123: Hourglass of WAB
A look at available WAB for college hoops, plus some UNC thoughts.
As the calendar turns to January, the men’s college basketball schedule transitions to conference play. In fact, there is only one non-conference game remaining for rest of the regular season1 out of a possible 3,078 total games between D-I teams.
The 18 teams in ACC men’s college basketball have already played 42 conference games. SMU and Clemson are a perfect 3-0 in league play.
The 16 teams in the SEC have yet to play any conference games. Tomorrow or Saturday, January 4 will be the first day of league play with eight games scheduled.
Now that we’re in league play, any chances a team has to improve its résumé is going to happen in a closed loop system. Any transfer of credibility between conferences is all but over.
Because any quality wins or damaging losses now happen within the confines of the conference. And as conferences continue to expand, league schedules are unbalanced and not all teams have identical opportunities.
So, which leagues are well-positioned to take advantage of the feedback loop that is conference games only?
One way to visualize this is using WAB or wins-above-baseline. As a reminder, WAB is a résumé metric that reflects how many more (or fewer) games a team has won against its schedule than a bubble-quality team would be expected to win against that same schedule.
Any loss or win is assigned a value between -1 and 1. A win is a positive value and a loss is a negative value. The values are location adjusted, for example, a road win is often worth more than a home win and a home loss is often more damaging than an away loss.
Think of it like an hourglass showing the amount of time a team has left to improve its résumé. Here is a breakdown of the top six leagues in college basketball as of today.
Each conference has a shape and the SEC’s shape is a pristine hourglass. A win at Auburn is worth 0.93 WAB+. The “worst” loss would be at home against South Carolina or worth -0.77 WAB-.
The 16 teams in the SEC only play 18 league games apiece too. And the SEC won ~89% of its non-conference games to create a massive advantage in league play.
The same cannot be said about the other listed leagues. Here are the “best” wins and “worst” losses in league play as of today for the other five listed leagues:
Big 12: 0.93 WAB+ at Houston and -0.85 WAB- at home against Oklahoma State
Big 10: 0.86 WAB+ at Illinois and -0.90 WAB- at home against Minnesota
Big East: 0.83 WAB+ at Marquette and -0.93 at home against Seton Hall
ACC: 0.92 WAB+ at Duke and -0.92 at home against Boston College
Mountain West: 0.68 WAB+ at Utah State and -0.98 WAB- at home against Air Force
We can gain some clarity if we take a closer look at comparing two teams. For example, take the two Carolinas.
North Carolina is 8-6 overall, 46th in the NET, and around 60th in WAB as of today.
South Carolina is 10-3 overall, 85th in the NET, and around 78th in WAB as of today.
The Gamecocks have an advantage playing in the SEC. It’s easier to gain “quality” wins and harder to pick up “bad” losses. I think it’s fair to say UNC would much rather play South Carolina’s schedule than its own, even if it’s tougher.
While UNC does have the benefit of playing rival Duke twice, there is much more room for the Heels to gather “bad” losses. Any home loss to Boston College (-0.92), California (-0.89), Miami (-0.88), Virginia (-0.83), or even NC State (-0.81) could torpedo UNC’s résumé.
Imagine each league has its own currency. The exchange rate for the ACC or any league compared to the SEC is not good.
Convenient comparisons
While we’re on the topic of North Carolina basketball, I’ve read and heard comparisons of Hubert Davis to Matt Doherty and even John Bunting as of late.
I understand all three played and attended the university, and all three became head coaches, but I find this to be too convenient of a comparison.
Davis has coached 27 more games than Doherty, and won 70 percent of those games compared to Doherty’s win percentage of 55. Did we forget that Doherty coached a team that lost to the EA Sports All Stars by 31 points in an exhibition2!
Bunting coached a completely different sport, but I won’t spend a lot of time with this comparison given Bunting’s sub-400 win percentage.
I do think Carolina basketball is in a tricky spot right now. And of course, criticism is warranted for the head coach because that’s how it goes.
The team simply looks unmotivated. I can’t quantify that, but the chemistry or respect between teammates or decision making is flat out poor. This clip from the loss at Louisville that ends in a live-ball turnover sums it up3.
Right now, I think fans and media are clinging to things that are stylistically different with Carolina basketball. Yes, the offensive rebounding is no longer a part of its identity.
It’s correct to point out the offensive rebounding rate is the lowest it has been since an 8-20 season in 2001-02. It’s also unfair to not share that the team’s adjusted offensive efficiency is better right now than it was in six different seasons under Roy Williams.
The style can be different. But if the results aren’t there, yes it becomes a problem.
My half-baked analysis is I struggle to understand how Carolina beats any good team on a poor shooting night. One of the advantages of offensive rebounding is that it often leads to higher percentage shots and more chances to score.
If turnovers and rebounds are the dials of shot volume, I’ve yet to see how UNC finds an advantage in shot volume against quality teams4.
Carolina has not prioritized forcing turnovers in any season under Hubert Davis. Take a look at this clip where a pass across the court ends in a clean three-point shot. The shot is missed, but there is no deflection or even attempt to steal the ball.
The Heels have ranked in the 300s in every season in defensive turnover rate under Davis, which doesn’t mean the defense is “bad” every season, but it does mean teams can gain an advantage in shot volume.
For example, consider the following formula for scoring chances in the loss at Louisville using shot attempts, free throw attempts, and turnovers. This weights three-point attempts as worth more than twos, and weights free throw attempts to count for the “trips” in a crude way or using the .475 multiplier.
(3PTA * 1.5) + 2PTA + (.475 * FTA) - TO Louisville attempted 26 three-point shots, 32 two-point shots, 39 free throws, and committed eight turnovers.
North Carolina attempted 16 three-point shots, 44 two-point shots, 23 free throws, and committed 12 turnovers.
Louisville’s scoring chances are 81.5 compared to UNC’s 66.9.
This is independent of making shots. This is simply possessions where a team has a chance to score.
It’s a tough needle to thread when your offense has to be super efficient every night to win. There is a lot of variance in shooting, especially in single elimination tournaments.
Anyhow, the comparison might be convenient, I just hope it’s wrong.
Happy New Year, and thanks for reading this far. I’m continuing to make updates to this tiny app: byc.evidence.app
I’ve removed the college football data as the bowl season is coming to an end. I’ll likely revisit some of that data prior to the championship game later this month. For example, how funny is it that Michigan did not hit its preseason win total despite beating both Ohio State and Alabama?
Find the code for the charts in this post and other posts here. Oh, and please subscribe below if you so choose.
Illinois and Duke in Madison Square Garden on February 22.
Shoutout Curtis Staples. Incredible box score.
This was actually part of three consecutive turnover sequence. Cadeau bad pass, moving screen by Louisville, and Cadeau travel. Just a masterclass is bad basketball and questionable officiating.
It’s very important for this team to get to the foul line. The problem is there isn’t a post presence that actually does get to the foul line all that much. Seth Trimble has attempted more free throws (57) than Jalen Washington and Jae’Lyn Withers combined (52).