pre scouting

what can i do to pre scout regionals or worlds

A pretty basic prescouting system would involve a Google Drive Spreadsheet of teams, as well as different metrics you would like to keep a track of. If you have multiple scouts helping you, it is simple to assign blocks of teams to different kids and have them fill out the spreadsheet based on previous matches.

Personally, I do back scouting as I find out which teams are registered for my district events. I make a a spreadsheet with all of the teams, and as build season continues I try to keep updated on all of those teams, keeping in mind the parameters I find important for scouting. Excel and Google Spreadsheets are good for this, but you can also make a folder with individual documents for pictures and stuff, too.

For example, if u have an event in week 4. U might want to check if the teams in your Regional/District already played in week 1,2 or 3.
And how they performed there and what there strategy is.

What I do for pre scouting is make my own scouting sheet using ed law’s scouting data base (reply if you need a link). I copy his big world rank tab to an excel file. Then, you could either use copy and paste or use the “vlookup” function to sort and find the data for that team. His sheet for this year broke opr into 5 components. Tote, Can, Litter, Coop, and Foul. Then, you can use the rank function to make a ranking system. This can tell you what the strengths of robots are. For example, our team was really fast with totes but were usually bad with the cans. That was reflected in our opr data.

Just remember that opr can be messed up based off of how teams schedule’s were. The better the schedule a team gets, the better their stats look. This typically is only truly effective for the top 10 teams of an event but it helps. I made a pre scouting sheet for the Archimedes Division using that system I explained above. If you want, I can post that sheet on here. I also did a (Different type of style but possibly better) pre scouting sheet for IRI which was a lot of fun.

I would (in order of most to least valuable):

  1. Watch videos of teams from their previous event(s)
  2. Acquire scouting data/notes from other teams at their previous event(s)
  3. Find out where they were selected (if at all) in alliance selection
  4. Look at stats such as OPR, w/l record…
  5. See how teams did in previous years
  6. Look at reveal video/pictures

Some really good resources you can use are blue alliance website and the CD forum for showing casing your robot. However you do need to be careful when looking at reveal videos for other teams robots. These are usually under controlled circumstances and may not be how it can perform in competition.

The constant is always to watch a lot of videos on TBA. That being said, I don’t pay a whole lot of attention to events other than a team’s most recent event. Pre-scouting is also only good until you have a couple matches of at-event scouting data, unless a team is working out issues.

With the ability to do component OPRs on different outputs from the FMS, getting a better picture of how a team is getting its points at events you don’t attend is going to get even better (for some games, anyway).

may i please have the link to ed law scouting data

Scroll to the bottom of the original. You should see a paper called 2015 championship scouting database v2

Our team is working on a pre-scout programy that fetches data from TBA. We might release it later on. My advice is to make a spreadsheet. Use the championship data and data from TBA.

Remember that there is no better alternative to seeing a robot live or in video. If you can watch match video of the teams that you are scouting, that is the best option if it is available. Otherwise, it would be wise to check stats like OPR as they are an indicator, albeit a fairly poor one. If you have no access to team stats or video (or if they haven’t competed yet) you can watch other teams and events as they will give you an idea of how strategies have been playing out (especially if your team hasn’t competed yet). Remember: strategy wins matches, so you not only need to know about the other teams at an event but how the game dynamic will be played.

Prescouting in a nutshell:

  1. Figure out where the teams attending your event(s) are competing prior to said event(s), if at all.
  2. Find webcast and match schedule of said prior events. (Being there in person is better, but that often isn’t going to work out.)
  3. Sit down with your existing scouting system and hope the webcast(s) are decent. Plan to stay all day (if you don’t have other, more important things–like school). You’re also testing your own scouting system here.
  4. Wait a day or two and check Youtube and TBA for videos of matches you want a better view of (probably most of the ones with teams you’re interested in).
  5. Find reveal videos of anybody who didn’t compete.
  6. Pit scout all day Thursday…
  7. …and throw out all the prescouting for several teams on Friday morning because they completely changed their robot between the event you prescouted and the event you’re playing against them in. :yikes:

Prescouting regionals can be difficult, but there is a significant correlation of previous years’ OPRs with the current year, so you can start with that data. In some years, we’ve run regressions on component OPRs against our scouting data. IN 2014, we used that adjusted OPR data in our scouting system as “Match 0” to seed our drive team scouting data. By the end of the competition it was less than 10% of our scouting data.

The last 2 years, we’ve watched videos of the teams that are on our alliances during the flights to St. Louis. In 2014, we picked up the 2 rookie teams that we could assist the most. The value of the videos isn’t looking for your strongest opponents but rather your weakest alliance members. You have limited resources so use the info you dig up to use those resources most effectively. For us it was deploying our mentors to help other teams.

Almost exactly what we do. It’s easy and one sheet can be used for many events.