This page is a work in progress. Spinning it up to get information to our friends over at Altair.

FIRST expects all teams to develop their own strategies, data models, and scouting techniques. We need to discuss where the line is crossed between helpful tools and excessively helpful tools. As we explore Altair's software advantages and apply it to our strategies, we need to be mindful of making certain things easier and more understandable (visualization, consolidation, organization, data management principles), while not handing over a total solution to every team. Teams should be able to maneuver within the toolset and develop unique strategies that give them a winning edge. I think we can figure out where that fuzzy line is. Right now, FIRST hands out licenses to Tableau, which can be a nice visualization tool, but I think, in this case, a software license is an example of not being helpful enough. We have to keep small teams in mind that are harvesting from a smaller pool of skillsets. Please checkout EveryBot which illustrates the correct balance. The everybot will get a team rolling but does not hand them the win.

• PreSeason
  • Identify pros and cons of past season:
    • data choices
    • gathering choices
    • analysis choices
  • Develop technologies for
    • data collection
      • As a means for obtaining important data
      • As a means to develop team spirit, cooperation, communication skills, and overall fun
    • data consolidation
    • Reporting
• Season Kick Off - GAME has been Revealed - Many weeks before actual competitions
  • Develop initial strategies, data points, collection methods
  • Develop methods to simulate competition and adjust
  • Compete as soon as possible with other teams and attend a pre-season competition
• Competitions begin
  • Begin collecting data available on all competitions and teams.
  • Begin analyzing data and comparing against tournament results
• At Competition
  • Historical data vs competition data - weighted on spectrum with historical being more important at the beginning of the competition
  • Reports for Strategy and Drive Team ahead of every match
  • Post-mortem Match Reports
  • Mid competition reports
    • First Pick Robots - solid
    • Second Pick Robots - solid
    • First Pick Robots - diamonds in the rough
    • Second Pick Robots - diamonds in the rough
  • Mid competition - pick list scouting / data collection
  • Nearing Finals
    • Solid FIRST Pick List
    • Solid SECOND Pick List
    • Pick Lists informative enough for third pick

• Pre 2021
  • Match and Robot data was collected via students in the stands watching individual robots every match
  • Extra data was collected via pit scouting
  • Data was consolidated in a Student Developed Application
  • Visualization was limited to what the students predicted they needed ahead of competition
  • Pick Lists were still compiled via meeting.
• 2021 - 2022
  • Absent solid scouting team / strategy - comprehensive stand scouting abandoned for the following
  • Poll of BlueAlliance data every 6 minutes for all competitions in Michigan for every Team
    • This gave us a unique advantage by being able to see a teams progression over time
  • Scouting data was gathered via Slack Application available to all students and mentors
  • Reports were generated on demand for drive team for their matches
  • Pick List algorithm created and adjusted on the fly by Worthing

2021-2022 Successes:

• Picklists were good, but not excellent
• Students enjoyed scouting more via the less rigid Slack App
• Drive team was never surprised in their matches, which is a good rating for our reporting

2021-2022 Failures:

• Analysis and Development not student driven
• Picklists did not comprehensively identify our diamond in the rough robots
• Data analysis was non complicated and non sophisticated so assuredly it was lacking
• Data points were not effectively quantified and used productively in analysis
• No algorithms for learning patterns (artificial intelligence)
• Poor visualization in reports
• Dependence on connectivity caused major problems
• Data not effectively scrubbed for outlying, incorrect data
• No system for evolving data collection throughout the competition - we need to be able to change our model on the fly

Data collected for Deep Space

• States Match Scouting CSV
• States Pit Scouting CSV
• Worlds Scouting CSV

2022 Competition polling. The data in this sqlite database represents polling of thebluealliance every 6 seconds, it has some gaps due to unforeseen bugs.

• development.sqlite3