Chief Delphi

Chief Delphi (http://www.chiefdelphi.com/forums/index.php)
-   Fantasy FIRST (http://www.chiefdelphi.com/forums/forumdisplay.php?f=128)
-   -   [FF]: 2016 Season Long Fantasy FIRST (http://www.chiefdelphi.com/forums/showthread.php?t=137788)

EricH 10-01-2016 19:54

Re: [FF]: 2016 Season Long Fantasy FIRST
 
Apologies to an anonymous claimant for making them wait over a day for the posting:

1023 is on waiver in Pittsburgh, with a claim. Claim window closes 1/13 at 8 PM EST.

Navid Shafa 11-01-2016 13:32

Re: [FF]: 2016 Season Long Fantasy FIRST
 
The same District Ranking Method used last year will be implemented this season.

EricH 12-01-2016 23:35

Re: [FF]: 2016 Season Long Fantasy FIRST
 
Quote:

Originally Posted by Navid Shafa (Post 1521153)
The same District Ranking Method used last year will be implemented this season.

I noticed that. OTOH, W-L-T is back as the primary method of ranking. (And then there's the extra RPs...)

Usually, we go W-L-T at 2-0-1, then add on ranking points for just the top 15 or 16 teams. (Last year, of course, was an exception.) I'm thinking there are a few ways to go here.

1) W-L-T, + standard rank points for the teams that seed high enough to make elims on their own if the cards line up.
2) RP total, + standard rank points.
3) District Ranking.
4) W-L-T, + District Ranking.
5) RP total, + District Ranking.

Of the 5, I'm leaning towards #5, as it gives out the most total points and is the easiest to calculate. OTOH, it'll really bias towards districts and smaller events with their 12+ matches/team. My second option would be #2, due to it being more standard.

Navid Shafa 13-01-2016 21:55

Re: [FF]: 2016 Season Long Fantasy FIRST
 
Quote:

Originally Posted by EricH (Post 1522450)
Usually, we go W-L-T at 2-0-1, then add on ranking points for just the top 15 or 16 teams. (Last year, of course, was an exception.) I'm thinking there are a few ways to go here.

I'm not recalling how this was done in 2014 and prior. Would that be W-L-T points plus additional rank points starting at 16 for 1st seed and down to 1 point for the 16th seed? Is this what you are referring to as Standard rank below?

Quote:

Originally Posted by EricH (Post 1522450)
1) W-L-T, + standard rank points for the teams that seed high enough to make elims on their own if the cards line up.
2) RP total, + standard rank points.
3) District Ranking.
4) W-L-T, + District Ranking.
5) RP total, + District Ranking.

Of the 5, I'm leaning towards #5, as it gives out the most total points and is the easiest to calculate. OTOH, it'll really bias towards districts and smaller events with their 12+ matches/team. My second option would be #2, due to it being more standard.

Since we have the table on the last sheet here from last season, this should be easy enough for us to use while working on automated FF scoring via the API for this year. Without ease of calculation being an issue, I am strongly against most options, save for 3.

Trying to use W-L-T as a stand-alone is a poor metric, because most teams may be designing strategies that net them a higher average RP, but net them a lower match score. Trying to use RP is a little better, but it creates an inflation of ranking points (relative to award and elimination points). This inflation may adversely affect Fantasy teams that made databases/picks based on a certain weighting of on-field performance.

The fact that District Ranking is normalized and is comparable between events has me leaning this route. Trying to add additional rank points to the district ranking defeats the purpose of normalization IMO.


Quote:

Originally Posted by EricH (Post 1520532)
1023 is on waiver in Pittsburgh, with a claim. Claim window closes 1/13 at 8 PM EST.

Who got lucky? ;)

EricH 13-01-2016 22:24

Re: [FF]: 2016 Season Long Fantasy FIRST
 
Quote:

Originally Posted by Navid Shafa (Post 1523198)
I'm not recalling how this was done in 2014 and prior. Would that be W-L-T points plus additional rank points starting at 16 for 1st seed and down to 1 point for the 16th seed? Is this what you are referring to as Standard rank below?

It is what I'm referring to as Standard rank. From the posted rules:
1st seed, 20 points
2nd-3rd seed, 12 points
4th-8th seed, 6 points
9th-12th seed, 3 points
13th-16th seed, 2 points

Subject to change based on the game given, naturally.

Quote:

Since we have the table on the last sheet here from last season, this should be easy enough for us to use while working on automated FF scoring via the API for this year. Without ease of calculation being an issue, I am strongly against most options, save for 3.

Trying to use W-L-T as a stand-alone is a poor metric, because most teams may be designing strategies that net them a higher average RP, but net them a lower match score. Trying to use RP is a little better, but it creates an inflation of ranking points (relative to award and elimination points). This inflation may adversely affect Fantasy teams that made databases/picks based on a certain weighting of on-field performance.

The fact that District Ranking is normalized and is comparable between events has me leaning this route. Trying to add additional rank points to the district ranking defeats the purpose of normalization IMO.
Just as a thought, the defenses may adversely affect real FRC teams that developed offseason drivetrains based on a certain assumption of field terrain. ;)

The main reason we went to the district ranking in the first place is because FIRST wasn't tracking W-L-T, so we had no way to calculate that. District rank only affects district advancement, which was also the case last year, but we have W-L-T back this year. Remember, we're scoring at events, NOT between them. FIRST doesn't normalize between events. (Except for the district ranking.)

The only two times we HAVEN'T used straight W-L-T were 2012 (the bridges were added in) and 2015 (no W-L-T to count). In all other years, it's been W-L-T + rank. That's a big part of why I favor the options without the district ranking--this is, after all, a Traditional league, though being the only league kind of led to that tag being dropped.


How about using Option 4? Keep the W-L-T, but swap out the rank bonus points for the district ranking, which gives all the teams points. If we really wanted to, we could assign a negative value to the losses.



Quote:

Who got lucky? ;)
Swamp Life, for 2526.

Navid Shafa 13-01-2016 22:41

Re: [FF]: 2016 Season Long Fantasy FIRST
 
Quote:

Originally Posted by EricH (Post 1523222)
How about using Option 4? Keep the W-L-T, but swap out the rank bonus points for the district ranking, which gives all the teams points. If we really wanted to, we could assign a negative value to the losses.

I think it's the lesser of the evils, fine by me. ;)

Quote:

Originally Posted by EricH (Post 1523222)
Swamp Life, for 2526.

Updated Rosters


Quote:

Originally Posted by EricH (Post 1523222)
Just as a thought, the defenses may adversely affect real FRC teams that developed offseason drivetrains based on a certain assumption of field terrain. ;)

Tangential: I would strongly argue that teams of this caliber are unlikely to blindly copy-paste their offseason drive trains. Teams willing to put in the time for off-season R&D, and/or training are likely smart enough to design around a game.

Rangel 13-01-2016 23:31

Re: [FF]: 2016 Season Long Fantasy FIRST
 
What's the problem with option 1? You have your standard W/L/T and then extra RP would be reflected in the actual standings.

EricH 20-01-2016 00:46

Re: [FF]: 2016 Season Long Fantasy FIRST
 
OK, we have another waiver claim up:

Anonymous team has claimed 11 in South Florida. This one closes 1/22/16, we'll call it midnight PST.

Alpha Beta 21-01-2016 15:05

Re: [FF]: 2016 Season Long Fantasy FIRST
 
GKC just went from 57 to 58 teams. ;)

EricH 21-01-2016 19:22

Re: [FF]: 2016 Season Long Fantasy FIRST
 
Quote:

Originally Posted by Alpha Beta (Post 1527896)
GKC just went from 57 to 58 teams. ;)

OK, I see a lot of folks figured out that that's 1986 in GKC. Waiver closes 1/24, 8 PM EST.

For reference, the current priority list is:
Broconnor
The Breakfast Company
The ULCAN
FIRST Pick
Rotten Fruit Alliance
OrangeJuice
Falcon
Swamp Life

EricH 23-01-2016 00:01

Re: [FF]: 2016 Season Long Fantasy FIRST
 
Well, I do believe that I've got some waivers to award.

The Breakfast Company picks up 11 for 2905 in South Florida.

And ordinarily I'd let the waivers run for 1986, but when the top of the list makes a claim, there's no point. Broconnor drops 5730 to pick up 1986 in GKC.

Current priority list:
The ULCAN
FIRST Pick
Rotten Fruit Alliance
OrangeJuice
Falcon
Swamp Life
The Breakfast Company
Broconnor

EricH 21-02-2016 23:51

Re: [FF]: 2016 Season Long Fantasy FIRST
 
OK, folks, as a reminder, we're coming up on competition season. So, we've got a few notes...

--Lists will lock (no trades, unless someone is replacing a no-show) at midnight on Wednesday prior to an event beginning. So, Palmetto locks the day after Bag Day. Week 1 events lock 1 week later.
--Scoring: As discussed some time ago, we will be scoring W-L-T, and adding in some ranking points using the District ranking model.


And a waiver is up: 842, Las Vegas, closes 2/24 at midnight Eastern.

EricH 22-02-2016 01:21

Re: [FF]: 2016 Season Long Fantasy FIRST
 
And another waiver just went up: 5817, Orange County. We'll call that the same close time as before.

Navid Shafa 26-02-2016 15:08

Re: [FF]: 2016 Season Long Fantasy FIRST
 
Quote:

Originally Posted by EricH (Post 1544040)
And a waiver is up: 842, Las Vegas, closes 2/24 at midnight Eastern.

Quote:

Originally Posted by EricH (Post 1544092)
And another waiver just went up: 5817, Orange County. We'll call that the same close time as before.

Results?

tindleroot 26-02-2016 15:14

Re: [FF]: 2016 Season Long Fantasy FIRST
 
Just for reference, can someone list the point values for scoring? I know there's been discussion on the style but it has all been fairly ambiguous as to the actual point values. Thanks.


All times are GMT -5. The time now is 16:48.

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