Week 3 Score Breakdown: Up Up and Away

FIRST General Forum
1

Week 3’s over, and you know what it’s time for, another score breakdown!

The Numbers
From now on, all the numbers below are a direct comparison of the current week and the single week before (i.e. Week 3 vs Week 2). Cumulative match stats are available at the end.

This week, the average match score was 214 points, 23 points higher than week 2. 20% of that score is earned in auto (+0.7%), and 76.6% in Teleop (+0.1%). Only 3.6% of the match score is dedicated to foul points (-0.7%).

In playoff matches, 0.4% of the match score is made of rotor bonuses (+0.2%), and 0.1% of pressure bonuses (+0.1%).

In Autonomous, 30.9% is dedicated to mobility (-3.2%), 65.9% for Rotors (+2.7) and 3.1% (+0.5%) for High Fuel, with Low Fuel crawling in with 0.1%.

In Teleop, Rotor and Takeoff Points are just about tied with 49.4% and 49.1% respectively. Rotors went down by 3.1% and Takeoff Points accounted for 2.5% MORE of the match score. High Fuel climbed to 0.9% (+0.3%) and Low Fuel is being valued less at ~0% (-0.1%).

This week, 75 games (1.1%) were decided by fuel (10 more than last week). 142 games (2.2%) were decided by fouls, reduced by 22 games in comparison to last week.

Something else to note is that 1984 adjustment points were taken from the match scores this week (usually due to red cards). This is 1319 more adjustment points than last week, weighing in at an average of 0.45 points deducted from each game.

The Extras
These values are from the entire season, not just week 3.

A total of 133 (4.5%) alliance seeds have been dependent on ranking points earned by fuel, and 145 (4.9%) alliance seeds from ranking points earned from rotors.

The Opinions
Rotors are counting less, and fuel is counting more, but not by much. A very interesting turn is how Takeoff (climb) Points have increased in value by nearly 2.5%, meaning more teams are starting to climb rather than worry about delivering more rotors and fuel. As this trend continues, I think we’ll start saturating the takeoff points, meaning you either climb and keep up, or you lose.

It’s good to see fouls are decreasing in value, and hopefully this trend continues.

The ranking point situation seems quite dire. Wins, Ties and Losses seem to be the only really important factor, with the bonus ranking point earned from Fuel or Rotors only accounting for ~9.4% of all alliances out of the entire season.

**Code: ** https://github.com/JacisNonsense/frc_2017_statistics
**Week 3 vs Entire Season: ** https://gist.github.com/JacisNonsense/b036ae11e0fcb4aed8dd680b9b0c2c8d
**Season Totals: ** https://gist.github.com/JacisNonsense/6089443a19b7e09401597d72f7b075d1
**Raw Data (comparative to week 2): **
Format: total (avg) +/- total_difference (+/- average difference)

=======================    
      ALL MATCHES          
=======================    
        Match Points:    
                     total: 933823 (213.79)    + 49108   (+ 22.8)
                      auto: 186971 ( 42.80)    + 16582   (+  6.0)
                    teleop: 715461 (163.80)    + 38245   (+ 17.6)
                      foul:  33375 (  7.64)    - 4400    (-  0.5)
                    adjust:  -1984 ( -0.45)    - 1319    (-  0.3)
     Mobility Points:    
                      auto:  57830 ( 13.24)    - 305     (+  0.7)
        Rotor Points:    
                      auto: 123180 ( 28.20)    + 15480   (+  4.9)
                    teleop: 353520 ( 80.93)    - 1720    (+  4.2)
      Takeoff Points:    
                    teleop: 351500 ( 80.47)    + 36000   (+ 12.4)
    Fuel High Points:    
                      auto:   5817 (  1.33)    + 1335    (+  0.4)
                    teleop:   6138 (  1.41)    + 2198    (+  0.6)
     Fuel Low Points:    
                      auto:    144 (  0.03)    + 72      (=  0.0)
                    teleop:    233 (  0.05)    - 282     (-  0.1)
          Foul Count:    
                      foul:    955 (  0.22)    - 85      (=  0.0)
                 tech_foul:   1144 (  0.26)    - 159     (=  0.0)
=======================    
 QUALIFICATION MATCHES     
=======================    
   Bonus Rank Points:    
                     rotor:     42 (  0.01)    + 29      (=  0.0)
                  pressure:     68 (  0.02)    + 54      (=  0.0)
=======================    
    PLAYOFF MATCHES        
=======================    
        Bonus Points:    
                     rotor:   3300 (  3.90)    + 1600    (+  1.9)
                  pressure:    760 (  0.90)    + 460     (+  0.5)
=======================    
    SCORE BREAKDOWN        
=======================    
        Total Points:    
               Match (auto)    :     20.0%    + 0.7     
               Match (teleop)  :     76.6%    + 0.1     
               Match (foul)    :      3.6%    - 0.7     
               Bonus (rotor)   :      0.4%    + 0.2     
               Bonus (pressure):      0.1%    + 0.1     
         Auto Points:    
                       Mobility:     30.9%    - 3.2     
                          Rotor:     65.9%    + 2.7     
                      Fuel High:      3.1%    + 0.5     
                       Fuel Low:      0.1%    + 0.1     
       Teleop Points:    
                          Rotor:     49.4%    - 3.1     
                        Takeoff:     49.1%    + 2.5     
                      Fuel High:      0.9%    + 0.3     
                       Fuel Low:      0.0%    - 0.1     
=======================    
   MISCELLANEOUS DATA      
=======================    
    Wins due to Fuel:       75 (1.1%)    + 10      (+  0.1)
   Wins due to Fouls:      142 (2.2%)    - 22      (-  0.3)
2

One of the most interesting developments, in my opinion, is the scoring of fuel in tele-op beyond 40 kpa, as shown by regional winners 254, 971, and 4990. I don’t see the merit in getting 30 extra fuel beyond 40 in tele-op instead of a third rotor- it’s less points for a riskier bet. If you can just match your opponents’ 3 rotors and get >2 climbs, the 40 kpa +20 bonus already gets you the match. I guess it worked though, so I can’t criticize too much.

3

When averaged out, this is very true, but ranking points played a huge role in seeding for teams that could reliably get 40 kPa. I would say ranking points are essentially the differentiating factor for elite teams, but not something the average team can depend on for a boost like last year.

Thanks for doing these statistics each week!

4

118 scored over 60 Kpa all regional long at lone Star Central. Hit 90+ multiple times

5

This comes down to defense. It’s very hard to defend fuel, and almost no one is practicing how. It’s very easy to defend gears, and everyone can practice how. The gear scoring is “chunky” and if you knock a single rotor out your opponents grasp, the match is yours.

Try to find a video of Finals 2 at San Francisco. 4990 holds a master class on smart gear defense, while 3x gear cycle robots trip over themselves trying to make 4 rotors happen. Meanwhile 254 happily chomps down fuel points all match. 971 wasn’t on the field, but if they were, all they would have needed to do was climb to guarantee victory.

Yes. Getting picked & going into eliminations requires a functional climb. Want to get picked at a district or regional? Climb every time. No fouls by humans or robots. Done.

At SF, it seemed like the refs were going pretty easy on the pilot actions during qualifying. Unsure how much of this is pilots/teams getting their acts together, and how much of this is just easing up from the refing side. Unsticking [velcro-tipped…] ropes from the velcro latches was a huge pain point for all the pilots and a lot of pilots were reaching over the side to get things loose, which I had previously understood as a yellow-card worthy offense?

6

That would be an interesting thing to look at, if someone has the time and inclination to do so. For the teams that climb successfully in all of their matches for a competition, where do they end up? Quarterfinals? Semis? Event winners? All in all, it’s not an extremely useful metric, but it might be interesting.

7

I think the point here was more what it takes to get picked (or do the picking), not what it takes to win the event. The metrics to track here would be to correlate the percentage of successful climbs with whether or not a team played after lunch on closing day. Once you’re in the playoffs, three climbing robots is closer to a necessary condition of victory than a sufficient one.

8

My friend and I noticed a very interesting thing while watching San Fransisco. If your alliance has any hope of getting 40kpa, then it is almost always more worth it to get the kpa and have your third bot play defense than it is to get the third rotor (especially in playoff rounds). The only time this doesn’t work is if the other alliance can get 4 rotors, but that is still a rare occurrance. If this strategy doesn’t become more popular in the next couple weeks, I will be surprised.

9

S07. Keep your limbs safe at all times. During the MATCH, the PILOT may neither…
D. contact any part of a deployed (i.e. any part of the ROPE is below the deck of the AIRSHIP) ROPE.
Violation: YELLOW CARD (emphasis mine)

10

Currently there is no way to know if a team climbed or not, we can only know how many robots climbed per alliance match (i.e. 0, 1, 2 or 3), as touchpads are not bound to a particular alliance station

11