Chief Delphi

Chief Delphi (http://www.chiefdelphi.com/forums/index.php)
-   General Forum (http://www.chiefdelphi.com/forums/forumdisplay.php?f=16)
-   -   Week--1 All regionals OPR (http://www.chiefdelphi.com/forums/showthread.php?t=104208)

MrForbes 05-03-2012 20:14

Re: Week--1 All regionals OPR
 
To validate the Hybrid numbers...yes, 1726's ranking looks reasonable.

cgmv123 05-03-2012 22:46

Re: Week--1 All regionals OPR
 
1 Attachment(s)
Quote:

Originally Posted by Bongle (Post 1139703)
It would be pretty cool to make scatter plots of the winning alliances (bridge power on x axis, teleop on another for example) and see if patterns emerge.

I don't think this is exactly what you are thinking of, but the radar plots sure do look cool. :D

Note that there are 2 charts on the "All Teams" sheet and 3 on the "Alliances as Whole" Sheet

Obviously, teleoperated baskets account for most of the points, though it is interesting how the Hybrid OPR is higher than the Bridge OPR for all teams and when you take just the winning alliance members. (1.765 vs. 1.506 & 3.968 vs. 3.896)

Attachment 12216

Wayne TenBrink 05-03-2012 23:14

Re: Week--1 All regionals OPR
 
Quote:

Originally Posted by cgmv123 (Post 1139853)
I don't think this is exactly what you are thinking of, but the radar plots sure do look cool. :D

Note that there are 2 charts on the "All Teams" sheet and 3 on the "Alliances as Whole" Sheet

Obviously, teleoperated baskets account for most of the points, though it is interesting how the Hybrid OPR is higher than the Bridge OPR for all teams and when you take just the winning alliance members. (1.765 vs. 1.506 & 3.968 vs. 3.896)

Attachment 12216

It doesn't look like the overall OPR takes the coopertition bridge adjustment factor into account. The "Bridge" OPR column does, but the overall OPR shown is not the sum of the components.

Bongle 06-03-2012 06:40

Re: Week--1 All regionals OPR
 
Quote:

Originally Posted by Wayne TenBrink (Post 1139868)
It doesn't look like the overall OPR takes the coopertition bridge adjustment factor into account. The "Bridge" OPR column does, but the overall OPR shown is not the sum of the components.

Fixed in v20. The alliance scores used to build the OPR stat now get the CP bridge adjustment applied as well.

cgmv123 06-03-2012 07:27

Re: Week--1 All regionals OPR
 
Quote:

Originally Posted by Wayne TenBrink (Post 1139868)
It doesn't look like the overall OPR takes the coopertition bridge adjustment factor into account. The "Bridge" OPR column does, but the overall OPR shown is not the sum of the components.

I just copied and pasted the numbers and made the charts. Don't yell at me.

Josh Drake 06-03-2012 07:49

Re: Week--1 All regionals OPR
 
2751 had a simple strtegy that mainly focussed on balancing the bridges, so I feel good about being ranked high in that category, and realize we lack in the others. Still working on those.

Adam Freeman 06-03-2012 08:06

Re: Week--1 All regionals OPR
 
I think the Hybrid and Bridge OPR data will be pretty good indicators of a teams ability to perform those tasks.

But teams that just hit their Hybrid shots, pop in a few buckets, assist their partners in balancing, then go to the Coop bridge will not rank very high in Telop OPR.

This game completely changes when you get to Eliminations. You NEED to be able to score quickly and accurately in Teleop before you go to balance. But, that has not really been the case in Qualifications (at least in Week 1).

Teams that focus on Teleop scoring in qualifications will rank high in Teleop OPR, but that may not be the focus of the best teams.

It will be interesting to see how this data pans out as the game evolves. I would assume as teams get more competitive you will need to score more Teleop baskets to win matches.

Jared Russell 06-03-2012 08:09

Re: Week--1 All regionals OPR
 
Quote:

Originally Posted by Adam Freeman (Post 1139985)
I think the Hybrid and Bridge OPR data will be pretty good indicators of a teams ability to perform those tasks.

But teams that just hit their Hybrid shots, pop in a few buckets, assist their partners in balancing, then go to the Coop bridge will not rank very high in Telop OPR.

This game completely changes when you get to Eliminations. You NEED to be able to score quickly and accurately in Teleop before you go to balance. But, that has not really been the case in Qualifications (at least in Week 1).

Teams that focus on Teleop scoring in qualifications will rank high in Teleop OPR, but that may not be the focus of the best teams.

It will be interesting to see how this data pans out as the game evolves. I would assume as teams get more competitive you will need to score more Teleop baskets to win matches.

I tend to agree. I think that Hybrid OPR stands on its own as a pretty objective measure. But you probably need to look at Teleop OPR + Balance OPR (adjusted for Coopertition Balancing) together to get a good indicator of a robot's post-hybrid performance.

Nemo 06-03-2012 08:22

Re: Week--1 All regionals OPR
 
Quote:

I got the numbers below. It would be useful if people could do some mild validation. For example: Were 2751, 967, and 1986 all very bridge-happy? Did 3597 sink pretty much every single one of their shots in the high basket in hybrid mode, or score more than 2 balls per hybrid mode, or aid their allies in scoring lots of balls in hybrid mode?
967 was fairly bridge happy, plus we usually had somebody good on the opposing alliance who could reliably get onto the coopertition bridge with us.

Chris Hibner 06-03-2012 08:24

Re: Week--1 All regionals OPR
 
Adam makes a very interesting point. There were several matches at Kettering where we knew that having a two robot balance would guarantee a win, but neither partner could manipulate the bridge, and we always went for the coop bridge. We spent the vast majority of those matches helping partners on the alliance bridge, then balancing on the coop bridge. I don't even know if we shot a ball in teleop in those matches (maybe one or two).

This will change in the later competitions. First, just getting a two bot balance won't be good enough. Also, a lot of teams will work on getting a working bridge manipulator. OPR will probably be more representative in later competitions, but it probably is too bad now (with the coop points thrown in).

alschnelle 09-03-2012 14:05

Re: Week--1 All regionals OPR
 
Here are some videos for team 341.

http://www.youtube.com/watch?v=I8rQe4ECXqM
http://www.youtube.com/watch?v=I8rQe4ECXqM

Ether 12-03-2012 17:51

Re: Week--1 All regionals OPR
 
Quote:

Originally Posted by MagiChau (Post 1139468)
The @frcfms twitter feed keeps track of foul points in matches.

Does anyone know:

1) why there was (and is) no Twitter data for Oregon and Traverse City, and

2) is this data lost forever ? and

3) is there some person to whom this should be reported ?



Navid Shafa 12-03-2012 19:45

Re: Week--1 All regionals OPR
 
Using the OPR executable it is possible to query the individual regionals, but I've been unable to figure out how to get it to output everything from all the regionals this year.

In previous years, If I wanted to compare cross-regional, I did it by querying each regional and writing down the data myself.

How do you calculate the week one OPR rankings across all the regionals? Do you look at the individual regionals and splice them together yourself to make the final post? Is it possible to enter a separate code, rather than an individual regional code to compare teams from all events?

stundt1 12-03-2012 20:19

Re: Week--1 All regionals OPR
 
Week 2 opr rankings are up!

http://www.chiefdelphi.com/forums/sh...41#post1143141


All times are GMT -5. The time now is 07:22.

Powered by vBulletin® Version 3.6.4
Copyright ©2000 - 2017, Jelsoft Enterprises Ltd.
Copyright © Chief Delphi