Data Analytics on Winning Teams

Updated: Apr 11, 2019

A study was conducted on robotics competitions of recent years to find whether or not there was a correlation between the time since a team's establishment (team age) and winning (awards and ranking).


DATA:

The data was pulled from a GitHub repository organized by The Blue Alliance, an organization that compiles team information, match videos and results from the FIRST Robotics Competition. The years observed were 2017 and 2018 since previous years contained inconsistent data fields.

https://github.com/the-blue-alliance/the-blue-alliance-data/tree/master/events


TOOLS:

To analyze the data, we used Python 3 and the Pandas Dataframe Library, a free software library written for the Python programming language for data manipulation and analysis. It offers data structures and operations for manipulating numerical tables and time series. The primary integrated development environment (IDE) used for this project was Jupyter Notebook, an open-source web application that allows users to create and share documents that contain live code, equations, visualizations and narrative text. Also used were software from the Anaconda program, a free and open-source distribution of the Python and R programming languages for scientific computing, that aims to simplify package management and deployment.


METHODS:

The general basis of the analysis was to find how many teams were younger than the winning team. Then by comparing that number to the amount of teams that competed in the competition (not including the winning team), we could see on average, what percentage of competing teams the winning team was older than. Age was a simply numerical comparison as it's denoted by the team number. The lower the FRC number, the older they are (they were created first). Although our focus was the big picture, we looked at each individual award as well as ranking for more specific data.


CODE:

Shown below are the scripts used to find the teams that won awards and/or ranked 1st. An in-depth analysis of the code will be included in a separate post.

Script used to analyze awards:

Script used to analyze rankings:


RESULTS:

Our findings showed a strong positive correlation between team age and winning. Generally, the older teams won more awards. Since "older" may be a subjective term, in this experiment "older" denotes a team that is older than at least 50% of competing teams. For example, teams that ranked 1st in competitions in 2018 were typically older than 65.98% of competing teams while teams that ranked 1st in competitions in 2017 were typically older than 66.7% of competing teams. One obvious outlier is that teams who won "rookie" awards were generally younger not older. The data is quite consistent throughout the years. The specific percentages for teams that won awards are displayed below in the form of tables and graphs:




APPLICATIONS:

This experiment was conducted to gain a better understanding on the qualities of a winning team. In 2017, Team Optix #3749 was older than 56.83% of competing teams and in 2018, Team Optix #3749 was older than 60.93% of competing teams. Additionally, we hope to further this kind of research by applying this same idea on other factors such as the amount of funding that teams receive. We hope that this project spreads and showcases the applications of data analytics and data science in robotics and we hope that our research will be helpful for the community.


This project was conducted by Andrew Liu with coding assistance from Kevin Cruise, President of the Del Norte Computer Science Club, and Maxwell Chang, Electronics for Team Optix, as well as domain assistance from Akul Arora, President of Team Optix.

77 views

Join the robotics community.

arrow&v

DNHS FIRST Robotics Team 3749

  • White Instagram Icon
  • LinkedIn
  • White Twitter Icon
  • Facebook Clean

© 2018 by Team 3749