Chief Delphi

Chief Delphi (http://www.chiefdelphi.com/forums/index.php)
-   General Forum (http://www.chiefdelphi.com/forums/forumdisplay.php?f=16)
-   -   Toughest Division ? (http://www.chiefdelphi.com/forums/showthread.php?t=136642)

SciBorg Dave 15-04-2015 16:06
Toughest Division ?
 
What is the toughest Division ? and why?
I am still looking over them.

MaGiC_PiKaChU 15-04-2015 16:10
Re: Toughest Division ?
 
Quote:
Originally Posted by SciBorg Dave (Post 1471523)
What is the toughest Division ? and why?
I am still looking over them.
this year it will be pretty close, because most *good* teams , will be able to put 7 stacks together... what will change the game is canburglars

JABot67 15-04-2015 16:10
Re: Toughest Division ?
 
According to the graph at http://championship-notifier.evanforbes.net:3000/, especially if you look at the top 24 teams in each division, Carson takes the cake. At least with regards to OPR of top teams.

microbuns 15-04-2015 16:11
Re: Toughest Division ?
 
The division of death has to be Carson.

67, 225, 254, 1519, 1730, 2085, 4488, 5254, and 5406 are all incredibly strong teams that would likely be in a high picking position in any other division.

Hot_Copper_Frog 15-04-2015 16:13
Re: Toughest Division ?
 
Quote:
Originally Posted by JABot67 (Post 1471530)
According to the graph at http://championship-notifier.evanforbes.net:3000/, especially if you look at the top 24 teams in each division, Carson takes the cake. At least with regards to OPR of top teams.
This is true, but Archimedes has a significantly "higher" OPR floor and a mid-range OPR ceiling, which makes for a deeper talent pool come alliance selections. Those third picks can make or break alliances at this level.

Dave Campbell 15-04-2015 16:21
Re: Toughest Division ?
 
Quote:
Originally Posted by Hot_Copper_Frog (Post 1471535)
This is true, but Archimedes has a significantly "higher" OPR floor and a mid-range OPR ceiling, which makes for a deeper talent pool come alliance selections. Those third picks can make or break alliances at this level.
Exactly right. Third pick in all divisions will be very critical. Depth of field is important unless a division has 17 robots able to "win" all matches by themselves.

Rachel Lim 15-04-2015 16:27
Re: Toughest Division ?
 
Quote:
Originally Posted by Hot_Copper_Frog (Post 1471535)
This is true, but Archimedes has a significantly "higher" OPR floor and a mid-range OPR ceiling, which makes for a deeper talent pool come alliance selections. Those third picks can make or break alliances at this level.
Looking at the top 32 teams, which would be the number in elims per division, I'd say Carson still has the lead. Ranked by their best OPR, the 32nd top team in Archimedes has an OPR around 42, compared to Carson's 48. Their average is also 44 vs 40.5. Curie is also gong to be very interesting, with 1114 and 148 both there.


But on Einstein, it's the first 0.2sec that matter, and Newton seems to be ahead in that aspect. I'm sure many teams are aiming to take away that advantage, though.

King Boopington 15-04-2015 20:40
Re: Toughest Division ?
 
I think Carson is the deepest, but Curie is the most top heavy.

MisterG 15-04-2015 23:05
Re: Toughest Division ?
 
Code:
                OPR(top 24)                        Worldrank(top 24)                       
Division        Average        Median        Max        Min        Average        Median        Min        Max
Carson                60.43        51.54        115.63        42.14        97        98        2        203
Tesla                57.72        55.83        81.57        41.87        93        77        13        212
Curie                56.97        51.85        121.67        42.02        99        95        1        207
Newton                56.43        51.60        98.18        37.10        110        101        4        293
Archimedes        55.06        51.43        81.08        39.51        110        98        15        249
Galileo                54.58        48.16        107.32        40.72        123        128        3        231
Hopper                52.92        47.41        88.72        37.99        125        132        7        277
Carver                51.61        48.52        75.06        40.37        126        125        22        236
Boom Carson. So much for that Carson curse I have been going on about ;)

Thanks to 2834 and their awesome data!

Abhishek R 15-04-2015 23:07
Re: Toughest Division ?
 
Quote:
Originally Posted by MisterG (Post 1471802)
Code:
                OPR(top 24)                        Worldrank(top 24)                       
Division        Average        Median        Max        Min        Average        Median        Min        Max
Carson                60.43        51.54        115.63        42.14        97        98        2        203
Tesla                57.72        55.83        81.57        41.87        93        77        13        212
Curie                56.97        51.85        121.67        42.02        99        95        1        207
Newton                56.43        51.60        98.18        37.10        110        101        4        293
Archimedes        55.06        51.43        81.08        39.51        110        98        15        249
Galileo                54.58        48.16        107.32        40.72        123        128        3        231
Hopper                52.92        47.41        88.72        37.99        125        132        7        277
Carver                51.61        48.52        75.06        40.37        126        125        22        236
Boom Carson. So much for that Carson curse I have been going on about ;)

Thanks to 2834 and their awesome data!
Does that use season average OPR or max OPR? It's probably more accurate to use max OPR.

MisterG 15-04-2015 23:16
Re: Toughest Division ?
 
I think it may be weighted average. All of the data is from the week 7 2834 data.

OPR is from column O of the Worldrank tab which is the same thing that comes up in the query search.

We often take the average of average and max when we are doing rankings, we call this third quartile (which I think is not strictly accurate) and we feel that it does a good job of predicting future results.

Feel free to PM me if you want the spreadsheet that I used to do the calcs.

Abhishek R 15-04-2015 23:19
Re: Toughest Division ?
 
Quote:
Originally Posted by MisterG (Post 1471810)
I think it may be weighted average. All of the data is from the week 7 2834 data.

OPR is from column O of the Worldrank tab which is the same thing that comes up in the query search.
I see, I was thinking it might be better to use Column G of the OPR Results tab because that uses the team's max OPR achieved throughout the course of the season, which usually happens to correlate with improvement over the season (thus often being the OPR of the event the team last competed at). That way, it's more reflective of the most recent/current ability of the team. Still interesting data to look at either way.

Dunngeon 15-04-2015 23:25
Re: Toughest Division ?
 
Quote:
Originally Posted by Rachel Lim (Post 1471554)
But on Einstein, it's the first 0.2sec that matter, and Newton seems to be ahead in that aspect. I'm sure many teams are aiming to take away that advantage, though.
I'd have to agree, Newton has the most established teams with very quick bin grabbers (1678, 3310 ect.) going into the event. If I had to pick right now, I would probably pick the Newton winners to be one of the tougher alliances on Einstein. However all the predictions with bin grabbers really can't be made until Thursday or Friday when teams will have them integrated and tested.

On a divisional level, in terms of pure scoring power as others have said, Carson takes the cake. Unfortunately, other divisions don't have a very big advantage by having a higher mid and lower pack this year. Due to the nature of the game (small field, score ceiling) 2 high level robots are all that is needed to win (if you get cans, big if) and in many cases will be more efficient.

Wayne TenBrink 16-04-2015 08:56
Re: Toughest Division ?
 
There are really two separate questions here:

1) Which division is most likely to have the highest QS (therefore, where will it be toughest to seed #1)?

2) Which division is most likely to win on Einstein?

For question #1, average OPR matters.

For question #2, the performance of the winning alliance (normally as determined by the #1 seed and their #1 pick) matters.

Carson gets my vote on question #1. Newton gets my vote on question #2.

Ed Law 16-04-2015 08:57
Re: Toughest Division ?
 
Quote:
Originally Posted by MisterG (Post 1471810)
I think it may be weighted average. All of the data is from the week 7 2834 data.

OPR is from column O of the Worldrank tab which is the same thing that comes up in the query search.

We often take the average of average and max when we are doing rankings, we call this third quartile (which I think is not strictly accurate) and we feel that it does a good job of predicting future results.

Feel free to PM me if you want the spreadsheet that I used to do the calcs.
The world OPR takes into account all events from week 1 to week 7. It is not just the max OPR of each team. It allows us to consider the interaction of teams from different district/areas when they travel to different events. This takes care of the argument about stronger/weaker events when comparing across regionals/districts. Another way to look at the World OPR is how well each team do for the whole season. If a team starts off slow and improves in later weeks, the world OPR will be like an average of their whole season. The world OPR may not be the best indication of how a team would do. I personally use max OPR of each team in estimating what a team can potentially do at World.

Citrus Dad 16-04-2015 22:44
Re: Toughest Division ?
 
Quote:
Originally Posted by MisterG (Post 1471810)
I think it may be weighted average. All of the data is from the week 7 2834 data.

OPR is from column O of the Worldrank tab which is the same thing that comes up in the query search.

We often take the average of average and max when we are doing rankings, we call this third quartile (which I think is not strictly accurate) and we feel that it does a good job of predicting future results.

Feel free to PM me if you want the spreadsheet that I used to do the calcs.
In the past I felt the average OPR was a reasonable indicator, and I would adjust the average OPR by an improvement factor over the season. And often teams would have lower OPRs at later events.

That's not the case this year. The improvement event to event has been dramatic. I think some teams have improved as much at 200% over 3 events. The average is really hit by how early a team first entered the season. Some teams that entered in Week 4 were much better simply by being able to watch earlier events. So this year we'll rely entirely on max OPR. However, I need to net out the coopertition OPR because it doesn't contribute to playoff scoring (or only at 10% of the 40 coop points at most.)

Spoam 17-04-2015 00:12
Re: Toughest Division ?
 
Quote:
Originally Posted by Citrus Dad (Post 1472421)
That's not the case this year. The improvement event to event has been dramatic. I think some teams have improved as much at 200% over 3 events.
I'd say this year "as much as 200%" is even an understatement. I know we went from a capped 5 our first event to 3 full stacks now. And I know some teams have made even larger jumps (OPR jumps of 60+).

Quote:
Originally Posted by Citrus Dad (Post 1472421)
However, I need to net out the coopertition OPR because it doesn't contribute to playoff scoring (or only at 10% of the 40 coop points at most.)
If you want this (or if anyone else does), we have an online OPR calculator that includes component OPRs and an overall OPR with coop and fouls already netted out (no coop for the exact reason you described). You can search for any event, division, or active team and it has team lists for every division sortable by highest no-coop OPR.

For example, the list for Newton is here.


All times are GMT -5. The time now is 02:55.

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