Chief Delphi

Chief Delphi (http://www.chiefdelphi.com/forums/index.php)
-   Rules/Strategy (http://www.chiefdelphi.com/forums/forumdisplay.php?f=6)
-   -   Average Winning Score Perspective (http://www.chiefdelphi.com/forums/showthread.php?t=110445)

pandamonium 05-01-2013 13:31

Average Winning Score Perspective
 
This is an exercise I like to try with my team and this year I want everyone else to provide input.

Premise: Each persons design and strategy is based around what they perceive match scores to be.

Task: Try to predict the average score of the winning alliance for each point of the season listed below

Week 1 Regional Qualifications
Week 1 Regional Eliminations
Week 3 Regional Qualifications
Week 3 Regional Eliminations
Week 6 Regional Qualifications
Week 6 Regional Eliminations
Championships Qualifications
Championships Eliminations

Bonus: Predict the high score of the season

Please provide some form of logical explanation to explain your educated guess.

Racer26 05-01-2013 13:34

Re: Average Winning Score Perspective
 
Wk1 Q: 40-50
Wk1 E: 60-80
Wk3 Q: 50-60
Wk3 E: 80-100
Wk6 Q: 50-60
Wk6: E: 100-120
C Q: 80-100
C E: 120-140.

Highscore: 190.

Rationale: Most #1-#2 alliances in Elims anywhere will have 2 bots that can 30pt climb. As regionals progress, more alliances will have two 30pt climbers, Wk1 Quals are weak, and one strong robot will singlehandedly win most matches, but after about Wk3, Regional Quals don't get much stronger, as the fields just aren't deep enough.

Highscore rationale: Consider 1114 and 2056 (who in my estimation are virtually guaranteed to play at least one match allied this season, given that both are going to the same 3 regionals+CMP, and historically have allied as the #1 alliance all but one time that both were at the same regional). Both will be able to 30pt climb. Both will be strong Frisbee throwers that can score most of their frisbees. Assuming they score all 6 colored DISCs in the pyramid goal (30pts) plus 2x30pt climb, plus 3x6pt auto shots each is another 36pts, for 126pts, Between them, probably 2/3 of the remaining 45 discs will be scored for 3 pts. Total 216.

Max score possible is (3x30pts CLIMBs) + (13x6pt Auto Mode Top Goal shots [3 robots x their initial 3, plus the 2 under the pyramid, plus the 2 closest to the ALLIANCE STATIONS]) + (6x5pt Pyramid Goal) + (45x3pt shots with the ALLIANCE STATION DISCS) + (2x3pts CENTER FIELD DISCS), assuming you don't steal any of the opposing alliance's DISCS. 339 Points, before you consider stealing DISCS, and assuming you can fit 60 discs in the Top Goal.

XXShadowXX 05-01-2013 13:43

Re: Average Winning Score Perspective
 
Highest score I see possible is somewhere around 200 (243, if I did my math right), but I doubt we'll see it.

Racer26 05-01-2013 13:53

Re: Average Winning Score Perspective
 
I expect we will see a few teams with 5-7 DISC in the top goal Autonomous modes.

Racer26 05-01-2013 13:59

Re: Average Winning Score Perspective
 
Other thoughts related to max scores: Frisbees are more-or-less rigid. They should be far more consistent to shoot than the foam basketballs of 2012. Its entirely plausible that full-field shots (ie. Park your robot in front of the FEEDER STATION, where it's protected from defense, and shoot TOP GOALs from there.) will be a thing.

The biggest thing limiting scores in 2012 was collecting the balls. If you can have 51 DISCs put straight into your ROBOT, and score them without moving, you'll be at a MASSIVE advantage.

pandamonium 05-01-2013 13:59

Re: Average Winning Score Perspective
 
Auto: 16*6= 96points + 30points red/blue disk points = 126 + 90 climbing points = 216 points

That is not even counting the other non autonomous white disks or the massive points accumulated from penalties.

karomata 05-01-2013 14:08

Re: Average Winning Score Perspective
 
Quote:

Originally Posted by 1075guy (Post 1207935)
Wk1 Q: 40-50
Wk1 E: 60-80
Wk3 Q: 50-60
Wk3 E: 80-100
Wk6 Q: 50-60
Wk6: E: 100-120
C Q: 80-100
C E: 120-140.

Highscore: 190.

Rationale: Most #1-#2 alliances in Elims anywhere will have 2 bots that can 30pt climb. As regionals progress, more alliances will have two 30pt climbers, Wk1 Quals are weak, and one strong robot will singlehandedly win most matches, but after about Wk3, Regional Quals don't get much stronger, as the fields just aren't deep enough.

Highscore rationale: Consider 1114 and 2056 (who in my estimation are virtually guaranteed to play at least one match allied this season, given that both are going to the same 3 regionals+CMP, and historically have allied as the #1 alliance all but one time that both were at the same regional). Both will be able to 30pt climb. Both will be strong Frisbee throwers that can score most of their frisbees. Assuming they score all 6 colored DISCs in the pyramid goal (30pts) plus 2x30pt climb, plus 3x6pt auto shots each is another 36pts, for 126pts, Between them, probably 2/3 of the remaining 45 discs will be scored for 3 pts. Total 216.

Max score possible is (3x30pts CLIMBs) + (13x6pt Auto Mode Top Goal shots [3 robots x their initial 3, plus the 2 under the pyramid, plus the 2 closest to the ALLIANCE STATIONS]) + (6x5pt Pyramid Goal) + (45x3pt shots with the ALLIANCE STATION DISCS) + (2x3pts CENTER FIELD DISCS), assuming you don't steal any of the opposing alliance's DISCS. 339 Points, before you consider stealing DISCS, and assuming you can fit 60 discs in the Top Goal.

I agree with your assumptions, although I feel that by the end of champs and at IRI (I know that doesn't really count), we will see higher scores than you have predicted

Quote:

Originally Posted by XXShadowXX (Post 1207940)
Highest score I see possible is somewhere around 200 (243, if I did my math right), but I doubt we'll see it.

I wouldn't put that possibility out of the question. In 2011 everyone said that no alliance would fill the rack and win both 1st and 2nd in the rat race, yet on the Newton field 148's Alliance proved us all wrong.

The way I see it, we can only be sure it's impossible if 148 can't do it. ;)

Tetraman 05-01-2013 14:23

Re: Average Winning Score Perspective
 
Most scores season long will be around 40-60, with around 60-100 in elimination rounds depending on the alliance. I expect a LOT of blowouts in qualification matches.

Racer26 05-01-2013 14:27

Re: Average Winning Score Perspective
 
Quote:

Originally Posted by Tetraman (Post 1207975)
Most scores season long will be around 40-60, with around 60-100 in elimination rounds depending on the alliance. I expect a LOT of blowouts in qualification matches.

Concur. Simply due to how one really consistent shooter (say 1114, 2056, 217, 148, 254, 1625 and scads more) can easily rack up 60 pts on their own, between 30pts for climbing, and another 15 or so in a good auto.

Furthermore, I think its going to be harder to beat #1 alliances than usual, since by and large, this game requires little in the way of help from alliance mates.

Herbblood 05-01-2013 16:05

Re: Average Winning Score Perspective
 
The highest possible score is higher beacuse of the other teams frisbees, but i estimate that most teams in qualifications will make around 60 without fouls, and in elminations towards 90 and above. At championships i assume lots of over a hundred.

Donut 05-01-2013 16:24

Re: Average Winning Score Perspective
 
The average winning match score will probably be something like 60 points, though I think it will be skewed by extremely high match scores of elite teams. Median will probably be something more like 40.

I have to agree with the potential for a lot of blowouts this year. Climbing seems like a fairly difficult challenge that elite level teams will meet but most teams will face plant trying to accomplish. An elite team can easily score 60+ in a match without trying just by climbing for 30, dumping a few colored disks in the top at the end, and scoring a few high goal disks in autonomous. Meanwhile the opposing alliance will dump some 1 point frisbees in the low goal and maybe get 10 points if one of the teams was smart enough to go for the simple level 1 chin up instead of trying to climb to the top.

The high score for the season will be over 200. An alliance of all top notch robots will get 90 for 3 level 3 robots, at least 20 points in autonomous, and 90+ in teleop from colored disks scored on top of the pyramid and general shooting in the middle and high goals.

MagiChau 05-01-2013 16:58

Re: Average Winning Score Perspective
 
I say in week 1 qualification we will see fairly low winning scores. Probably 20-30 as the mean is my prediction. Elimination goes up to 40-50. I am very scared of blowouts for the end game. I agree with the sentiment that we will return to #1 Alliance dominance with the top 2 shooters and possibly the top 2 climbers.

Ian Curtis 05-01-2013 18:16

Re: Average Winning Score Perspective
 
Way cool that people (besides me) care about this! Here is the Predict The Average Score threads from the past several years.

Why did you choose the average winning score? Historically there is a pretty significant difference between the distribution of winning scores and the distribution of losing scores (in terms of shape and mean/median) -- but isn't it enough to score the losing alliance's score + 1?

Joon Park 05-01-2013 21:19

Re: Average Winning Score Perspective
 
Quote:

Originally Posted by 1075guy (Post 1207955)
Other thoughts related to max scores: Frisbees are more-or-less rigid. They should be far more consistent to shoot than the foam basketballs of 2012. Its entirely plausible that full-field shots (ie. Park your robot in front of the FEEDER STATION, where it's protected from defense, and shoot TOP GOALs from there.) will be a thing..

More consistent, but also more difficult. Considering the distance, I don't think we'll ever see consistent full-field shots, especially with the pyramids in the way.

pandamonium 05-01-2013 21:56

Re: Average Winning Score Perspective
 
Quote:

Originally Posted by Ian Curtis (Post 1208189)
Way cool that people (besides me) care about this! Here is the Predict The Average Score threads from the past several years.

Why did you choose the average winning score? Historically there is a pretty significant difference between the distribution of winning scores and the distribution of losing scores (in terms of shape and mean/median) -- but isn't it enough to score the losing alliance's score + 1?

In a perfect world yes, but I like to build in a little bit of room for error. I find this statistic to be more accurate as a reliable goal. It balences out low outliers like aliances that score 0...


I have the feeling that the elite teams will be very very elite this year I can not wait to watch championships and posibilities of 3 30 point hangs.


All times are GMT -5. The time now is 04:34.

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