ADVERTISEMENT

NET is absurd

KennyHeisman8

Heisman Winner
Gold Member
Oct 31, 2021
8,235
7,075
113
Looking through, there are some major head scratchers.

Pitt:
Q1- 3-2
Q2- 4-4
Q3-1-0
Q4- 7-1

Texas A&M
Q1-2-4
Q2- 2-1
Q3- 3-1
Q4 8-1

They are 47, we are 61.

Oral Roberts:

Q1- 0-4
Q2- 1-0
Q3-2-0
Q4- 13-0

They are 41

It’s astonishing how ridiculous these are
 
Looking through, there are some major head scratchers.



Pitt:

Q1- 3-2

Q2- 4-4

Q3-1-0

Q4- 7-1



Texas A&M

Q1-2-4

Q2- 2-1

Q3- 3-1

Q4 8-1



They are 47, we are 61.



Oral Roberts:



Q1- 0-4

Q2- 1-0

Q3-2-0

Q4- 13-0



They are 41



It’s astonishing how ridiculous these are

This is the first year I've paid attention to it and the more I'm learning about it, the more I'm realizing they place a high value on the predictive "Vegas" metrices. As for Oral Roberts, I was stunned to see Max Abmas still there. He is one of the legit best players in the nation and for the life of me, I cant figure out why he hasnt transferred.

As for the Vegas metrices, take ACC Net for example. Convince me this isnt simply who would be favored in Vegas.

1. UVa
2. Duke
3. UNC
4. Miami
5. NC State
6. VT
7. Pitt
8. Clemson (3 Q3/4 losses)
9. Wake
10. Syr
11. BC
12. ND
13. FSU
14. GT
15. Lou

Take resume completely out of it, dont tell me that isnt pretty close to a perfect order of the teams based on who Vegas would pick in games.

But I think this is also why you dont see the committee using the team's individual NET rank. They use it to evaluate how good the team's wins and losses truly are. Example: in the RPI days, you could game it by beating, say 23-9 Liberty. Perhaps their RPI was 85. But Vegas doesnt think they are very good so their NET is 145. So you dont get the credit for being an RPI good team. You are penalized because Vegas thinks they should have been an easy team to beat.
 
Texas A&M's efficiency numbers are good partially because they beat South Carolina by 41, Mizzou by 18, Auburn by 16. They're also ahead of us in CBS's RPI and waaaaaaay ahead of us in ESPN's BPI (30 vs 70.)

IMHO this encourages you as a coach to never empty your bench and to try to press inferior teams at the end of games even to run up the score to boost your efficiency ratings.
 
Texas A&M's efficiency numbers are good partially because they beat South Carolina by 41, Mizzou by 18, Auburn by 16. They're also ahead of us in CBS's RPI and waaaaaaay ahead of us in ESPN's BPI (30 vs 70.)

IMHO this encourages you as a coach to never empty your bench and to try to press inferior teams at the end of games even to run up the score to boost your efficiency ratings.

Coaches are going to catch on real quick and start running up the score because the margin of victory clearly matters. If you stare at the efficiency sites like Tovik and Pomeroy daily and compare it to how teams are ranked and moving up and down in the NET daily because the NET is just another efficiency site like the other two, it is so plainly obvious margin of victory is very important and the NET obviously uses it to rank teams. And blowing out good teams is really important.

If we could have just 2 games back it would be the West Virginia and Michigan 30 point beatings they put on us. The Florida State loss and Vandy loss hurt as well, but not as much as the beatings the other 2 put on us and how they are hurting our efficiency metrics and our overall NET ranking because of it.

It wont be long before coaches figure it out and start running up the score.
 
TBF you probably don't get as much credit for running up the score against St. Francis since it's adjusted efficiency, right? But definitely if you manage to get a big lead against a Pitt or Michigan, you need to keep your foot on the gas.
 
TBF you probably don't get as much credit for running up the score against St. Francis since it's adjusted efficiency, right? But definitely if you manage to get a big lead against a Pitt or Michigan, you need to keep your foot on the gas.

Again though, your individual NET isnt really considered. They use it to see the types of teams you are beating. If they were using these Vegas metrics to pick the field, you'd have riots. They have to base it mostly on resume (ie what Vegas thinks about your wins).
 
TBF you probably don't get as much credit for running up the score against St. Francis since it's adjusted efficiency, right? But definitely if you manage to get a big lead against a Pitt or Michigan, you need to keep your foot on the gas.
I think margin of victory is being overstated in this thread. Pomeroy has long capped MOV so that winning by 50 is really no better than winning by 20. He also factors in the difference in rating of the teams. Beating an equal peer by 20 is a big deal but beating YSU by 30 is not.

If Pitt beats UNC tonight by 20, why should that not benefit our rating more than beating them by 1?
 
TBF you probably don't get as much credit for running up the score against St. Francis since it's adjusted efficiency, right? But definitely if you manage to get a big lead against a Pitt or Michigan, you need to keep your foot on the gas.


It matters against the good teams a lot more, because the better teams have the much higher efficiencies, the higher ratings, the higher NET and rpi, etc...

Your efficiency isn't going to move a lot when you beat a 300 level team by 30+ points and that team has bad efficiencies, its irrelevant. Your efficiency is going to move quite a bit when you beat a quality high level team by 30+ points. And even more so when you do it on the road or even a neutral location.


The rpi works in the same respect. You beat a high level rpi team, you get a big jump. Its one reason Pitt just saw a decent jump in the rpi, because Wake Forest and Miami are rated high in the rpi and we just beat both of them. If those were road games, the jump would have been even higher.


If you watch the efficiency sites daily like torvik and pomeroy, you will notice the trend. NET follows these 2 sites in the efficiency ranks.


Torvik for example has UNC favored by 6.6, which is less than the odd makers. If Pitt goes in there tonight and wins by 10-15 for example, you will see a sizeable jump in the offensive efficiency, the defensive efficiency, and the overall team ranking after the game because its also on the road. And Pitt will get a decent jump in the efficiency rankings and net ranking along with the rpi.


Torvik has Pitt winning 6 of its last 9 games for example. The lines are in place for each game already. If Pitt wants to move up in the rankings, Pitt needs to beat these teams by more than what Torvik expects the game results to be. The more points Pitt beats these teams by, the higher Pitt moves up on the efficiency sites and on NET.
 
I think margin of victory is being overstated in this thread. Pomeroy has long capped MOV so that winning by 50 is really no better than winning by 20. He also factors in the difference in rating of the teams. Beating an equal peer by 20 is a big deal but beating YSU by 30 is not.

If Pitt beats UNC tonight by 20, why should that not benefit our rating more than beating them by 1?
All fair points but since NET isn't a public formula, I do not know what they are looking at or if there are caps on margin. But even if there is, that isn't the same as a cap on efficiency in terms of total possessions vs points, right?.
 
All fair points but since NET isn't a public formula, I do not know what they are looking at or if there are caps on margin. But even if there is, that isn't the same as a cap on efficiency in terms of total possessions vs points, right?.


Even if there was a cap, lets say 20 or 25, there is still a strong correlation to scoring margin and the efficiency sites and the overall rankings. So the cap is almost irrelevant unless you are destroying good to great teams.

In other words, the more that a team beats the spread on the efficiency sites spreads (not odds makers), the more the teams overall ranking is going to be effected on the efficiency sites and on the NET.

That's why when Pitt beat North Carolina and Virginia, Pitt hardly moved on torvik, pomeroy, and on NET when it it was 2 great wins. Why? Because the efficiency metrics of a site like tovik had the spreads very small already, much smaller than the oddsmakers. And Pitt barely won. Thus, little movement in the actual ranking. Efficiency sites are not odds makers.

When Pitt significantly beats the spread on the efficiency sites actual lines on the games, then, you are going to get significant movement.


Efficiency sites don't care about wins and losses like they do about efficiency. So you could play a bunch of teams that are rated high in efficiency, lose every game by a close score, and still be rated high on the efficiency sites. The same cannot happen with the RPI, you actually have to win too to get a high rpi ranking.


Thats why if Pitt is up 20 on a great team, inserts all subs and the lead goes to 10 at games end when its already over, its a real bad idea if you want to gain ground on the efficiency sites and on the net and actually improve your ranking. The win margin against good to great teams is very important.
 
Oh Im sure they use some type of efficiency metric to create odds.

But it clearly isnt the same as what someone like Torvik is using.


Why would you say that, when the Vegas odds are so similar in almost all cases to the point spreads the efficiency sites come up with? I mean just yesterday I posted a bunch of Vegas odds versus Pomeroy numbers, and iirc the biggest difference was 1.5 points, there was only one game where it was higher than one point.

If they aren't doing something similar then there are an awful lot of coincidences going on out there.
 
  • Like
Reactions: levance2
Why would you say that, when the Vegas odds are so similar in almost all cases to the point spreads the efficiency sites come up with? I mean just yesterday I posted a bunch of Vegas odds versus Pomeroy numbers, and iirc the biggest difference was 1.5 points, there was only one game where it was higher than one point.

If they aren't doing something similar then there are an awful lot of coincidences going on out there.

For ACC games this evening,




Tonight the odds Im looking at for UNC are -9 on espn. On Torvik they are -6.6.

Tonight the odds Im looking at for NC State are -10 on espn. On Torvik they are -14.4

Tonight the odds for Georgia Tech are -2. On Torvik they are -2.9



I dont look at odds all that often outside of Pitt games. Again, I do not gamble. But there are and have been larger differences than the 1.5 points that you are suggesting. And that is for certain when it comes to Torvik lines because a gambler on this site tried calling out Torviks Pitt lines previously when I posted them against the oddsmakers and he lost all 4 games in a row he tried to take in response to me.
 
ADVERTISEMENT