First NET rankings released..

[muguru]

This is an embarrassment. If this is because there isn't enough data yet, then hold this until there is..MU at 99, gotta like Notre Dame at 19 with their impressive home loss to Radford. I think the computer is broken.

https://www.ncaa.com/rankings/basketball-men/d1/ncaa-mens-basketball-net-rankings

[BM1090]

Yeah I'll wait to see where things are at in January

[TAMU, Knower of Ball]

RPI had wonky results this early too. I'm assuming not enough data

[TVDirector]

based upon this, GA Southern should blow us out of the gym tomorrow.

[Juan Anderson's Mixtape]

Interesting that they also list "previous rank" even though that wasn't previously released. 

Some pretty wild fluctuations between rank and previous rank.  I expect that will continue until conference play is well underway.

[Herman Cain]

Offensive and Defensive Efficiency are factors in the tool.  So MU has to use the cupcake games as a way to build their stats in those areas.

[Warrior1969]

based upon this, GA Southern should blow us out of the gym tomorrow.

Won't happen....we play Charleston Southern 

[brewcity77]

I don't get why they released this so soon. Anyone looking at it is going to scoff.

Also, as far as the "previous rankings" my guess is they've been tallying this all year, probably on a weekly basis, and decided to release it this week. Why they didn't wait until mid-January is beyond me. Nonsensical to put this out now.

[Jay Bee]

I don't get why they released this so soon. Anyone looking at it is going to scoff.

Also, as far as the "previous rankings" my guess is they've been tallying this all year, probably on a weekly basis, and decided to release it this week. Why they didn't wait until mid-January is beyond me. Nonsensical to put this out now.

It all makes me continue to worry that results in past years would have looked like a complete sh1t show and that's a big reason why they won't provide 'em. I think we've replaced something stupid with something else stupid. Is it better or worse? Can't fricken tell because they won't truly define what the hell it is.

[LoudMouth]

Won't happen....we play Charleston Southern 

If we don't beat Georgia Southern tomorrow then Wojo is officially on the hot seat   

[fjm]

The ESPN article on NET is actually pretty solid. Ripping it apart pretty well.

https://es.pn/2TQIAqX

[forgetful]

From Nate Silver: "These are the worst rankings I've ever seen in any sport, ever."

[UWW2MU]

From Nate Silver: "These are the worst rankings I've ever seen in any sport, ever."

Nate Silver and Matt Norlander got into it a little bit on twitter.  Found it mildly entertaining, as Norlander seemed to be a defender of the system.

[fjm]

Nate Silver and Matt Norlander got into it a little bit on twitter.  Found it mildly entertaining, as Norlander seemed to be a defender of the system.

Someone defended THIS?:

1. Ohio State (6-0)
2. Virginia (6-0)
3. Texas Tech (6-0)
4. Michigan (6-0)
5. Gonzaga (6-0)
6. Duke (5-1)
7. Michigan State (5-1)
8. Wisconsin (5-1)
9. Virginia Tech (5-0)
10. Loyola Marymount (7-0)
11. Kansas (5-0)
12. Belmont (5-0)
13. Nevada (6-0)
14. Nebraska (5-1)
15. Iowa (5-0)
16. Auburn (5-1)
17. Maryland (6-0)
18. Houston (4-0)
19. Notre Dame (5-1)
20. Purdue (5-1)
21. North Carolina (6-1)
22. Radford (4-1)
23. Pittsburgh (6-0)
24. Kansas State (6-0)
25. San Francisco (6-0)

(Kentucky #61, MU #99)

[cheebs09]

Someone defended THIS?:

1. Ohio State (6-0)
2. Virginia (6-0)
3. Texas Tech (6-0)
4. Michigan (6-0)
5. Gonzaga (6-0)
6. Duke (5-1)
7. Michigan State (5-1)
8. Wisconsin (5-1)
9. Virginia Tech (5-0)
10. Loyola Marymount (7-0)
11. Kansas (5-0)
12. Belmont (5-0)
13. Nevada (6-0)
14. Nebraska (5-1)
15. Iowa (5-0)
16. Auburn (5-1)
17. Maryland (6-0)
18. Houston (4-0)
19. Notre Dame (5-1)
20. Purdue (5-1)
21. North Carolina (6-1)
22. Radford (4-1)
23. Pittsburgh (6-0)
24. Kansas State (6-0)
25. San Francisco (6-0)

(Kentucky #61, MU #99)

I believe the defense is it's a small sample size and probably shouldn't have been released yet, but it will sort itself out with more data.

Not that I agree with it, but that's been the argument.

[oldwarrior81]

Loyola Marymount came up big at home against Central Connecticut State, pounding them by 2 points.  That alone should give them a top 25 ranking.
But when combined with the dismantling of the Westcliff University Warriors, they certainly appear worthy of a top 10.

Although they did beat Georgetown.

[SaveOD238]

Someone defended THIS?:

5. Gonzaga (6-0)

10. Loyola Marymount (7-0)

12. Belmont (5-0)

22. Radford (4-1)

25. San Francisco (6-0)

It's nice to see a little mid- and low-major representation here.  But I'm not convinced that THREE teams from the WCC deserve to be ranked right now.  LMUs best wins are Georgetown and UNLV, while SFUs are Arizona St and uh...Harvard?

[brewcity77]

Someone defended THIS?:

1. Ohio State (6-0)
2. Virginia (6-0)
3. Texas Tech (6-0)
4. Michigan (6-0)
5. Gonzaga (6-0)
6. Duke (5-1)
7. Michigan State (5-1)
8. Wisconsin (5-1)
9. Virginia Tech (5-0)
10. Loyola Marymount (7-0)
11. Kansas (5-0)
12. Belmont (5-0)
13. Nevada (6-0)
14. Nebraska (5-1)
15. Iowa (5-0)
16. Auburn (5-1)
17. Maryland (6-0)
18. Houston (4-0)
19. Notre Dame (5-1)
20. Purdue (5-1)
21. North Carolina (6-1)
22. Radford (4-1)
23. Pittsburgh (6-0)
24. Kansas State (6-0)
25. San Francisco (6-0)

(Kentucky #61, MU #99)

I will defend it to this point: It's too early to know what value it has. It's entirely possible this will be a workable, viable metric. It's possible that it will prove to be better than not just the RPI, but Pomeroy, Sagarin, Torvik, and any of the other metrics.

Right now I'm finding this very useful: http://www.barttorvik.com/trank.php?lastx=10&year=2019&hteam=Duke#Duke

That's the efficiency of all teams for the last ten games. It doesn't factor in predictive stuff like Pomeroy currently does. Frankly, at this time of year, citing a team's Pomeroy rank, or their offensive/defensive ranks at Pomeroy, is usually either ignorant or lazy. The Torvik last 10 looks at what has actually happened this season while current Pomeroy numbers are based on last year's results & player projections.

The Last 10 there is also flawed: San Francisco at #4, Kentucky at #81, there are definitely some outliers, but unlike the predictive nature of Pomeroy, it's actual data on this year's games. Which is also what we see from the NET. So I'm happy to look at it objectively because it seems to heavily weight efficiency (maybe too much so) but it's too early to know how accurate it is.   But what the last 10 (and likely the NET) has is a metric that shows Marquette is better defensively than offensively. That is unquestionably true, but predictive models like Pomeroy say the opposite.

Every model is flawed this early. Pomeroy uses too much past & predictive data, the NET overvalues victory margin, RPI heavily benefits major conference teams, whatever model you use has problems, especially after less than a month of games. My problem is putting anything out period. Wait until January at least so there's some conference games mixed in.

[brewcity77]

I just listened to Ken Pomeroy's take on his podcast. He said he thinks it was brilliant to release it now, in part because it is so inaccurate. He feels if the NCAA waited for the data to normalize, say into January, it would look similar to the other metrics and wouldn't generate the discussion it has by coming out this week.

After listening to his explanation, I'm inclined to agree, though I still want to see what the model looks like come January.

[MU_CHI]

Has anyone thought to run this after a month of last year's season and compare it to the final NET rankings? Perhaps that could shed some light on how this thing will shake out.

[brewcity77]

Has anyone thought to run this after a month of last year's season and compare it to the final NET rankings? Perhaps that could shed some light on how this thing will shake out.

The NCAA has refused to release the equation and also refused to release NET results for past seasons.

[Nukem2]

Wonder how much losing by 10+ affects the ranking?  Guess we'll never know.

[brewcity77]

Wonder how much losing by 10+ affects the ranking?  Guess we'll never know.

At some point someone will reverse engineer it, just like they did with RPI. I think the problem is that while they limit the margin of victory, they don't seem to have limited offensive and defensive efficiency. So while you might only get a 10-point "victory" margin no matter how much you win by, it still counts your 100-60 victory in 80 possessions as 1.25 ppp on offense and 0.75 ppp on defense. You benefit from a larger victory margin despite the supposed cap.

[Cheeks]

If RPI was still a thing, we would have this

1) Kansas
2) Georgia Southern
3) Tennessee
4) St. John's
5) Loyola Marymount
6) Duke
7) Texas
8) Oklahoma
9) Auburn
10) Radford
11) Ohio State
21) Wisconisn
22) K-State
33) Villanova
41) Buffalo
65) Marquette

Sagarin

MU at 28th

ESPN BPI

MU 22nd

[avid1010]

nate silver trashing it tells me more than anyone saying it just needs more data....i'm sure it will rate better as the year goes on, as it can't get worse, but hearing silver say they had highly qualified people working years on a better model that was thrown out for this one is concerning.