paper: 2015 CMP misc data and analysis

FIRST General Forum
#1

**

http://www.chiefdelphi.com/media/papers/3132

76 teams, 127 matches, ARCHIMEDES
75 teams, 125 matches, CARSON
76 teams, 127 matches, CARVER
76 teams, 127 matches, CURIE
76 teams, 127 matches, GALILEO
76 teams, 127 matches, HOPPER
76 teams, 127 matches, NEWTON
76 teams, 127 matches, TESLA

8 Subdivisions, 1014 matches, 607 teams

CMP Qual Match Alliance Score Comparison presents a percentile plot comparing the qual match alliance scores of the 8 subdivisions.

[Subdivision Partial OPRs](Subdivision Partial OPRs) provides team OPR data computed from the qual match final scores in the subdivisions to allow you to analyze how team OPRs changed over the course of qual matches. It includes an example plot for the Newton subdivision.

**

#2

**
posted:

CMP subdivision Qual Match final score OPR percentiles plot

CMP subdivision Qual Match auto score OPR percentiles plot

CMP subdivision Qual Match component OPRs and stats

http://www.chiefdelphi.com/media/papers/3132

**

#3

I’m a little confused at the autonomous OPRs in the plot. I would expect them to generally fall in the range 0 < x < 20.

#4

Yeah, I graphed the wrong column. That was the team average score, not auto.

That’s what I get for doing this when I’m half awake.

Thanks for the heads-up. [strike]I’ll fix it.[/strike] It’s been fixed.

#5

**
posted:

CMP Playoffs Alliance Score Percentile plot

http://www.chiefdelphi.com/media/papers/3132

**

#6

**
posted:

Ax-b residuals percentile analysis (plus example plot) weeks 1-7 plus CMP subdivisions. Compare qual match actual alliance scores to OPR-computed alliance scores for all 117 events.

http://www.chiefdelphi.com/media/papers/3132

**

#7

**
posted:

Ax-b residuals statistics for all qual matches in 2015: 117 events (weeks 1-7 plus 8 CMP subdivisions), 9878 qual matches, 2872 teams.

http://www.chiefdelphi.com/media/papers/3132

**

#8

**
posted:

Weeks 1-7 plus CMP Qual Match Alliance Final Scores percentile data ready for plotting.

http://www.chiefdelphi.com/media/papers/3132

**

#9