New S&P+ rankings: #2 OSU, #5 PSU, #20 App State, #30 TCU

[Judge Smails]

"Penn State hasn't played anyone."

Appalachian State is better than TCU.

https://www.sbnation.com/college-football/2018/9/23/17892662/ncaa-football-rankings-2018-week-5

130-team S&P+ rankings: Penn State’s turn to try and crash the top tier
22
Updated S&P+ rankings show two of the country’s five best teams squaring off in Week 5. Can Penn State end the emerging No. 5 curse?
By Bill Connelly@SBN_BillC Sep 23, 2018, 12:07pm EDT

 

[Judge Smails]

Speaking of which, why wasn't Oklahoma punished for going to OT against Army? Stayed at #5.

 

[jimjim47]

Michigan is 6 and looking forward to PSU coming to AA.

 

[LaJolla Lion]

Michigan is 6 and looking forward to PSU coming to AA.

Good for UM, glad you guys think so highly of PSU. PSU has OSU, MSU, Iowa, and IU....all before UM so I can give 2 sh!ts about that game right now as it may be meaningless by then.

 

[Howie'81]

Michigan is 6 and looking forward to PSU coming to AA.

We'll beat the Buckeyes first and take care of you guys later.

 

[Raffycorn]

S&P is a per-play efficiency metric, so it’s not surprising that it likes us. One of the big issues with the offense in 2016 and 2017 was its inefficiency, and the creators of the metric spent a great deal of time talking about the interplay of the Penn State offense’s consistent tendency to have a few inefficient plays buoyed by a big run by Barkley or a pass for lots of yardage. The 2018 offense’s ability to get consistent yardage and stay ahead of the sticks is an improvement as far as the things that contribute to its S&P.

Also, for all of those who have spent so much time wringing hands about the defense, it’s ranked #22. I swear after hearing some posters over the past few days, you’d think we were one of the worst units in college football.

 

[LaJolla Lion]

S&P is a per-play efficiency metric, so it’s not surprising that it likes us. One of the big issues with the offense in 2016 and 2017 was its inefficiency, and the creators of the metric spent a great deal of time talking about the interplay of the Penn State offense’s consistent tendency to have a few inefficient plays buoyed by a big run by Barkley or a pass for lots of yardage. The 2018 offense’s ability to get consistent yardage and stay ahead of the sticks is an improvement as far as the things that contribute to its S&P.

Also, for all of those who have spent so much time wringing hands about the defense, it’s ranked #22. I swear after hearing some posters over the past few days, you’d think we were one of the worst units in college football.

The defense is ranked 22nd against some pretty poor competition outside of App State IMO. They have been poor, (Ill and Pitt in particular gave up huge chunks on the ground) but they are young and they will get better. Look how many frosh, sophs, and RS frosh are starting or are a big part of the 2 deep. Bad angles, wrong gaps, and just bad whiffs on tackles. Those things do get better in time, but we need a lot of young guys to grow up really fast.

 

[Judge Smails]

Michigan is 6 and looking forward to PSU coming to AA.

 

[Raffycorn]

The defense is ranked 22nd against some pretty poor competition outside of App State IMO. They have been poor, (Ill and Pitt in particular gave up huge chunks on the ground) but they are young and they will get better. Look how many frosh, sophs, and RS frosh are starting or are a big part of the 2 deep. Bad angles, wrong gaps, and just bad whiffs on tackles. Those things do get better in time, but we need a lot of young guys to grow up really fast.

One of the major advantages of the S&P+ is that it’s adjusted to correct for the relative strength of your opponents. So Penn State’s defensive S&P+ ranking already takes its opponents into account and puts our schedule on the same footing as everyone else’s (that being said, almost every team in the country has played 3 or more average at best opponents at this point so we’re hardly alone in that regard).

 

[PineIslandLion]

"Penn State hasn't played anyone."

Appalachian State is better than TCU.

https://www.sbnation.com/college-football/2018/9/23/17892662/ncaa-football-rankings-2018-week-5

130-team S&P+ rankings: Penn State’s turn to try and crash the top tier
22
Updated S&P+ rankings show two of the country’s five best teams squaring off in Week 5. Can Penn State end the emerging No. 5 curse?
By Bill Connelly@SBN_BillC Sep 23, 2018, 12:07pm EDT

In terms of how reliable and rational this ranking is, Miss St @ #9 & Kentucky @ #15 is all you need to know.

 

[Judge Smails]

In terms of how reliable and rational this ranking is, Miss St @ #9 & Kentucky @ #15 is all you need to know.

It's predictive, it's not based on resume.

 

[PineIslandLion]

It's predictive, it's not based on resume.

I understand that this is predictive. My point is, how rational is this prediction given that #15 (who is undefeated) just pasted #9, on the field?

 

[Howie'81]

The defense is ranked 22nd against some pretty poor competition outside of App State IMO. They have been poor, (Ill and Pitt in particular gave up huge chunks on the ground) but they are young and they will get better. Look how many frosh, sophs, and RS frosh are starting or are a big part of the 2 deep. Bad angles, wrong gaps, and just bad whiffs on tackles. Those things do get better in time, but we need a lot of young guys to grow up really fast.

True, but college football is a very unpredictable thing week to week. I'm of the belief that OSU is largely untested too, and also has some questions on defense.

 

[LaJolla Lion]

True, but college football is a very unpredictable thing week to week. I'm of the belief that OSU is largely untested too, and also has some questions on defense.

OSU mods on 247 who aren't your typical homers have some very serious questions about the back 7 of their own defense and the OL. There is only one team in the nation on cruise control....Bama. This game can go either way the spread kind of reflects that. OSU has a bit more success recently, but the talent gap is closing with each passing year.

Edited...Clemson is too, but the ACC will not present them with a top 10 challenge from here on out.

 

[Howie'81]

OSU mods on 247 who aren't your typical homers have some very serious questions about the back 7 of their own defense and the OL. There is only one team in the nation on cruise control....Bama. This game can go either way the spread kind of reflects that. OSU has a bit more success recently, but the talent gap is closing with each passing year.

You are EXACTLY where I'm at going into this game.

 

[mcgunns]

S&P is a per-play efficiency metric, so it’s not surprising that it likes us. The 2018 offense’s ability to get consistent yardage and stay ahead of the sticks is an improvement as far as the things that contribute to its S&P.

PSU actually ended 4th last year FYI. That team was ****ing good and could have beaten anyone. That Michigan State game is heartbreaking.

I understand that this is predictive. My point is, how rational is this prediction given that #15 (who is undefeated) just pasted #9, on the field?

Was it irrational last year to leave Georgia ahead of Auburn after Auburn won the game by 3 TDs? Was it suddenly rational two weeks later when Georgia won the rematch by 3 TDs?

 

[Grant Green]

I understand that this is predictive. My point is, how rational is this prediction given that #15 (who is undefeated) just pasted #9, on the field?

It's supposed to tell you who the better team is when playing many games against each other, not just one game.

 

[PineIslandLion]

PSU actually ended 4th last year FYI. That team was ****ing good and could have beaten anyone. That Michigan State game is heartbreaking.


Was it irrational last year to leave Georgia ahead of Auburn after Auburn won the game by 3 TDs? Was it suddenly rational two weeks later when Georgia won the rematch by 3 TDs?

Based on the collective bodies of work late in season for both Georgia & Auburn, yes, it was way more rational. But based on only 4 games in, when Kentucky has not yet lost and won decisively vs Miss State, no, it’s not rational. If Kentucky is sitting at 2-2 at this point, with one of the victories vs MSU, I realize it’s an entirely different point. But, they’re not 2-2 & they won a head-to-head match up.

 

[PineIslandLion]

It's supposed to tell you who the better team is when playing many games against each other, not just one game.

Miss St victories vs SFA, U La Lafeyette & Kansas St.
Kentucky victories vs Central Mich, Murray St & Florida, in addition to their beat down head to head vs Miss St.
I understand what a predictive ranking means, but based on results (& competition) to date, there’s 0 logic to Miss State being ranked higher than Kentucky, in any poll.

 

[PSU2UNC]

The defense is ranked 22nd against some pretty poor competition outside of App State IMO. They have been poor, (Ill and Pitt in particular gave up huge chunks on the ground) but they are young and they will get better. Look how many frosh, sophs, and RS frosh are starting or are a big part of the 2 deep. Bad angles, wrong gaps, and just bad whiffs on tackles. Those things do get better in time, but we need a lot of young guys to grow up really fast.

This is a fair point, but if you are comparing PSU to OSU, OSU has played against S&P numbers 30, 98, 107 and 118. PSU has played 20,75,102 and 120. Neither is a murderer's row, but for some reason no one is faulting OSU for giving up 31 points (in 3Q) to Oregon State.

 

[Grant Green]

Miss St victories vs SFA, U La Lafeyette & Kansas St.
Kentucky victories vs Central Mich, Murray St & Florida, in addition to their beat down head to head vs Miss St.
I understand what a predictive ranking means, but based on results (& competition) to date, there’s 0 logic to Miss State being ranked higher than Kentucky, in any poll.

It's not a poll. That is what you are missing. A poll rewards teams for their resume.
This is a power ranking. It is intended to assign value for each team, and frankly more useful for betting spreads than anything else. There are a number of factors that they use that go beyond the score of a game and the quality of the opponent.

Miss St would still be a TD+ fav over Kentucky on a neutral field if the game were played again next week.

 

[PineIslandLion]

It's not a poll. That is what you are missing. A poll rewards teams for their resume.
This is a power ranking. It is intended to assign value for each team, and frankly more useful for betting spreads than anything else. There are a number of factors that they use that go beyond the score of a game and the quality of the opponent.

Miss St would still be a TD+ fav over Kentucky on a neutral field if the game were played again next week.

I realize it’s effectively a statistical ranking ( and predictive in nature) & not a poll. While I’ve not done a deep dive on the predictive analytics they’re considering, I’ve looked at the opponents to date & the scores of these games. Based on this, and the head-to-head results, I’ll continue with my opinion that nothing which has happened the 1st 4 games should project to Miss St being higher in this ranking than Kentucky.

 

[Grant Green]

I realize it’s effectively a statistical ranking ( and predictive in nature) & not a poll. While I’ve not done a deep dive on the predictive analytics they’re considering, I’ve looked at the opponents to date & the scores of these games. Based on this, and the head-to-head results, I’ll continue with my opinion that nothing which has happened the 1st 4 games should project to Miss St being higher in this ranking than Kentucky.

Then I would suggest you find a large sum of money and go bet Kentucky at -1.5 vs SC, because if Kentucky is power rated higher than Miss St, they should be over a TD favorite vs SC.

 

[mcgunns]

Does anybody make a Vegas poll based strictly on public point spreads?

 

[Grant Green]

Does anybody make a Vegas poll based strictly on public point spreads?

Every pro handicapper has a power ranking for each team. Some guys put theirs online. You can google it, but you will have to sort through some crap (like espn and SI) to find the good stuff. Here is one that I have developed with some help from others. If you want a point spread, just subtract the two PR values and factor in home field, injuries, etc.

1 Alabama 101.3
2 Ohio State 95.7
3 Clemson 95.0
4 Georgia 93.5
5 Oklahoma 92.1
6 Auburn 89.4
7 Michigan 88.1
8 Washington 87.9
9 Penn State 86.0
10 Notre Dame 85.9
11 Wisconsin 85.3
12 Stanford 84.9
13 LSU 83.2
14 Mississippi State 82.9
15 Oklahoma State 82.8
16 Oregon 82.7
17 Texas A&M 82.5
18 TCU 82.3
19 Miami (FL) 81.7
20 West Virginia 81.5
21 Michigan State 81.3
22 Texas 80.2
23 Missouri 79.2
24 Boston College 79.1
25 UCF 79.0
26 Boise State 78.7
27 South Carolina 78.0
28 Iowa 77.6
29 USC 77.5
30 Washington State 77.2
31 Florida 77.0
32 NC State 76.5
33 Memphis 76.2
34 Arizona State 76.0
35 Duke 75.6
36 Virginia Tech 75.6
37 Texas Tech 75.2
38 Utah 75.1
39 California 75.1
40 Georgia Tech 74.9
41 Kentucky 74.6
42 Utah State 74.4
43 Houston 74.1
44 Purdue 74.0
45 Florida State 73.5
46 Syracuse 73.2
47 Vanderbilt 73.0
48 Fresno State 72.9
49 BYU 72.4
50 Maryland 72.4
51 Arizona 72.3
52 Ole Miss 72.2
53 Virginia 71.6
54 Wake Forest 71.6
55 Minnesota 71.4
56 Northwestern 71.2
57 Indiana 71.1
58 San Diego State 70.9
59 USF 70.8
60 Pittsburgh 70.8
61 Iowa State 70.4
62 North Texas 70.1
63 Baylor 70.0
64 Colorado 69.8
65 Florida Atlantic 69.7
66 Tennessee 69.0
67 North Carolina 68.7
68 Louisville 68.6
69 Nebraska 67.9
70 Toledo 67.9
71 Buffalo 67.4
72 Kansas State 67.3
73 Marshall 67.0
74 Appalachian State 66.8
75 Louisiana Tech 66.5
76 UCLA 66.0
77 Army 65.8
78 Arkansas State 65.3
79 Cincinnati 65.3
80 Eastern Michigan 64.6
81 Northern Illinois 64.4
82 Arkansas 64.3
83 Navy 63.9
84 Temple 63.7
85 Troy 63.5
86 Ohio 63.4
87 Middle Tennessee 63.1
88 Wyoming 63.0
89 Oregon State 62.7
90 Air Force 62.6
91 UAB 62.2
92 Tulsa 62.2
93 UNLV 62.1
94 Tulane 62.1
95 Western Michigan 61.4
96 Hawaii 60.9
97 Miami (OH) 60.7
98 Kansas 60.6
99 Nevada 59.9
100 Southern Miss 58.3
101 FIU 58.1
102 SMU 58.0
103 Illinois 57.8
104 Rutgers 57.0
105 Western Kentucky 56.7
106 UL-Monroe 56.3
107 New Mexico 56.1
108 East Carolina 56.0
109 Georgia Southern 55.1
110 Ball State 55.1
111 Central Michigan 54.9
112 Akron 54.9
113 South Alabama 54.8
114 Colorado State 54.8
115 Liberty 53.8
116 Coastal Carolina 53.4
117 UTSA 52.2
118 Massachusetts 52.1
119 Old Dominion 52.0
120 Bowling Green 51.8
121 Kent State 51.7
122 Georgia State 49.6
123 UL-Lafayette 49.6
124 San Jose State 48.5
125 New Mexico State 48.2
126 Texas State 47.8
127 Charlotte 47.3
128 Rice 47.2
129 Connecticut 44.9
130 UTEP 42.9

 

[PineIslandLion]

Then I would suggest you find a large sum of money and go bet Kentucky at -1.5 vs SC, because if Kentucky is power rated higher than Miss St, they should be over a TD favorite vs SC.

You continue to miss my point. To the contrary, would suggest you take some betting action this upcoming Saturday given how reliable the S&P rankings apparently are.

 

[mcgunns]

You continue to miss my point. To the contrary, would suggest you take some betting action this upcoming Saturday given how reliable the S&P rankings apparently are.

This is a mathematical system that has proven itself to be extremely accurate in the past. You have an opinion based on looking at a final game score. You have no point.

 

[Grant Green]

You continue to miss my point. To the contrary, would suggest you take some betting action this upcoming Saturday given how reliable the S&P rankings apparently are.

I understand that this is predictive. My point is, how rational is this prediction given that #15 (who is undefeated) just pasted #9, on the field?

You made your point clear.
You based your analysis of the ranking system on one data point. The system may end up getting this ranking wrong in the end. it's far from perfect and it really is just a tool. But to dismiss it based on the results of one game is not rational.