Jean-Baptiste Rochambeau
Banned
- Messages
- 2,077
Says you!! What about CavMan?
Deleted member 2897
Guest
Because? The SEC is so tough it is difficult to dominate one of their teams, even if you are a dominating team.
I know that was tongue in cheek, but the season just started, so the SEC has predominantly played non-SEC teams so far.
Deleted member 2897
Guest
Here is the Domination Ratio after Week 6.
Reminder the Domination Ratio is the ratio of Total Offense to Total Defense. That is to say, if your offense gains an average of 500 yards per game and your defense gives up an average of 300 yards per game, your Domination Ratio is 167% and you'd rank #12 in the country.
Some notes:
* Georgia Tech is now 4th in the country. If we play 100% this week (ie our offense gains what our defense gives up and vice versa), we'll slip down to maybe #10-12 in the country. Still pretty good.
* Look how far down this list folks go to make sure the SEC keeps some teams ranked. Kentucky is getting votes in the polls, and they rank #90 - their defense gives up 50 yards more per game than their offense does. On average, they are getting outplayed pretty well, but yet they get votes. #GoSEC.
* Only 3 SEC teams are actually ranked in the polls now - interesting. It mirrors this list, with only 4 SEC teams found in the top 25.
* 4 ACC teams are ranked in the polls (+ Notre Dame), but 5 here, and they are nearly all different. Georgia Tech, Duke, Louisville show up on this list, but NC State and Miami are ranked and they are not.
Reminder the Domination Ratio is the ratio of Total Offense to Total Defense. That is to say, if your offense gains an average of 500 yards per game and your defense gives up an average of 300 yards per game, your Domination Ratio is 167% and you'd rank #12 in the country.
Some notes:
* Georgia Tech is now 4th in the country. If we play 100% this week (ie our offense gains what our defense gives up and vice versa), we'll slip down to maybe #10-12 in the country. Still pretty good.
* Look how far down this list folks go to make sure the SEC keeps some teams ranked. Kentucky is getting votes in the polls, and they rank #90 - their defense gives up 50 yards more per game than their offense does. On average, they are getting outplayed pretty well, but yet they get votes. #GoSEC.
* Only 3 SEC teams are actually ranked in the polls now - interesting. It mirrors this list, with only 4 SEC teams found in the top 25.
* 4 ACC teams are ranked in the polls (+ Notre Dame), but 5 here, and they are nearly all different. Georgia Tech, Duke, Louisville show up on this list, but NC State and Miami are ranked and they are not.
Deleted member 2897
Guest
Here is the Domination Ratio after Week 6.
Reminder the Domination Ratio is the ratio of Total Offense to Total Defense. That is to say, if your offense gains an average of 500 yards per game and your defense gives up an average of 300 yards per game, your Domination Ratio is 167% and you'd rank #12 in the country.
Some notes:
* Georgia Tech is now 4th in the country. If we play 100% this week (ie our offense gains what our defense gives up and vice versa), we'll slip down to maybe #10-12 in the country. Still pretty good.
* Look how far down this list folks go to make sure the SEC keeps some teams ranked. Kentucky is getting votes in the polls, and they rank #90 - their defense gives up 50 yards more per game than their offense does. On average, they are getting outplayed pretty well, but yet they get votes. #GoSEC.
* Only 3 SEC teams are actually ranked in the polls now - interesting. It mirrors this list, with only 4 SEC teams found in the top 25.
* 4 ACC teams are ranked in the polls (+ Notre Dame), but 5 here, and they are nearly all different. Georgia Tech, Duke, Louisville show up on this list, but NC State and Miami are ranked and they are not.
The game against University-7 will be very interesting this weekend. If we lose, this ratio should fall a lot. If we win, we will finally be ranked commensurate with how we've been playing.
MWBATL
Helluva Engineer
- Messages
- 6,534
Having seen Duke play their last two games, I have to question the validity of this metric. Those may be the stats, but Duke must have racked up some ridiculous numbers against someone terrible to have such a nice domination ratio.
Suggest this be calculated using only P5 teams. Might look VERY different
Suggest this be calculated using only P5 teams. Might look VERY different
Sideways
Helluva Engineer
- Messages
- 1,589
This is true but that one SEC team would probably beat all 7 of the ACC teams. Just saying, they are really tough on defense and are well coached. The only one of the 7 ACC teams that would give them a run for their money is Clemson. Clemson is capable of beating them if the stars are aligned. The rest of them? Well, they might make it interesting...for a while.
Deleted member 2897
Guest
Well, I wasn't trying to insinuate that all those teams were better than Alabama. I was just trying to insinuate that the ACC once again is playing much better football than the SEC.
Sideways
Helluva Engineer
- Messages
- 1,589
Yep. The SEC is similar to the Big 8 back in the day, well, actually a little before my time when it was disparagingly called "Oklahoma and the 7 Dwarfs". Maybe not quite as bad as that because after all you got Auburn in the West and UGA in the East playing very good football but just about everyone else is having pretty mediocre seasons. Florida, Tennessee, LSU, and Ole Miss are in danger of having historically bad seasons. One wise guy on the LSU board said that Troy was in danger of hurting their strength of schedule playing LSU. Interesting times.
THWG
Helluva Engineer
- Messages
- 4,157
I think that LSU is in danger of hurting Troy's strength of schedule.
85Escape
Helluva Engineer
- Messages
- 1,450
LOL. But, actually, true. How can you take that much talent and squander it...and still be paid millions?
ramblinjacket
Ramblin' Wreck
- Messages
- 802
Is this metric any good at predicting W/L's? For example if a team held a 40% advantage over another team that should be a sure win right?
Also kind of surprised there is an abundance of teams over 100%. Maybe I don't fully understand the maths but that makes me pause.
Deleted member 2897
Guest
Is this metric any good at predicting W/L's? For example if a team held a 40% advantage over another team that should be a sure win right?
Also kind of surprised there is an abundance of teams over 100%. Maybe I don't fully understand the maths but that makes me pause.
I think half should be greater than 100% and half less than.
I think in the long run, the trend should be there for W/La, but our game versus Tennessee proves their are limitations sadly.
smathis30
Ramblin' Wreck
- Messages
- 732
aww heck. Ill post my computer rankings here as i have no other place.
Scores are First column, Adjusted Offense, Defense, SOS, and TOM. Stuff on the far right is the raw data. Worst, avg. and best game. "Score" is the game score needed to maintain their current rank. Projected spreads for the week at the bottom
ramblinjacket
Ramblin' Wreck
- Messages
- 802
I just thought there were <130 FBS teams but 80 of them >=100%. That's why I felt the math was suspect. Are there FCS teams in there too?Is this metric any good at predicting W/L's? For example if a team held a 40% advantage over another team that should be a sure win right?
Also kind of surprised there is an abundance of teams over 100%. Maybe I don't fully understand the maths but that makes me pause.
- Messages
- 19,555
Yep, I like this stat. Thanks for continuing to track and share.
I will point out for others that this does not include special teams, if I understand it correctly. So, when we allow someone to house a punt return (Pitt), that play not only doesn't affect the calculation but it in turn gives the ball back immediately to our offense to inflate our domination ratio. All ST and related field position issues are hidden in this.
AE 87
Helluva Engineer
- Messages
- 13,026
He's including all games, including games vs FCS teams.
ibeattetris
Helluva Engineer
- Messages
- 3,604
Yep. These can have a pretty big impact too. A team with a great def + st keeping good field position can actually limit the total yards your offense can gain (similar to how our offense limits potential yards our defense can surrender). I forget if it is FEI or SP, but one of them uses a ratio of yards gained out of total possible. So if you get the ball on the 50 and score a TD you score a 1. If you get it on the 20 and take it to the opponent 40, you get 0.5. There are then bonus points added for drives > X yards.
These advanced stats really try to focus on pure offense production while reducing noise added by defense and special team. Nothing is perfect, but I enjoy looking at all these different stats to see how different people approach looking at the game in numbers.
ibeattetris
Helluva Engineer
- Messages
- 3,604
aww heck. Ill post my computer rankings here as i have no other place.
Scores are First column, Adjusted Offense, Defense, SOS, and TOM. Stuff on the far right is the raw data. Worst, avg. and best game. "Score" is the game score needed to maintain their current rank. Projected spreads for the week at the bottom
You have an obvious bug in your code. GT isn't number 1
.All jokes aside, your spreads actually came out looking close to Vegas lines. You have the source on github?
smathis30
Ramblin' Wreck
- Messages
- 732
You have an obvious bug in your code. GT isn't number 1
All jokes aside, your spreads actually came out looking close to Vegas lines. You have the source on github?
I do it all by hand as ESPN changed the html codes so i no longer can do that. Working on it, but my work schedule is tough right now. If you want spreads on games give a holler. I always bet on games that have the highest differential from the vegas ones.
ibeattetris
Helluva Engineer
- Messages
- 3,604
Not sure what stats you use/need, but http://www.cfbstats.com/2017/national/index.html is a pretty good resource.