Big East Poll , NET Rankings and Team Sheets

[Shooter McGavin]

The formula also probably weighs wins and losses more than we think.  A loss may have dropped us significantly even with the same efficiency metrics. 

[PGsHeroes32]

Wait ... is this true?

We were favored to beat a team at home by 20+, only beat them by 5 ... and we didn't drop even a single spot in NET?

I didn't really think we'd go from 8 to 40 or something crazy like that, but I admit I thought we'd drop some. I guess I just don't understand how it works.

As stated it was more so probably our gap closed on teams behind us. But just look whose behind us. Its Creighton who just got beat badly to a poor team and fell from 5-9 only. Gonna be hard to drop a ton when you win like we did.

Also even in Kenpom we only dropped two spots from 5 to 7. With Kenpom/NEt having different formulas all the metrics seem to be evening out now. The effecency is still there even if yesterday hurt it a bit.

[MarquetteMike1977]

Wait ... is this true?

We were favored to beat a team at home by 20+, only beat them by 5 ... and we didn't drop even a single spot in NET?

I didn't really think we'd go from 8 to 40 or something crazy like that, but I admit I thought we'd drop some. I guess I just don't understand how it works.

Agree NET surprised me too. Saint Thomas went from 185 to 165. And Marquette went from a tight 8 to a looser 8. Haha.

[Carl]

Agree NET surprised me too. Saint Thomas went from 185 to 165. And Marquette went from a tight 8 to a looser 8. Haha.

The freshman 15, so to speak

[MarquetteMike1977]

The freshman 15, so to speak

LOL yes Saint Thomas lost their Freshman 15 so to speak and probably should have changed Marquette was a solid lock 8 to a not a lock 8 haha.

[Carl]

4 days for a hard look in the mirror. They'll tighten it back up for conference play

[MarquetteMike1977]

4 days for a hard look in the mirror. They'll tighten it back up for conference play

Ah yes agree!!

[Carl]

Or else they'll hear about it from their mothers over Christmas break

Okay I'll stop..

[MU82]

Thanks for the info, folks.

In the past few days I've learned that computers cream their pants and that a "bad win" often isn't all that bad to the sex-crazed machines.

[Herman Cain]

Big East NET rankings as of games of December 15, 2023
New  Old
6   6   UConn   
8   8   Marquette   
9   9   Creighton   
46   46   Villanova   
55   56   Butler
58   59   Providence   
67   67   St. John's
68   68   Xavier   
112   113   Seton Hall   
237   240   Georgetown   
266   265   DePaul

December 16 Team Sheets. MU is 8
https://www.warrennolan.com/basketball/2024/net-teamsheets-plus

[Herman Cain]

Big East NET rankings as of games of December 16, 2023
New  Old
5   6   UConn   
7   8   Marquette   
11   9   Creighton   
42   46   Villanova   
51   55   Butler   
63   67   St. John's
67   68   Xavier   
69   58   Providence   
113   112   Seton Hall   
203   237   Georgetown   
274   266   DePaul

December 17 Team sheets . MU is 7
https://www.warrennolan.com/basketball/2024/net-teamsheets-plus

[FairWeatherEagle]

I haven't looked but bump probably due to Baylor falling off a cliff.

[Milkshakes]

Why did Providence fall 11 points?

[brewcity77]

Why did Providence fall 11 points?

Failed to cover against Sacred Heart (22-point favorites), but also because teams behind them had notably better than expected results: SMU, Texas Tech, St. John's, Iowa, Texas for example. It's not just about what you do, but what the teams around you do.

[Milkshakes]

Thanks. I still find the NET confusing.  We did t get dinged for failing to cover against St Thomas.  I guess those behind us didn't over perform. 

[brewcity77]

Thanks. I still find the NET confusing.  We did t get dinged for failing to cover against St Thomas.  I guess those behind us didn't over perform.

Except that's likely not true. I have no doubt our NET took a significant hit because of that result. But because you only see the rank number, it may not look that way. You have to think of it like kenpom, where the number that matters is the Adjusted Efficiency Margin, not the rank. Using 30 as a completely arbitrary cap, NET could've looked like this before and after the St. Thomas game:

1) Arizona 30.00
2) Houston 29.85
3) Purdue 29.65
4) BYU 29.20
5) Baylor 27.80
6) UConn 27.75
7) Iowa State 27.70
8) Marquette 27.50
9) Creighton 25.25
10) Tennessee 25.20
11) Alabama 25.15
12) Indiana State 24.90

So imagine the St. Thomas result was a massive hit to their rating. Marquette wins, but loses a full 2 points in the NET. You have the following next day rankings and ratings (we'll assume none of the others changed since none of them played 12/14):

1) Arizona 30.00
2) Houston 29.85
3) Purdue 29.65
4) BYU 29.20
5) Baylor 27.80
6) UConn 27.75
7) Iowa State 27.70
8) Marquette 25.50
9) Creighton 25.25
10) Tennessee 25.20
11) Alabama 25.15
12) Indiana State 24.90

Marquette's rating changed massively, but because of the gap they had built in, their rank didn't change at all. The two aren't automatically linked. But it's very possible you could have big gaps like that. Look at some of the Pomeroy disparities between positions:

1) Houston +31.21
4) Arizona +29.55
5) BYU +25.07
8) Auburn +23.25
21) Kentucky +20.37
38) Florida +15.89
143) Longwood +1.83
205) Chattanooga -2.66

In this case, the gap between 1-Houston and 4-Arizona (1.66) is pretty similar to the gap between 5-BYU and 8-Auburn (1.82), but the one-rank gap between 4-Arizona and 5-BYU (4.48) if applied further down is a 16-rank gap between 5-BYU and 21-Kentucky (+4.70) and another 17-rank gap if applied again to 21-Kentucky and 38-Florida (4.48).

We only have a ranked picture of NET because they don't share the formula, but there are almost certainly gaps like this where the disparity between two teams might be one in rank, but the gap further down the rankings would equate to 15, 20, or more spots. Look at the 78-rank gap between Longwood and Chattanooga (4.49). That's a titanic difference in ranks for virtually the same efficiency difference as 4/5 have.

[Milkshakes]

Thanks Brew. This is extremely helpful.

[brewcity77]

Thanks Brew. This is extremely helpful.

My favorite moments on Scoop 

[lawdog77]

My favorite moments on Scoop 

I give your post a C. I expected and received detailed insight from you.

[MU82]

Great explanation, brewski.

I will say I'm still getting a kick out of the fact that even though our only game this past week was a sh1tshow win over St. Thomas, we actually moved up from 8 to 7 in the NET.

[brewcity77]

I give your post a C. I expected and received detailed insight from you.

I try to keep my target to meeting expectations. The less people expect, the less you disappoint them 

[Herman Cain]

Except that's likely not true. I have no doubt our NET took a significant hit because of that result. But because you only see the rank number, it may not look that way. You have to think of it like kenpom, where the number that matters is the Adjusted Efficiency Margin, not the rank. Using 30 as a completely arbitrary cap, NET could've looked like this before and after the St. Thomas game:

1) Arizona 30.00
2) Houston 29.85
3) Purdue 29.65
4) BYU 29.20
5) Baylor 27.80
6) UConn 27.75
7) Iowa State 27.70
8) Marquette 27.50
9) Creighton 25.25
10) Tennessee 25.20
11) Alabama 25.15
12) Indiana State 24.90

So imagine the St. Thomas result was a massive hit to their rating. Marquette wins, but loses a full 2 points in the NET. You have the following next day rankings and ratings (we'll assume none of the others changed since none of them played 12/14):

1) Arizona 30.00
2) Houston 29.85
3) Purdue 29.65
4) BYU 29.20
5) Baylor 27.80
6) UConn 27.75
7) Iowa State 27.70
8) Marquette 25.50
9) Creighton 25.25
10) Tennessee 25.20
11) Alabama 25.15
12) Indiana State 24.90

Marquette's rating changed massively, but because of the gap they had built in, their rank didn't change at all. The two aren't automatically linked. But it's very possible you could have big gaps like that. Look at some of the Pomeroy disparities between positions:

1) Houston +31.21
4) Arizona +29.55
5) BYU +25.07
8) Auburn +23.25
21) Kentucky +20.37
38) Florida +15.89
143) Longwood +1.83
205) Chattanooga -2.66

In this case, the gap between 1-Houston and 4-Arizona (1.66) is pretty similar to the gap between 5-BYU and 8-Auburn (1.82), but the one-rank gap between 4-Arizona and 5-BYU (4.48) if applied further down is a 16-rank gap between 5-BYU and 21-Kentucky (+4.70) and another 17-rank gap if applied again to 21-Kentucky and 38-Florida (4.48).

We only have a ranked picture of NET because they don't share the formula, but there are almost certainly gaps like this where the disparity between two teams might be one in rank, but the gap further down the rankings would equate to 15, 20, or more spots. Look at the 78-rank gap between Longwood and Chattanooga (4.49). That's a titanic difference in ranks for virtually the same efficiency difference as 4/5 have.

Excellent Analysis.

[PointWarrior]

But Providence had a solid win over Sacred Heart....0

Failed to cover against Sacred Heart (22-point favorites), but also because teams behind them had notably better than expected results: SMU, Texas Tech, St. John's, Iowa, Texas for example. It's not just about what you do, but what the teams around you do.

[Herman Cain]

Big East NET Rankings as of games of December 17, 2023
New  Old
5   5   UConn   
7   7   Marquette   
9   11   Creighton   
47   42   Villanova   
53   51   Butler   
64   63   St. John's
65   67   Xavier   
71   69   Providence   
95   113   Seton Hall
196   203   Georgetown   
270   274   DePaul   

December 18th Team Sheets . MU is 7
https://www.warrennolan.com/basketball/2024/net-teamsheets-plus

[Herman Cain]

Big East Poll Rankings December 18. 2023
AP
5. U Conn
6 . MU
12. Creighton
Others Receiving Votes
Providence

Coaches
5. U Conn
6. MU
14. Creighton