Cracking the NCAA SRS Code

[Montana Gym Rat]

As Dannyboy03867 has pointed out in another thread about playoff selection process, it appears that the NCAA has developed their own 'proprietary' Simple Rating System. There was already an SRS out there and has been published for at least 4 years or so, developed by Chase Stuart it would appear. Chase's SRS was based on a Margin Of Victory factor that could be calculated based upon point differential with a tweak for home/away/neutral. Each team was given a MOV and their SOS was an average of the MOV's of their opponents.

The NCAA's SRS does not take margin of victory into account, and they utilize a Win-Loss Differential instead. The SOS factor is still calculated the same way.
http://www.ncaa.org/wps/wcm/connect/public/ncaa/fcs+selections/fcs+selections" onclick="window.open(this.href);return false;

So the Win-Loss Differential is based on home/neutral/away and the NCAA subdivision of the opponent from this link. And I've heard coaches/articles keep referring to this as a type of RPI system for college football.

Taking these items into account, I assume that the NCAA SRS goes something like this:
- Road Wins Against FCS Opponent = 1.4
- Neutral Wins Against FCS Opponent = 1.0
- Home Wins Against FCS Opponent = 0.6
- Road Losses Against FCS Opponent = -0.6
- Neutral Losses Against FCS Opponents = -1.0
- Home Losses Against FCS Opponents = -1.4

Now, what is the tweak to account for FBS and Division II teams? Note that the link says games against Division III & non-NCAA teams will not be counted. (Which definitely means Marist won out over Butler in the SRS, by the way.)

Maybe something like this?
- Road Wins Against FBS Opponent = 1.8
- Neutral Wins Against FBS Opponent = 1.4
- Home Wins Against FBS Opponent = 1.0
- Road Losses Against FBS Opponent = -0.2
- Neutral Losses Against FBS Opponent = -0.6
- Home Losses Against FBS Opponent = -1.0

- Road Wins Against D-II Opponent = 1.0
- Neutral Wins Against D-II Opponent = 0.6
- Home Wins Against D-II Opponent = 0.2
- Road Losses Against D-II Opponent = -1.0
- Neutral Losses Against D-II Opponent = -1.4
- Home Losses Against D-II Opponent = -1.8


These are guesses on my part based on the current RPI system and the references to RPI. I'm going to try my hand at calculating the NCAA SRS with these parameters. Stay tuned.

 

[G-BEARS]

As Dannyboy03867 has pointed out in another thread about playoff selection process, it appears that the NCAA has developed their own 'proprietary' Simple Rating System. There was already an SRS out there and has been published for at least 4 years or so, developed by Chase Stuart it would appear. Chase's SRS was based on a Margin Of Victory factor that could be calculated based upon point differential with a tweak for home/away/neutral. Each team was given a MOV and their SOS was an average of the MOV's of their opponents.

The NCAA's SRS does not take margin of victory into account, and they utilize a Win-Loss Differential instead. The SOS factor is still calculated the same way.
http://www.ncaa.org/wps/wcm/connect/public/ncaa/fcs+selections/fcs+selections" onclick="window.open(this.href);return false;

So the Win-Loss Differential is based on home/neutral/away and the NCAA subdivision of the opponent from this link. And I've heard coaches/articles keep referring to this as a type of RPI system for college football.

Taking these items into account, I assume that the NCAA SRS goes something like this:
- Road Wins Against FCS Opponent = 1.4
- Neutral Wins Against FCS Opponent = 1.0
- Home Wins Against FCS Opponent = 0.6
- Road Losses Against FCS Opponent = -0.6
- Neutral Losses Against FCS Opponents = -1.0
- Home Losses Against FCS Opponents = -1.4

Now, what is the tweak to account for FBS and Division II teams? Note that the link says games against Division III & non-NCAA teams will not be counted. (Which definitely means Marist won out over Butler in the SRS, by the way.)

Maybe something like this?
- Road Wins Against FBS Opponent = 1.8
- Neutral Wins Against FBS Opponent = 1.4
- Home Wins Against FBS Opponent = 1.0
- Road Losses Against FBS Opponent = -0.2
- Neutral Losses Against FBS Opponent = -0.6
- Home Losses Against FBS Opponent = -1.0

- Road Wins Against D-II Opponent = 1.0
- Neutral Wins Against D-II Opponent = 0.6
- Home Wins Against D-II Opponent = 0.2
- Road Losses Against D-II Opponent = -1.0
- Neutral Losses Against D-II Opponent = -1.4
- Home Losses Against D-II Opponent = -1.8


These are guesses on my part based on the current RPI system and the references to RPI. I'm going to try my hand at calculating the NCAA SRS with these parameters. Stay tuned.

It can't work like this. It says, "The teams’ ratings are independent in that one team’s rating depends on its opponents’ ratings, which depend on their opponents’ ratings, etc., based on the “network” of college football games played each week during the football season." So a win against NDSU would be worth more than a win against someone like Idaho State, which isn't the case in your example.

 

[HelenaHandBasket]

As Dannyboy03867 has pointed out in another thread about playoff selection process, it appears that the NCAA has developed their own 'proprietary' Simple Rating System. There was already an SRS out there and has been published for at least 4 years or so, developed by Chase Stuart it would appear. Chase's SRS was based on a Margin Of Victory factor that could be calculated based upon point differential with a tweak for home/away/neutral. Each team was given a MOV and their SOS was an average of the MOV's of their opponents.

The NCAA's SRS does not take margin of victory into account, and they utilize a Win-Loss Differential instead. The SOS factor is still calculated the same way.
http://www.ncaa.org/wps/wcm/connect/public/ncaa/fcs+selections/fcs+selections" onclick="window.open(this.href);return false;

So the Win-Loss Differential is based on home/neutral/away and the NCAA subdivision of the opponent from this link. And I've heard coaches/articles keep referring to this as a type of RPI system for college football.

Taking these items into account, I assume that the NCAA SRS goes something like this:
- Road Wins Against FCS Opponent = 1.4
- Neutral Wins Against FCS Opponent = 1.0
- Home Wins Against FCS Opponent = 0.6
- Road Losses Against FCS Opponent = -0.6
- Neutral Losses Against FCS Opponents = -1.0
- Home Losses Against FCS Opponents = -1.4

Now, what is the tweak to account for FBS and Division II teams? Note that the link says games against Division III & non-NCAA teams will not be counted. (Which definitely means Marist won out over Butler in the SRS, by the way.)

Maybe something like this?
- Road Wins Against FBS Opponent = 1.8
- Neutral Wins Against FBS Opponent = 1.4
- Home Wins Against FBS Opponent = 1.0
- Road Losses Against FBS Opponent = -0.2
- Neutral Losses Against FBS Opponent = -0.6
- Home Losses Against FBS Opponent = -1.0

- Road Wins Against D-II Opponent = 1.0
- Neutral Wins Against D-II Opponent = 0.6
- Home Wins Against D-II Opponent = 0.2
- Road Losses Against D-II Opponent = -1.0
- Neutral Losses Against D-II Opponent = -1.4
- Home Losses Against D-II Opponent = -1.8


These are guesses on my part based on the current RPI system and the references to RPI. I'm going to try my hand at calculating the NCAA SRS with these parameters. Stay tuned.

It can't work like this. It says, "The teams’ ratings are independent in that one team’s rating depends on its opponents’ ratings, which depend on their opponents’ ratings, etc., based on the “network” of college football games played each week during the football season." So a win against NDSU would be worth more than a win against someone like Idaho State, which isn't the case in your example.

G-Bears, there are 2 components as noted:

"Within the NCAA SRS, the rating of a team will be calculated largely by two components: a strength-of-schedule measure (SOS) and a win-loss differential (WL)."

What MGR is referencing is the W-L differential component and not the SOS which is what you are referencing.

 

[IdaGriz01]

Okay, I've been bouncing back and forth between the two NCAA rating threads. The following statement surely refers to a team's overall rating.

The teams’ ratings are independent in that one team’s rating depends on its opponents’ ratings, which depend on their opponents’ ratings, etc., based on the “network” of college football games played each week during the football season.

Why? Because the SOS component of a team's rating does depend upon secondary relationships through the full-schedule opponent SRS averages. Conversely -- as Gym Rat suggests -- the "WL Differential" probably does not depend upon who the opponents are, just the situation of the matchup (location and division level). Of course, we do not know that for sure, but he makes a strong case. So you could just as easily say "team x" for that calculation.

So the statement, as it stands, is clearly misleading, if not just plain incorrect.

 

[G-BEARS]

Gotcha thanks