User Pass
Home Sign Up Contact Log In
Forum > Pee Wee Leagues > Pee Wee Gold League > Yardage vs points more important in ranking an offense/defense?
Page:
 
Gerr
offline
Link
 
Originally posted by PING72

I'd say:

60% record
35% stats
5% SOS


I could easily do this. Stats would be broken down at 15% Offense, 15% defense, and 5% ST. And I could easily break down the offense and defense as you suggested above as right now, I have it 50/50 between yards and scoring.
 
PING72
offline
Link
 
I'd suggest trying a few different things, posting the results and letting people look over them and see which looks best.
 
Gerr
offline
Link
 
I still have to finish setting up everything in Excel, but will see what happens after today's games.
 
DirtyMike
offline
Link
 
Originally posted by The Eagle
I think ranking by

For record, I think you go through the schedule and give each team some weighted point value for each win (and subtract a bit for each loss), based on the SoS of each opponent. The weighting factor would depend on how you're weighting stats, so that the record tally is kinda in proportion with the stats tally. The idea would be you get more points for beating a good opponent and less for a lousy one; similarly, you lose a lot of points if you get beat by an easy opponent, and you don't lose as much if you get beat by a high-ranking team. Doing this, it'll be very difficult for teams with the record to get the same tallied record statistic. I've seen a similar formula for rankings on the web before; but I just did some quick google searching and couldn't find it.

Another idea (never seen this before...but wouldn't be surprised if it's out there already) is to simulate the rest of the season based on current rankings and stats (like the fantasy predictor in the 2.6GD script), and use the final estimated record (real + "fantasy prediction") to rank everybody. Then you'd have to use something like the current O and D rankings to sort teams with the same predicted final record.


Eagle.. you and I think alike. I have been a fan of the ELO Based Gaming Ladder System and mentioned it here many times. This system does not care if your #1 team ran up 700 yards versus a #32 team. This is a ranking strictly based on strength of opponent. For the most accurate rating of team strength extrapulated out through the entire season this would be a clear and decisive ranking. If someone will PM me their email address I will send a spreadsheet i have started that is complete through Week 2. I would love someone to continue through for the rest of the schedule. If not I will complete it in the next few days. As Follows;

All Teams Begin Season with 1000 Pts

Points Scaling Always = 120 Split between 2 teams
Win vs Equal Opponent = 60 Pts
Loss vs Equal Opponent = - 60 Pts
Win vs Opponent with 1 - 100 Pts Greater = +70 Pts
Win vs Opponent with 1 - 100 Pts Less = +50 Pts
Loss vs Opponent with 1 - 100 Pts Greater = - 50 Pts
Loss vs Opponent with 1 - 100 Pts Less = - 70 Pts
Win vs Opponent with 101 - 150 Pts Greater = +80 Pts
Win vs Opponent with 101 - 150 Pts Less = +40 Pts
Loss vs Opponent with 101 - 150 Pts Greater = -40 Pts
Loss vs Opponent with 101 - 150 Pts Less = -80 Pts
Win vs Opponent with 151 - 200 Pts Greater = +90 Pts
Win vs Opponent with 151 - 200 Pts Less = +30 Pts
Loss vs Opponent with 151 - 200 Pts Greater = -30 Pts
Loss vs Opponent with 151 - 200 Pts Less = -90 Pts
Win vs Opponent with 201 - 250 Pts Greater = +100 Pts
Win vs Opponent with 201 - 250 Pts Less = +20 Pts
Loss vs Opponent with 201 - 250 Pts Greater = -20 Pts
Loss vs Opponent with 201 - 250 Pts Less = -100 Pts
Win vs Opponent with 251 - 300 Pts Greater = +110 Pts
Win vs Opponent with 251 - 300 Pts Less = +10 Pts
Loss vs Opponent with 251 - 300 Pts Greater = -10 Pts
Loss vs Opponent with 251 - 300 Pts Less = -110 Pts
Win vs Opponent with 300 Pts or Greater = +120 Pts
Win vs Opponent with 300 + Pts Less = +5 Pts
Loss vs Opponent with 300 Pts or Greater = 0 Pts
Loss vs Opponent with 300 + Pts Less = -120 Pts
Draws = Win Pts - 60 Pts
Edited by DirtyMike on Jun 7, 2011 14:18:10
Edited by DirtyMike on Jun 7, 2011 14:15:44
Edited by DirtyMike on Jun 7, 2011 14:13:40
Edited by DirtyMike on Jun 7, 2011 14:13:00
 
DirtyMike
offline
Link
 
Example through Week 2, TOP 10

..........................................Preseason Ranking Points.....Week 1 Ranking Points.....Week 2 Ranking Points
Atlantic City Surf..................................1000.......................................1060......................................1120
Brigate Nere Italia..............................1000.......................................1060.......................................1120
Chicago 85 Bears..............................1000.......................................1060.......................................1120
Ogopogo Lake monsters.................1000.......................................1060.......................................1120
Fighting Gamecocks.........................1000.......................................1060.......................................1100
Little Green Men.................................1000.......................................1060.......................................1100
Maine Grenadiers..............................1000.......................................1060.......................................1100
Pain Nation Midgets..........................1000.......................................1060.......................................1100
SAS Chicken Stranglers...................1000.......................................1060.......................................1100
Team Tapout......................................1000 .......................................1060.......................................1100
Springfield Test Dummies..............1000 940 1020
Crazytown Invisible Killer Clown....1000 940 1020
Lunch Box...........................................1000 1060 1000
Montpellier Wildcats.........................1000 1060 1000
Providence Nose Pickers................1000 1060 1000
Crib Midgets.......................................1000..................................1060 1000
Baytown Bulldogs.............................1000 940 1000
Cascade Surely Goats 1000 940 1000
Rockford Revival 1000 940 1000
The Insane Asylum 1000 940 1000
Arizona Spitoons 1000 1060 980
Devils Advocates 1000 1060 980
Short Bus Window Lickers 1000 940 900
Texas Shorthorns 1000 940 900
Dirty Inc 1000 940 900
Shy Town Bears 1000 940 900
South Carolina Muddwhomps 1000 940 900
Twin Peaks 1000 940 900
Boendby Pee Wee Bengals 1000 940 880
Addict Nation PW 1000 940 880
Kill Devil Hill Flyers 1000 940 880
Woodinville Truth 1000 940 880
Edited by DirtyMike on Jun 7, 2011 14:28:03
Edited by DirtyMike on Jun 7, 2011 14:27:50
Edited by DirtyMike on Jun 7, 2011 14:26:57
 
The Eagle
offline
Link
 
good stuff, Mike. ELO is exactly where I saw that concept (and that's why I couldn't find it earlier today)
 
durettug
offline
Link
 
without actually looking at data, teams, or rankings, I would think that determining who is the better team should be based off a combination of win/loss record tied into the teams strength of schedule and the team's opponents SoS. Personally I think Off/Def stats are skewed by the blowouts...

For instance, look what college basketball does with their API when determining who is the better team to get into the NCAA tourney...

ELO works awesome for the final weeks of the season and end of season ranks when everybody plays the same teams, but is not accurate at the beginning of the season when a team may have 3 blowout wins. Say that team is the grenadiers, PNM or 85 Bears who are always near the top, their rank would be below what the should be. Imo
Edited by durettug on Jun 7, 2011 15:47:16
 
Page:
 


You are not logged in. Please log in if you want to post a reply.