IRI Endorsements / Support

If you want to send a specific endorsement / data for a team that has applied to IRI that you think could be overlooked, those submissions are welcome. They help us make better decisions.

Please be specific and state reasons for your endorsement (beyond "awesome, really nice, etc. :smiley: ). We work with data.

send to pc@indianaroboticsinvitational.org.

How does one quantify the performance of a defensive robot? Number of shots blocked/prevented? Amount of experience playing defense? Number of penalties committed while playing defense?

It would seem to me that the most logical way to evaluate a defensive robot is simply by watching matches. I guess this could be simplified down to draft position i.e. the higher they were drafted (at higher levels of play - district champs/champs) the better defender they are?

DPR is even worse than OPR.

Very curious to see how the selection committee views defense for more than just personal reasons.

Would it be helpful for a team to send you archived scouting data from our events, without endorsing a specific team? We’d comment it up so that it would be as readable as possible.

No need to quantify defense, as most very good offensive robots could play very very good defense if need be. A “defense only” robot will never be more than a 3rd pick…when all the good teams on both sides of the field are gone.

I can’t completely agree; however, I do know that it was difficult for my team to compile useful scouting data on defensive robots.

In Newton, we saw three robots that played very effective (read: annoying) defense – 5150, 48, and 5499. We had seen what the Hybrid Hornets can do at MSC, where their defense came very close to stopping an extremely powerful alliance in the Finals. 48’s legendary defense continued this year, played up to and perhaps beyond the limit most referees will tolerate – just ask 16! (I wonder if 5150 picked them just to avoid playing against them?) 5499 played great shut-down D in qualifications against several high shooting teams, including Theory 6, and my own Average Joes. I was surprised that they were not picked.

None of the above is easy to quantify with simple tablet-clicks.

I hope yah don’t fah-get the numbah 1 team, Team 1254!

  • Let the defense robot be Team D.
  • Identify matches where Team D played targeted and strategic D for the majority of the match; let the teams defended by Team D be Teams O_1, O_2,… O_n.
  • Look at the average scouted points per match of team O_k. Compare this with the points scored in the match where Team D defended O_k.

Basically, you’re statistically evaluating the impact Team D had on the teams they defended. Obviously not a perfect metric, but it at least couples some numbers to the eyeball test. The eyeball test can be notoriously weak when evaluating defense.

I really like this method. I had never thought to perform this kind of analysis. Thanks Karthik!