top of page
Writer's pictureRocket Haverland

"Computer Trickers"?: The Curious Case of the Big 12

Two weeks ago ACC coaches Brad Brownell and Jeff Capel claimed the Big 12 was manipulating the NET rankings. And to be honest, at first, I thought they were stupid and bitter. How could a team gain the NET ranking system? They just want ACC road wins to count as Q1 wins like they do in the B12, but are overlooking how much better the conference is. It was an erroneous claim... until I looked into it. TL/DR at the end.


(Voice of public opinion: "But why would teams want to manipulate the NET anyway? The committee looks at your resume, not your metrics.")


That is the first question that needs to be answered, why would a team want to manipulate the NET rankings? In the end, your NET ranking doesn't matter. You could have a terrible NET ranking, but you will get in if your wins look good.


Brownell and Capel, specifically, described how it is a league-wide issue. It benefits the teams you are playing. A high NET ranking means that the teams you play now have a higher quadrant win or loss. So it's not one B12 team gaining the NET, it is the entire league to make sure that nearly every game possible within conference play is a Q1 game. (I have seen a few people point out how NET rankings don't matter, and I wanted to disprove that one point).


(Voice of public opinion: "How could a team manipulate their NET ranking? It's based on their efficiency.")


Cappell and Brownell pointed to the easy B12 schedules, the majority of the B12 has.


(Voice of public opinion: "The NET is just like Kenpom, it's just an efficiency-based metric that accounts for your opponent's skill. So playing worse opponents does not inflate your Net ranking."


So scheduling an easy non-con doesn't affect it... Right? For the most part that is true. I tracked the difference in Kenpom from the NET and the average team was off about 8 spots between the two, so the philosophy behind the two rankings is nearly the same. There is no possible way to play more efficiently without being better.


Well, I think the B12 found out how to do just that. Run up the score against bad opponents. Most P6 teams when playing sub-100 ranked teams, kinda just coast after getting up 15 with 10 minutes left in the game. Look at the score differential in the two games of Purdue v. Eastern Kentucky and Uconn v. Stonehill.




These are two great teams and when playing terrible opponents the game was over by the second half. You can see the lead peaks at the 30-minute mark and from there Purdue and Uconn coast. They use up the shot clock, muck around, put in the backups, and eventually put in the practice squad.


Now look at the score differential of Texas Tech over Sam Houston St., Baylor over North Western St., and Cincinnati over FGCU.





You notice how the B12 teams just kept running up the score. The green dot at the top is when Barrtorvik declared a 100% chance of winning. Just like Purdue and Uconn the B12 won their games at around the 30-minute mark. But Purdue was content winning by 27. Baylor, on the other hand, won by 51 points!!!! Texas Tech and Cincinnati were gonna win by 20 at the 30-minute mark like Purdue, but then ran the score difference up to 40 points!


An efficiency metric will look at Cincinnati's 40-point win over a 240th-ranked FGCU as more efficient as Connecticut's 40-point win over 350th-ranked Stone Hill. The NET will account for the fact the B12 played worse opponents but it won't account for the fact they ran up the score.


It's the equivalent of Lebron and myself beating my little brother 21-0. I was just as efficient as LeBron was against my little brother, but computers can't account for the fact I had to try way harder. The only way to see if I'm as good as Lebron is to play Lebron.


The B12 made sure to have everybody besides Baylor and Kansas not schedule any good opponents. Only Baylor and Kansas played more than 2 top-40 opponents in the non-con for the B12.


(Voice of public opinion: "But other conferences don't have many top-40 non-con matchups either. The only reason Baylor and Kansas had them is because they are a big name, that's the case for the other conferences.")


The Big 12 had 19 top-40 matchups total, Purdue alone had five (and won all five). ACC (34), B10 (29), BE (22 with 3 less teams), SEC (29), P12 (29).


(Voice of public opinion: "Looking at the score differential doesn't prove the B12 was running up the score against lesser opponents. They probably still played just as well in their tougher matchups.)


Below is the B12's efficiency ranking in the non-conference, according to Barttorvik.



Now here is the B12's efficiency ranking in the non-conference against top-100 ranked teams. I will be referencing this ranking a lot, so I will use the acronym NCT100OAE (Non-conference top-100-opponent adjusted-efficiency) ranking.


(And to explain in simpler terms what this NCT100OAE is: it is the computer rankings if you only looked at games against top-100 opponents that aren't conference opponents).



There is a massive difference between the two photos. You can West Virginia, Kansas State, and Kansas which played a lot of top-100 teams had nearly the same ranking. But for the rest of the conference they were ranked 20, 30 and in UFC's case 100 spots lower.


It's a small sample size but that's the goal for the B12. If they keep this sample size as small as possible the efficiency metrics will be made up primarily of their blow-outs against terrible teams.


In fact, TCU, Cincinnati and Texas Tech only played 2 top-100-ranked opponents each. They, also, lost all six of those games. Texas Tech's best win (UT Arlington), Cincinnatti's (Georgia Tech), TCU's (Arizona State). But somehow all three are ranked in the top 45 in the NET. How is it possible that beating one of those three teams at home is a Q2 win and away is a Q1 win? It's stupid.


(Voice of public opinion: "How do we know that this is the case for specifically the B12? Every conference NCT100OAE ranking is probably much worse than its NET rankings.")


The average difference between NET ranking and NCT100OAE ranking for a B12 team was 36.4 spots. Which I feel means, that the average B12 team is 36 spots higher than where they should be in the NET. The average difference between Net ranking and NCT100OAE ranking for an ACC team is 9.5 spots; B10 (23.4), BE (-10.3), SEC (-10.5), P12 (15.9)


(Voice of public opinion: "But I'm sure that was the case for other years for the B12, they manipulated the net and still had teams go far the tournament.")


The average rank of the Big 12 in the NCT100OAE this year is 82; 2023 (46.1 average), 2022 (46.8), 2021 (70.3), 2019 (40.7), 2018 (47).


(Voice of public opinion: "That's just because they added BYU, Houston, Cincinnati, and UCF if you take out those teams it would look like past years.")


Their average does go down if you take out those teams but not my much, 78.


(Voice of public opinion: "Aside from deviation from the NET how does NCT100OAE ranking compare to other conferences? They are still the best conference.. right?")


Nope. Big 12 (82.4 average), ACC (86.1), B10 (83.4), BE (69.1), SEC (57.6), P12 (98.7).


(Voice of public opinion: "What if you only look at the top 4 teams of the conference? We still have Kansas, Baylor and Houston who had a great non-conference. Obviously, the top of the B12 is still the best of the best and will go deep.")


The average rank of the top four Big 12 teams in NCT100OAE is 21. Better than the ACC (24.5), Big Ten (23.25), and P12 (30.25) but not the BE (9.25) and SEC (15.25).


The Big 12 might legitimately be having a down year and nobody can tell because as Gary Parish would say, "they are computer trickers".



The next quetion, however, is not if the B12 is manipulating the Net, but if doing so helps their resume.


Let's say we take the rankings of B12 teams in NCT100OAE on Barrtorvik and make that their new NET.


What would BYU's resume look like?


Well in the non-con they had Q1(1-1), Q2 (1-0), Q3 (0-0) and Q4 (9-0) which doesn't change. As you can see they loaded up on those Quad 4 games.


In conference they now are Q1(1-4), Q2 (4-3), Q3 (3-1), Q4 (1-0).


For a total of Q1 (2-5), Q2 (5-3), Q3 (4-1), Q4 (10-0). That's a pretty stark difference from their current resume of Q1 (6-6), Q2 (3-3), Q3 (3-0), Q4 (9-0).


Seton Hall has a resume of Q1 (5-7), Q2 (4-3), Q3 (2-1), Q4 (8-0) which is clearly better than BYU's new resume. BYU is a five-seed in Jerry Palm's bracketology, Seton Hall is one of the last 4 in.


What are the implications of this?


  1. The Big 12 is highly overrated, in my opinion. I think as a whole the conference is just as good as the Big 10 and ACC, while the SEC and the Big East are the two best conferences.

  2. Soon all conferences will be trying to "manipulate" the NET. This means only the best teams should schedule tough games, while lesser programs will be pressured into scheduling as many easy games while running up the score in those games.


TL/DR: Is the B12 manipulating the NET? Yes. How? By having everybody but Kansas and Baylor schedule only easy games and run up the score in those games. If you look outside those games the B12 is pretty average. This means the resumes of teams in the B12 look better than they should. The B12 is overrated this year but soon every conference will do this.









55 views0 comments

Recent Posts

See All

No final four future

Painter has subjugated Purdue to a fate far worse than failure. Mediocrity. There is a clear pattern that every fan has pointed out. His...

Yorumlar


bottom of page