Gael Force Scouting Database Opinions

Hi guys! I’m Kush Rastogi, Deputy President of Gael Force Robotics. As you can see from previous related posts here , our club has been working on developing an improved online scouting database that pools the resources of Robot Events to create an integrated scouting environment for teams.

One of the features we plan to include is team-related statistics. These would include multiple values, such as win percentage, OPR, CCWM, etc. to give teams multiple items to rank and compare teams according to their own standards, which is why we will include a variety of information that teams can go through.

One “rank” that we were discussing is a data point that takes the relative average of all the performance related statistics and displays them along with the other information. One way to do this is after some discussion, give some basic weightage to certain statistics. Another way to calculate this rank is to make every number relative: for each statistic, we would find the team’s relative score by dividing their score by the highest score of any team. This would be adjusted if lower is better for a statistic. In all, this rank would end up ranking teams that have the best all-around statistics on top.

One view is that this rank does not make sense mathematically and should not be included with any of the other ranks. This view proposes that other teams may assume that this number is just telling them who the best team is, when really, it’s a misleading number that represents nothing useful since everyone has different opinions, and may lead teams to make ill-advised decisions during a match.

Another view is that this number is meant for more passive comparison rather than active comparison during a match. The intent of this rank would be to display which team has the most hefty resume (of course, this does not mean that the team with the highest rank will necessarily be the best robot). If teams do not believe this statistic is useful, they have every right to ignore it.

We are wondering what you guys think of this rank. Please respond, telling us whether or not you think we should include this rank on our scouting website. Our decision will be based entirely on your thoughts.

I think that the best indicator of robot quality over multiple tournaments is definitely win percentage. However, if you guys already have this rank figured out, it probably isn’t that hard to implement, so go for it. It would be interesting to have a custom ranking capability where you can manually adjust the weights.

I think you guys should keep the comprehensive ranking and just put a disclaimer on it. As your poll reflects right now (4 keep it, 5 don’t include it) some teams would use it and other teams would not, but so long as teams are using it I think that it would be worth putting out there.

Back to that previous thread, I had a couple more ideas for statistics. Drop me a pm if you’re interested.