Overview: To best utilize the Scouting App, the raw data files can processed as a .csv file for importing into an Excel spreadsheet and also written out as HTML files for browsing on the local computer. The HTML is all saved on the local computer, as FTC venues typically don’t have reliable wifi and do not allow independent wifi hotspots or servers to be set up. Our experience is that it works best to have a laptop in the pits dedicated to aggregating and displaying the data for analysis and alliance selection assistance.
Two Data Output Formats: The scripts as they exist currently output data in two formats. Delta Robotics currently only uses the second output format, but the scripts still create the first format, just in case someone wants to use the data out of Excel.
- Export raw data to a .csv text file for easy importing into Excel (or other spreadsheet program) which can then be used for free-form data analysis and manipulation.
- Export data into HTML files for browsing with a web browser off the local computer. This format is comprehensive in that it compares each team against others, and ranks them based on “estimated” points. This estimated value is not the point total that the team alliance may have scored, but is an attempt to characterize the contribution the individual team made towards the final alliance score. Therefore these numbers should only be used for comparison and not actual points scored.
Installation: The following steps/tools should be installed in order to run the data processing tools:
- Download the Scouting2017_18_v1.1.zip installation files (approximately 39 meg).
- Included tools are based on installing on an x64 Windows image of Windows-7 or higher. By downloading and installing different versions of these tools you may be able to get these to run on other OS configurations, but that is up to you.
- Unzip ScoutingAppTools.zip into a directory.
- Install Tools
- Note: You probably will need to be running as admin in order to install some of these tools *.
- * packages/vc_redist.x64.exe – needed for PHP to function
- packages/php-7.1.0-Win32-VC14-x64.zip -unzip and rename to “php”. Then move to the “Windows(c:)” directory and you are done!
- Once tools are installed, you should just be able to put new data files from the scouting app into the ./data/ directory, then run “runit.bat” file to process data.
- Data files on your android device can be found in the following directory, and by connecting your device up to your computer via USB you should be able to easily navigate to the directory and move the .json data files to your computer’s data directory within the scouting folder you have unzipped the above files into :
- /Android/data/org.deltaroboticsftc.relicrecovery17_18/files/MatchData/DR-2015-Official – Relic Recovery
Data Analysis: The HTML output files try do some comparisons between teams and identify which scoring elements are high-scoring for a team and in which areas they do not score highly.
- The data displays both the estimated scoring contribution for each type of scoring opportunity, as well as the number of times the team scored that way.
- File naming standard for the output file from the scouting app is <teamNumber>-Match<matchNumber>.json. There is a dummy team 0000-Match0.json file in the ./data/ directory of the tools zip file as an example. Once match files start getting populated into the ./data/ directory, the dummy file can safely be removed.
- Once team scores start getting pulled in, the analysis will show various information as identified by colors and bold numbers.
- Within each scoring box, the smaller number in parenthesis is the number of times that element was scored by the team, and the larger number is the points scored.
- “Estimated Average Scoring Contribution” is an average of how the specific team actually scored across their matches, and the standard deviation value gives insight into how consistent a team was in scoring.
- A team’s detailed scoring page will show at a high level of detail how a team performed in each match.
- Note: The “Estimated Average Scoring Contribution” value is not an actual score, but what is estimated that the scored team might be able to contribute to an alliance, based on their actual observed performances.Utilizing these numbers can give insight into which team(s) may make for a good alliance partner to find areas where team strengths can be complimentary to each other.