to start and peak interest in scouting.
We are a small team as well and have always struggled with our scouting.
We had great luck this year crowd sourcing our scouting. We had some amazing programmers on the team who wrote a scouting app that would run on smartphones that would push data to a database server that would compile the raw data.
We would reach out to teams before events and solicit other teams that wanted to join our scouting alliance. We would share the raw data with each team that contributed scouting personnel to the effort. Generally, we would get 2 or 3 other smaller teams to join us. They could download the app onto their phones at the beginning of the tournament. It was set up so that the scouts would not necessarily need to sit in the same location in order to work together. The app would indicate which robot had been claimed for each match and allow users to claim the remaining ones.
We could also load pit scouting data into the same database.
We could push the data out to the participating teams in a simple tabulated .CSV file that could be read into excel or other programs for sorting, ranking or other analysis.
On my team a few people slowly gave up and sort of said they’d done enough and barely any people were doing it by the end of the regional. One of the reasons was we didn’t have a very clear idea of what information was being collected, only general ratings. Having very clear and data based scouting, as in how many cubes in scale, can help keep people focused to some extent. At champs, we worked with another team and everyone focused because they didn’t want to let down the other team, and they the system they had in place had very clear directions for data input.
This season I was planning to put in place a survey sheet using epicollect 5 (https://five.epicollect.net/), which I found allowed for easy, free set up, and easy data access. I took me 5 minutes to figure out, and another 5 minutes to make what I wanted. It also allows for offline data collection for a later upload. It sets up like a survey, and it pops up the questions 1 by 1, making it very clear what people need to put in. There are also a lot of apps made by FRC teams of varying quality which are worth looking into. Having a good system can really help.
How long before each match were robots claimed? Did you indicate that team x would be scouting Blue 1, 2 and 3 for Q30-60, or did people claim a robot right before the start of a match? Asking because I’d assume the app would be PWA-esque, as in it would still function without an active internet connection.
The blue alliance has a very good API that can be used to programmatically get data before and even at competitions. Obviously, this is usually unfeasible at the first competition your team goes too but it worked pretty well for my team. Below is a google drive link to the python code we used (partially) for scouting in 2018. Best of luck.
The match schedule was pre-loaded. But the teams for a given match would not show up in the app to be claimed until the previous match had finished. The scouting was not pre-assigned, so once the match became current, any of the scouts could claim any of the robots. At one point the programmers talked about excluding the scouting member’s own team (you can’t scout yourself) but I am not sure whether they implemented that or not. In the end we did get scouting data on our own robot which turned out to be very helpful.
I’ve scoured the land, and I’ve found mention of these Simbucks, but no real explanation. Could somebody pleas provide a more thorough clarification of Simbucks?
Basically, you give each member X amount (lets say 100 simbucks) to each scout. For every match, someone comes up with some stat that if it happens, people who bet Y back get Z amount (if all 6 bots cross the line, you get double back.) Its a good way to get people to pay attention to the bots and remember who does what as theyre gonna have to know the stats (and the person making the bet itself also) to know which ones are good bets and bad bets to take.
To begin with no matter how large or small a team is never determines how good of scouting data can be received. The only factor that matters is the student’s drive to be successful. If they don’t care about winning and are fine with dancing to cotton eye joe the whole comp, then fine you will never be able to convince them to sit down and write numbers on paper for hours on end. The fact is if they do care about winning then it is crucial to communicate that the best way to improve in ranking and match performance is good scouting. OP if you would like to understand how to do this best for your team feel free to dm me.