Data Quality First

Most of our scouting decisions are made around having the best possible data.

No Data is better than Bad Data

Bad data gives you incorrect insights that you can be (incorrectly) confident in. We would much rather have no data and know that we are lost than have data, think it’s good, but have it be vastly incorrect.

Only Scout what you Need

4476 has a very minimalist approach to the number of things we scout broadly. If it doesn’t apply to every team, target that scouting for the teams that need it. There is a very high correlation to the lower number of data points that a team has to take live during a match, and the quality of data. Even though one extra data point may not in theory make or break your data quality, it will effect the end result and it is a slippery slope.

Validate Your Data

We strive to run as much data validation as possible on our data to make sure that the data is actually correct!

Live Access To Data Whenever Possible

Having instant access to up to date data in a seamless way is massively useful to help plan matches. This is one of the main reasons why we have built a website viewer for all our data so that it can be accessed by a variety of devices from anywhere. This assists a lot with last minute changes in plans while in the queue line and making sure we are making informed decisions.

Brutally Honest and Descriptive Comments

Comments are a fun thing to do with our team, adding funny lines about teams performance good and bad help make matches memorable when reflecting back on comments. It’s also important that the comment also has good content behind it. Comments should not be generally describing quantitative observations.

Comments are meant for Internal Use Only

Our scouting comments are to help us objectively evaluate the performances of all robots at an event. They are not intended to be feedback to the team we are scouting.

If it isn’t Universally Applicable put it in a Comment

Reducing fields of scouting can be limited by having better comment searchability/readability and quality.