Home Inside the Numbers: How in hell is UVA on the bubble?
Sports

Inside the Numbers: How in hell is UVA on the bubble?

Contributors
basketball
(© nobeastsofierce – stock.adobe.com)

I spent otherwise valuable time combing through the numbers this afternoon to try to figure out why UVA is on the NCAA Tournament bubble, and this is my conclusion.

Something is seriously wrong with the metrics that the tournament poo-bahs use as they go about their work – NET, KenPom.com, KPI, the ESPN BPI and Strength of Record metrics, and Sagarin.

But you knew that already.

The NET, for example, as of this writing, has Virginia (18-7, 10-5 ACC) at 55.

KenPom.com, then, has Virginia at 51.

Those are not good numbers for an at-large aspirant.

How those numbers were arrived at is a mystery, as is so much about the world we live in, but that’s another story for another day.

This is the reigning, defending, undisputed champs’ resume:

Virginia: 18-7, 10-5 ACC
Q1: 3-3 Q2: 5-3 Q3: 5-1 Q4: 5-0
Road: 4-4 Neutral: 2-0
VS. NET 100-200: 4-1
VS. NET 200+: 4-0
SOS: 80 NCSOS: 191
NET: 55 KenPom: 51 KPI: 35 ESPN BPI: 37 ESPN SOR: 35 Sagarin: 30
Average rating: 40.5

Before I get too far into this, I’ll present the resumes of five teams with similar records, and then we’ll discuss.

Iowa: 19-8, 10-6 Big Ten
Q1: 8-6 Q2: 4-1 Q3: 1-1 Q4: 6-0
Road: 4-5 Neutral: 2-2
VS. NET 100-200: 4-1
VS. NET 200+: 4-0
SOS: 89 NCSOS: 219
NET: 27 KenPom: 22 KPI: 26 ESPN BPI: 24 ESPN SOR: 15 Sagarin: 22
Average rating: 22.7

Illinois: 16-9, 9-6 Big Ten
Q1: 5-7 Q2: 3-1 Q3: 2-1 Q4: 6-0
Road: 5-5 Neutral: 0-1
VS. NET 100-200: 3-0
VS. NET 200+: 6-0
SOS: 38 NCSOS: 207
NET: 33 KenPom: KPI: 44 29 ESPN BPI: 33 ESPN SOR: 25 Sagarin: 27
Average rating: 27.0

Texas Tech: 17-9, 8-5 Big 12
Q1: 2-8 Q2: 5-1 Q3: 2-0 Q4: 8-0
Road: 2-5 Neutral: 2-2
VS. NET 100-200: 0-0
VS. NET 200+: 8-0
SOS: 78 NCSOS: 172
NET: 20 KenPom: 20 KPI: 52 ESPN BPI: 19 ESPN SOR: 47 Sagarin: 23
Average rating: 30.2

Stanford: 17-9, 6-7 Pac-12
Q1: 3-5 Q2: 2-3 Q3: 6-1 Q4: 6-0
Road: 3-4 Neutral: 2-1
VS. NET 100-200: 5-1
VS. NET 200+: 6-0
SOS: 99 NCSOS: 216
NET: 34 KenPom: 40 KPI: 55 ESPN BPI: 57 ESPN SOR: 70 Sagarin: 74
Average rating: 55.0

VCU: 17-9, 7-6 A-10
Q1: 1-6 Q2: 2-2 Q3: 8-0 Q4: 6-1
Road: 4-4 Neutral: 0-2
VS. NET 100-200: 7-1
VS. NET 200+: 6-0
SOS: 46 NCSOS: 127
NET: 49 KenPom: 54 KPI: 56 ESPN BPI: 49 ESPN SOR: 63 Sagarin: 70
Average rating: 56.9

What do we see here?

Iowa is the highest-rated team in this group, with an average rating of 22.7, 17.8 points ahead of Virginia.

Head to head, the records are similar (Iowa is 19-8, Virginia is 18-7).

Strength of schedule: Virginia 80, Iowa 89.

Non-conference strength of schedule: Virginia 191, Iowa 219.

Give Iowa the edge against Q1 (8-6 vs. 3-3) and Q1 and Q2 (12-7 vs 8-6).

Keep in mind that the definitions of Q1 and Q2 are based on how NET defines what Q1 and Q2 are.

Both teams have played nine games against teams rated 100+ in NET.

Virginia is 4-4 in true road games; Iowa is 4-5.

NET values Big Ten games this year more than it values ACC games.

To wit: Iowa’s win at 12-13 Minnesota last week is a Q1 win, same as Virginia’s home win over 22-4 Florida State back on Jan. 15.

Muse on that for a minute.

(It’s driving me to drink.)

Texas Tech is third in this group with an average rating of 30.2.

I’m trying to understand how Texas Tech is 10.3 average rating points better than a Virginia team that has a better overall record (18-7 vs. 17-9), a better record against Quadrant 1 (3-3 vs. 2-8), a better record against Q1 and Q2 (8-6 vs. 7-9) and a better road record (4-4 vs. 2-5).

Just looking at those numbers, four different, important, data points, Virginia should maybe be a spot or two ahead of Texas Tech.

At the least, maybe, somehow, you say, Texas Tech is a spot or two ahead, based on whatever you want to point to.

Like, maybe, strength of schedule. Texas Tech’s is marginally better (78/172 vs. 80/191).

Texas Tech is 10.3 ratings points ahead of Virginia.

How?

Stanford is another interesting case.

Virginia has the better record (18-7 vs. 17-9).

Virginia has advantages in strength of schedule (80 vs. 99) and non-conference strength of schedule (191 vs. 216).

A better record vs. Q1 (3-3 vs. 3-5) and Q1 and Q2 (8-6 vs. 5-8).

A better road record (4-4 vs. 3-4).

Stanford is 21 spots ahead of Virginia in NET, and 11 clear in KenPom.

One last comparison: VCU.

Nobody thinks the Rams, at 17-9, at 7-6 in the A-10, having lost three straight, four of five, including a home loss to 14-12 George Mason, is anywhere near the NCAA bubble.

VCU is six spots ahead of Virginia in NET, at 49.

With a Q1 record of 1-6, a Q1/Q2 record of 3-8, a strength of schedule at 46, a non-conference strength of schedule at 127, and how the schedule is rated that strong is a mystery, because 15 of the Rams’ games are against NET 100+ opponents.

What the actual hell?

Of this group of six, Virginia would rank fourth, just from an average rating standpoint.

It’s clear from looking at the resumes side-by-side that there’s been some gaming of the system going on.

Credit to whoever it is in the conference front offices who figured things out in that respect.

We’re talking some next-level numbers manipulation.

But, bottom line here, if the committee were to leave out a defending champ with the resume that Virginia has on its side right now, we’d friggin’ riot.

Story by Chris Graham

Contributors

Contributors

Have a guest column, letter to the editor, story idea or a news tip? Email editor Chris Graham at [email protected]. Subscribe to AFP podcasts on Apple PodcastsSpotifyPandora and YouTube.