TL;DR: The data from FiM and NEF DCMP show that their are a lot of high number teams breaking into FIRST’s top tier as measured by participation in DCMP event and the DCMP Elims, in both districts not just in FiM as I had initially thought. This is a good thing. I, for one, welcome to our new robot overlords 
All,
Let me just bare my Michigan biases right from the start. I was looking at that FiMDCMP results and I was blown away. SO MANY new teams in the elims. I was all set to pat FiM on the back for a nice job developing their talent.
My friends back in Michigan are engineers so I don’t just complement them, I send data if I can manage it. My first attempt looked promising (FiMQuartileTeamNumber2016.jpg attached - first attachment below also here). Looking at that chart, I think it is amazing that the 50%tile team number (median team number) for FiM District is 5067, 3575 at the FiMDCMP tournament, and 2966 for the ELIMS of FiMDCMP.
I was about to hit send on a congratulatory e-mail to some friends back in Michigan when I recalled the words of the emcee at NEFDCMP saying how FiM gets all the accolades but it was NEFCDMP that held (and still holds) the World highscore for Stronghold. All we ever hear, he said, was FiM, FiM, FiM…
So, I decided to do a little data mining to see if FiM is all that different from other districts in this regard or not.
My first cut was to generate FiMAndNEFQuartileTeamNumber2016.jpg (second attachment below also here). My first look at that chart (again with my FiM bias fully showing) made me think, Yes, FiM is killing it; the FiM 50%tile charts are above the 75% charts for NEF for the District as a whole, the DCMP, and the DCMPElims. The same can be said of the FiM25% compared to the NEF50%. And there is nothing to compare the FiM 75% to. It’s all alone on the top of the graph.
Seems cut and dried but then I started thinking about it. The populations of the Districts are very different. If you have 99% new teams in one district and 99% experienced teams in another, then you’d expect that you the first district would have more new teams and the second in any subgroup of teams you select.
I decided that the real comparison would be to see where the teams that make it to DCMP and DCMPElims compare to the population of their associated district. PercentRank to the rescue. I took those 25%tile, 50%tile, & 75%tile teams that played in the DCMP and the DCMPElims and computed where those teams would be in the District population (using PercentRank to normalize for differing numbers of teams in the two districts).
The result is FiMAndNEFQuartileTeamDistrictPercentRank2016.jpg (third attachment below also here). The graphs for NEF and FiM are virtually the same. I haven’t done the other districts, but I expect they follow the same general trend (I have attached the spread sheet – someone can feel free to do the others).
My BOTTOM LINE? FIRST’s Top Tier is continually renewing itself.
Yes there are a lot of good teams in there with low numbers but it isn’t ONLY those low numbers. Many new teams are coming into the fold. In NEF the 75%tile team playing in the DCMP was a team number in the mid 3000s started in 2011 – that means a quarter of the teams at NEFDCMP where rookies in 2011 or after. For FiM, the 75%tile team had a number just over 5000 whose rookie season was 2014 – a quarter of teams at FiMDCMP playing in their 3rd season or fewer. That is amazing to me.
I, for one, welcome to our new robot overlords
Cheers,
Dr. Joe J.
FiMvNEFTeamCompositionComparison2016.xls (121 KB)
FiMvNEFTeamCompositionComparison2016.xls (121 KB)