Stats models and rankings

roadkill

Helluva Engineer
Messages
1,815
After week 7, Colley's (https://www.colleyrankings.com/currank.html) is starting to converge on a fairly decent alignment with the AP for the top 8 teams. This poll is not predictive but it tends to have very good agreement with the AP by the end of the season.
There is some notable divergence for a few teams that traditionally get the benefit of the doubt in the AP and Coaches. :rolleyes:

Gonna post this in the CFP thread as well.

TeamColley Rank (W-L only)AP Rank
Oklahoma16
Washington25
Michigan32
Florida St44
Ohio St53
North Carolina610
Penn St77
Texas88
James Madison926
Iowa1024
Georgie141
GT66NR
 
Last edited:

Heisman's Ghost

Helluva Engineer
Messages
4,828
Location
Albany Georgia
After week 7, Colley's (https://www.colleyrankings.com/currank.html) is starting to converge on a fairly decent alignment with the AP for the top 8 teams. This poll is not predictive but it tends to have very good agreement with the AP by the end of the season.
There is some notable divergence for a few teams that traditionally get the benefit of the doubt in the AP and Coaches. :rolleyes:

Gonna post this in the CFP thread as well.

TeamColley Rank (W-L only)AP Rank
Oklahoma16
Washington25
Michigan32
Florida St44
Ohio St53
North Carolina610
Penn St77
Texas88
James Madison926
Iowa1024
Georgie141
GT66NR
UGA is number 1 until someone beats them and certainly should be at the very least in the top four. Michigan, North Carolina, and FSU all have fine teams but they should not be ranked ahead of UGA. I don't know where we should be ranked but 66th is probably as good an assessment as any for the moment. For the love of God, James Madison is ranked in the top ten and Georgia is not? I don't know who this Colley crowd is but their credibility is somewhat suspect.
 

roadkill

Helluva Engineer
Messages
1,815
Interesting. Georgia State at 28 and Miami Oh is rated higher than Miami

UGA is number 1 until someone beats them and certainly should be at the very least in the top four. Michigan, North Carolina, and FSU all have fine teams but they should not be ranked ahead of UGA. I don't know where we should be ranked but 66th is probably as good an assessment as any for the moment. For the love of God, James Madison is ranked in the top ten and Georgia is not? I don't know who this Colley crowd is but their credibility is somewhat suspect.
By way of explanation, Colley's ranking is a formula based strictly on W-L. It was the subject of a mathematical paper by Wes Colley, Ph.D. It contains no human biases, such as "It Just Means More", "Eye Test", or "They were really good last year". It assigns a value to your wins and losses based on the wins and losses of the teams you played. So at a season's start, it cannot provide meaningful data. Towards the end of the year, it provides a pretty good "deservedness" ranking for the CFP.

Miami now has 2 losses, one of them to a team ranked 66th.

I don't know if you've examined Georgia's schedule so far this year, but Colley ranks it 137th so far. They have beaten no currently ranked teams. Their best win was against Kentucky which Colley currently has ranked at 47th. Also, aside from Kentucky, they have struggled to beat mediocre-to-bad teams. Personally, I would put them in the top 10, but not the top 4 right now based on their struggles. Their schedule gets a little tougher from here on out so it will give us better data. In contrast, Michigan has dominated everyone they've played.

For us, 66th seems about right with a 3-3 record. Right in the middle of FBS. We've beaten and lost to some decent teams. The Miami win is offset by the BG loss.
 

slugboy

Moderator
Staff member
Messages
11,472
Here’s a repost from the CFP thread (which got into rankings, and some of this). We’re not in the top 30 anywhere, but FSU and Duke—yes, Duke—show up well:

Here’s a different resume system—it’s Kelly Ford’s “most deserving”. It’s not a prediction—it’s who has played the best against their schedule

1-Most%20Deserving-3ec5f13.PNG


And here’s the weekly progression (you can see “week 0”, where only the teams that played an early game got rated)

2-Progression-6bc2305.PNG
 

slugboy

Moderator
Staff member
Messages
11,472
F+, as of week 7, would have BC and UVA games as pretty even, and would favor VT over us.

Rk
Team
Rec
FBS
FEI
OFEI
Rk
DFEI
Rk
ELS
Rk
GLS
Rk
ALS
Rk
76​
Louisiana​
4-2​
3-2​
-.11​
.30​
31​
-.46​
106​
.05​
129​
.27​
125​
1.47​
123​
77​
Marshall​
4-2​
3-2​
-.11​
-.39​
87​
.24​
53​
.05​
131​
.39​
120​
1.80​
114​
78​
Houston​
3-3​
3-3​
-.12​
.08​
44​
-.38​
100​
.19​
92​
1.03​
80​
2.72​
69​
79​
Florida Atlantic​
3-3​
2-3​
-.13​
-.53​
95​
.15​
59​
.19​
91​
.83​
98​
2.27​
92​
80​
Georgia Tech​
3-3​
2-3​
-.13​
-.13​
63​
-.01​
77​
.29​
74​
1.33​
57​
3.00​
54​
81​
Northwestern​
3-3​
2-3​
-.16​
-.46​
92​
.04​
71​
.61​
29​
1.76​
28​
3.09​
52​
82​
UTSA​
3-3​
3-3​
-.18​
-.12​
62​
-.34​
98​
.29​
73​
.83​
97​
2.05​
102​
83​
Georgia Southern​
4-2​
3-2​
-.19​
-.09​
58​
-.42​
102​
.14​
97​
.72​
105​
1.99​
105​
84​
Utah State​
3-4​
2-4​
-.21​
-.02​
55​
-.49​
109​
.28​
75​
1.42​
48​
3.27​
41​
85​
Boston College​
3-3​
2-3​
-.23​
-.16​
64​
-.37​
99​
.41​
58​
1.19​
71​
2.57​
74​
86​
Miami (OH)​
6-1​
5-1​
-.23​
-.79​
113​
.14​
60​
.11​
107​
.53​
113​
1.55​
120​
87​
Arizona State​
1-5​
0-5​
-.23​
-.68​
102​
.21​
56​
.24​
81​
1.32​
59​
3.24​
44​
88​
Toledo​
6-1​
5-1​
-.23​
-.24​
71​
.09​
65​
.06​
128​
.20​
129​
1.07​
127​
89​
Virginia​
1-5​
0-5​
-.24​
-.35​
82​
-.10​
85​
.38​
62​
1.54​
41​
3.31​
38​
 

ibeattetris

Helluva Engineer
Messages
3,604
Spent some time just poking around bcf toys today and some of the numbers are crazy.

There are still 4 school averaging over 4 ppd at this point in the season. Michigan, Washington, Oregon, Airforce
3 schools are still averaging less than 1 point per drive on defense. Michigan, Penn State, Duke (honorable mention to OSU and Iowa at 1.02 and 1.06 respectively).

Michigan's schedule outside of OSU and PSU is a complete joke, but they are completely dismantling everyone they have played

Iowa is insane. They might be the worst offensive team to ever have a chance at an 11-1 season. Coming in at 1.44 PPD and a brutal touchdown rate of 14.3%, they are still winning games due to having a defense limiting teams to 1.06 ppd. Combine that with an incredibly weak SOS and they might be cruising to double digit wins while barely averaging double digit points per game.

GT is currently rank 80 in FEI. Offense has been slipping some but defense has been improving. The one defensive stat we are truly putrid in is "Available Yards". The stat tracks what percent of available yards the defense gives up on average. So if a drive starts on the opponent 20, and they drive 40 yards, the Available Yards would be 50%.

GT is currently 123rd and surrendering 57% (the 70th ranked team is 47%). I doubt our team can shift from its bend but don't break nature in one bye week, but this is the metric I want to see improved most.
 

bobongo

Helluva Engineer
Messages
7,544
The one defensive stat we are truly putrid in is "Available Yards". The stat tracks what percent of available yards the defense gives up on average. So if a drive starts on the opponent 20, and they drive 40 yards, the Available Yards would be 50%.
Available yards? Who came up with that one? Honestly, of all the pointless stats (pun intended), that has to be one of the most useless and misleading I've ever seen.
 

slugboy

Moderator
Staff member
Messages
11,472
Spent some time just poking around bcf toys today and some of the numbers are crazy.

There are still 4 school averaging over 4 ppd at this point in the season. Michigan, Washington, Oregon, Airforce
3 schools are still averaging less than 1 point per drive on defense. Michigan, Penn State, Duke (honorable mention to OSU and Iowa at 1.02 and 1.06 respectively).

Michigan's schedule outside of OSU and PSU is a complete joke, but they are completely dismantling everyone they have played

Iowa is insane. They might be the worst offensive team to ever have a chance at an 11-1 season. Coming in at 1.44 PPD and a brutal touchdown rate of 14.3%, they are still winning games due to having a defense limiting teams to 1.06 ppd. Combine that with an incredibly weak SOS and they might be cruising to double digit wins while barely averaging double digit points per game.

GT is currently rank 80 in FEI. Offense has been slipping some but defense has been improving. The one defensive stat we are truly putrid in is "Available Yards". The stat tracks what percent of available yards the defense gives up on average. So if a drive starts on the opponent 20, and they drive 40 yards, the Available Yards would be 50%.

GT is currently 123rd and surrendering 57% (the 70th ranked team is 47%). I doubt our team can shift from its bend but don't break nature in one bye week, but this is the metric I want to see improved most.
For “available yards”, we have games like Ole Miss, Bowling Green, and Louisville to dig out from. We even gave up too many yards to SCST (if they aren’t excluded from the stats).
That stat will be horrible all year, even if we’re above average for the rest of the season.

Miami and Wake might be the two games we played good defense, and I’m not that sure about Wake. Miami got a ton of yards, but we tackled. Was that game a fluke?
 

slugboy

Moderator
Staff member
Messages
11,472
Available yards? Who came up with that one? Honestly, of all the pointless stats (pun intended), that has to be one of the most useless and misleading I've ever seen.
The idea is that giving up 8 yards from your opponent’s 20 means a lot less than giving up 8 yards from your 8 (or even your 30).

Points matter a lot more, but having the other team march down to your side of the field can wear you out. Ole Miss might be an example of where we bent until we broke (or worse)
 

bobongo

Helluva Engineer
Messages
7,544
The idea is that giving up 8 yards from your opponent’s 20 means a lot less than giving up 8 yards from your 8 (or even your 30).

Points matter a lot more, but having the other team march down to your side of the field can wear you out. Ole Miss might be an example of where we bent until we broke (or worse)
It's a superfluous mirror of time of possession. In that, we're # 116:

 

yeti92

Helluva Engineer
Messages
3,030
GT is currently rank 80 in FEI. Offense has been slipping some but defense has been improving. The one defensive stat we are truly putrid in is "Available Yards". The stat tracks what percent of available yards the defense gives up on average. So if a drive starts on the opponent 20, and they drive 40 yards, the Available Yards would be 50%.

GT is currently 123rd and surrendering 57% (the 70th ranked team is 47%). I doubt our team can shift from its bend but don't break nature in one bye week, but this is the metric I want to see improved most.
I'm most interested in seeing the metric called "Wins" improving myself.
 

ibeattetris

Helluva Engineer
Messages
3,604
For “available yards”, we have games like Ole Miss, Bowling Green, and Louisville to dig out from. We even gave up too many yards to SCST (if they aren’t excluded from the stats).
That stat will be horrible all year, even if we’re above average for the rest of the season.

Miami and Wake might be the two games we played good defense, and I’m not that sure about Wake. Miami got a ton of yards, but we tackled. Was that game a fluke?
Yeah this is pulled from bcftoys and removes ScSt. I don't imagine it would be much improved with their inclusion.

Available yards? Who came up with that one? Honestly, of all the pointless stats (pun intended), that has to be one of the most useless and misleading I've ever seen.
I am trying to imagine a scenario where someone presents me with novel information I don't understand, and instead of asking questions to widen my knowledge, I say "that has to be one of the most useless and misleading I've ever seen." I'd wager the only thing misleading would be your own understanding.

It's a superfluous mirror of time of possession. In that, we're # 116:

Ah yes, I was correct.

Let's say you have an opponent start at their 25 yard line in two separate drives. On one drive, they drive all the way down to your two yard line and fail to convert. No point, you take over at your two. In the other drive, they drive to your 40 go for it, and fail, no points. Both drives have led to 0 points, however, one drive led to an additional 38 yards of flipped field position.

@slugboy had another good example. Let's say you have one drive start at you 25 and it goes 20 yards. You are now punting and you captured 26.7% AY. You have another drive that starts at the opponent 50. You go 20 yards captured 40% AY and went from out of FG range to inside of FG. Those 20 yards will have taken the same time of possession, but one is worth way more than the other.

While you might be under some notion that all drives ending in 0 points are the same, that is emphatically false. Capturing available yards is a core part of flipping fields and scoring points. An offense that consistently flips fields coupled with a defense that prevents the opposing offense from doing the same will win more ball games than others. It is easy for a human to look at a drive and say "hey the offense didn't score, but look the forced the other team to start deep within their own territory". This is one single stat that helps capture this.

There is also the fact that "total yards" is a stat that is affected by a lot of things like number or drives and starting field position. GT under Paul Johnson in 2014 was 17th in total yards. We also ran the ball a lot and therefore on average had less drives per game than average. When considering available yards gained per drive, GT was second in the country.
We can even see some interesting examples of this in action
WkOpponentRFinalNG FinalDrAYDYPlPtTOSFPOVODEOPDOAYOPP
6MiamiW28-1728-1775113567028-7314.12.0240.6975.09
8North CarolinaL43-4843-48118305846743-175.522.42.043.910.7048.72
9PittsburghW56-2849-21126975036449-158.117.61.474.080.7227.86
The UNC game had the most yards by far. But look at the Miami game. We "only" had 356 yards, but we also only had 7 drives (BCF ignores the final drive as it was garbage time). The Pitt game had less yards by 80 in more drives, but we had better field position on average (58 AY per drive vs 75.5), so the total available yards for the game were less than vs UNC.
 

ibeattetris

Helluva Engineer
Messages
3,604
For “available yards”, we have games like Ole Miss, Bowling Green, and Louisville to dig out from. We even gave up too many yards to SCST (if they aren’t excluded from the stats).
That stat will be horrible all year, even if we’re above average for the rest of the season.

Miami and Wake might be the two games we played good defense, and I’m not that sure about Wake. Miami got a ton of yards, but we tackled. Was that game a fluke?
This is the defensive game splits so far this year:
WkOpponentRFinalNG FinalDrAYDYPlPtTOSFPDVDDEDPDDAYDPP
1LouisvilleL34-3934-39139354816339171.9-4-0.3130.5147.63
3Ole MissL23-4823-41106304605441-63-14-1.44.10.738.52
4Wake ForestW30-1630-161182542080164759.70.881.450.5095.25
5Bowling GreenL27-3820-3885483646231-68.5-12.6-1.573.880.6645.87
6MiamiW23-2023-20138704448220566.913.61.041.540.515.41
The Ole Miss And BG games are definitely killing, but in the remaining games we are still averaging ~51% which would be good for 92nd.
Can definitely see how the blessed TO stat is pretty darn important for limiting points when you are giving up a lot yards.
 

JacketOff

Helluva Engineer
Messages
2,951
It's a superfluous mirror of time of possession. In that, we're # 116:

It has literally nothing to do with time of possession, and I have no idea how you even came to that conclusion. Case in point: we held the ball for 40 minutes in the Ole Miss game, yet they out gained us by 80 yards.
 

bobongo

Helluva Engineer
Messages
7,544
Let's say you have an opponent start at their 25 yard line in two separate drives. On one drive, they drive all the way down to your two yard line and fail to convert. No point, you take over at your two. In the other drive, they drive to your 40 go for it, and fail, no points. Both drives have led to 0 points, however, one drive led to an additional 38 yards of flipped field position.

@slugboy had another good example. Let's say you have one drive start at you 25 and it goes 20 yards. You are now punting and you captured 26.7% AY. You have another drive that starts at the opponent 50. You go 20 yards captured 40% AY and went from out of FG range to inside of FG. Those 20 yards will have taken the same time of possession, but one is worth way more than the other.
Or, say you got the ball at the 50 and drove down to the one, where you left it on fourth down. You gobbled up 98% of your available yardage, and true you left them deep in their own territory, but you put zilch on the scoreboard, where it counts. My point is, that extra 2% you would have gotten had you scored barely moves the needle on available yardage but makes all the difference as to who wins the game. So it's 98% on the stat, but nothing on the board. Where it counts.
 

bobongo

Helluva Engineer
Messages
7,544
It has literally nothing to do with time of possession, and I have no idea how you even came to that conclusion. Case in point: we held the ball for 40 minutes in the Ole Miss game, yet they out gained us by 80 yards.
One game can produce all kinds of anomalies. Our ranking is low in both stats. How long you keep the ball and how much available yardage is gained are closely related on the whole.
 

roadkill

Helluva Engineer
Messages
1,815
I can see both sides of the argument over the utility of this stat. (After I finally overcame my semantic confusion over y'all's use of the term "available yards", which is just a function of starting field position x # of drives, when you actually mean the ratio of actual yards to available yards (DAY) as a defensive metric, or OAY as an offensive one.)

One can think of numerous examples which either support the utility of this metric or undermine it, simply by moving the starting field position. Case in point, in @ibeattetris example of a team that started from their 50, gained 20 yards, kicked a field goal. What if they actually started at their opponent's 25 (due to a blocked punt, for ex.) and did the same? Most would laud the other team's D as "winning" this drive by holding to a field goal, despite the fact that the team on offense gained 80% of available yards and scored 3 points.

It seems to me that a successful "bend but don't break" defensive scheme would normally have a relatively high DAY, which undermines the concept of using DAY to evaluate defensive quality. On the other hand, it seems like it would be somewhat useful in comparing teams with elite defenses that force a lot of 3-and-outs, or elite offenses that score a lot. But wouldn't ppd be better for that?

I think the problem may be that many advanced stats need to be used in the proper context, rather than in isolation. That's probably why the rating systems that use them generally combine a number of advanced stats into a formula to yield a single number for ranking purposes.
 

babuka

Georgia Tech Fan
Messages
43
Is there a method similar to Colley that starts everyone on equal level and only uses the final score?
 

JacketOff

Helluva Engineer
Messages
2,951
One game can produce all kinds of anomalies. Our ranking is low in both stats. How long you keep the ball and how much available yardage is gained are closely related on the whole.
The top 10 teams in time of possession this year:

RKTeamGTotalAvg
1Air Force620834
2Penn St.620734
3Liberty724034
4Utah620534
5Florida723834
6Georgia723533
7West Virginia620133
8Ohio723433
9Stanford619632
10Cincinnati619632


The top 10 in total offense:
RANKTEAMGPLAYSYDSYDS/PLAYYPG
1Oregon642933207.74553.3
2LSU748738537.91550.4
3Washington638432628.49543.7
4UCF641431007.49516.7
5Georgia750035667.13509.4
6Oklahoma645230366.72506.0
7Miami (FL)641130117.33501.8
8North Carolina646530086.47501.3
9Southern California744434427.75491.7
10Texas St.750534276.79


And the top 10 in scoring offense:
RANKTEAMGTDSFGPTSPPG
1Oregon6387291.0048.5
2Southern California7455331.0047.3
3LSU7419317.0045.3
4Oklahoma6358271.0045.2
5Penn St.6357266.0044.3
-Washington6364266.0044.3
7Florida St.6337253.0042.2
8Ole Miss63111250.0041.7
9Georgia73512281.0040.1
10Michigan737276.0039.4

As you can see, there is very little to zero correlation between time of possession and offensive potency. Only 1 team in the top 10 of TOP is also in the top 10 of either total offense or scoring offense: Georgia. The available yards stat is essentially an offensive (or defensive) efficiency metric that shows you how well your offense moves the ball on each drive, or how poor of a job your defense does at preventing the other team from moving the ball. It’s pretty simple really.

Of the top 10 TOP teams, here’s where they rank in the offensive available yards stat:
1. Air Force -> #7 OAY
2. Penn State -> #26 OAY
3. Liberty -> #9 OAY
4. Utah -> #105 OAY
5. Florida -> #36 OAY
6. Georgia -> #5 OAY
7. West Virginia -> #67 OAY
8. Ohio -> #55 OAY
9. Stanford -> #50 OAY
10. Cincinnati -> #49 OAY

In today’s era of college football, a high TOP is usually a signifier of an inefficient offense not capable of producing big plays.
 
Top