NET Debuts with VT at #34, Polls Love the Hokies, Computers Think VT “Bytes”

Uncategorized

With UVA bailing, the rankings changes for the Hokies were not large, mostly increasing as schools above lost. This week also brought the debut of NCAA NET Rankings. Wow, did social media blow up with those. There were some wacky things that we will cover later, but the Hokies debut at #34. This ranking is rather close to other models out there, so nothing to worry about for Hokie fans at the moment.

Human polls, except CBS, like the Hokies a lot. The human poll average is up to #18. Human polls are really starting to cluster overall, VT included. I think Andy Katz put it best when he said “the Hokies are proving their win over Villanova wasn’t a fluke.” I think that is the overall thought process in ranking the Hokies. As long as they don’t drop a bad loss, expect the Hokies to stay or climb in human polls. 

Computer polls are another story. The Hokies computer average slips to #39. Our EVTB model has them at #31. Our model accounts for margin of victory, and that is where the Hokies are falling short of being higher. Other efficiency models are really hammering the Hokies for their defensive efficiency. RPI as always is a wild right, as the Hokies fall all the way to #46.

OK, NET Rankings. As a reminder, this is the main tool for making the field of 68. VT’s rank of #34 places them as a bubble at-large team. Now, the NET rankings should be called into question at this time. Notably, Colgate is #16 after starting 1-1. That should immediately raise an eyebrow. The problem is many conferences are missing or lacking non-conference play. You are getting teams with weird SOS because of a lack of diversity in opponents. If the NCAA doesn’t account for this in the future, we may see weird results all year. Personally, I think they should have waited a few more weeks to release this. Lastly, Duke is #115 lol, which is obviously too high.

So how are we compared to NET? Not bad I say. We correlate 78% right now, which I am happy with for it being early in the season and all the schedule issues. Our model might be accounting for some of the oddities pretty well, because as much as I give shit to the NCAA for Colgate, the EVTB model has them at 5! Out model has the top 25 teams all within 7 spots of NET, except for Colgate, Minnesota and Florida. I am really interested to see how this evolves through the year. I could make predicting the field an exercise in chaos.

Virginia Tech Rankings (1/4/2021)
PollRankNotes
EVTB Ranking Model31Hokies improve slightly, but non-conference hurt SOS and MOV is low
NCAA NET34Debut rankings are… well… questionable. 2 quad 1 wins. Penn St is quad 2 loss
AP Top 2519Up 5 spots, increasing votes from 181 to 411. Exactly tied with Clemson
Coaches Poll20Up 6 spots to #20, increasing votes from 42 to 151.
Goodman Top 2517Goodman is really high on the Hokies. How can you not be with 2 top wins?
Rothstein 4517“Keve Aluma is the most underrated transfer in the country.”
Katz Power 3619Omitted last week. “The Hokies are proving their win over Villanova wasn’t a fluke.”
ESPN BPI38Slight dip from last week, but the Resume Rank climbs up to #13
CBSNo love, Clemson, NC St, Duke, Louisville all ranked. CBS is a joke.
Bracketology9 seedStrong 9 seed with games in INDY!
Kenpom42Hokies have steadily slipped in Kenpom for weeks. Hopefully not an omen.
Massey Composite37Rankings range from a high of 13 (Pugh) to a low of 76 (Stat Fox)
College Basketball Reference42Defensive ratings are killing he Hokies in computer rankings
RPI46Huge RPI slide for the Hokies
Reddit20The Internet’s Student Section poll is starting to mimic the AP. Hokies get 411 votes.
Average29Go Hokies!

*Notes about each poll we track*

EVTB Ranking Model:This is our proprietary ranking model. While it is useful to compare teams, its goal is really to project the NCAA field. It tends to be very bouncy in November and December and gets better as more games are played. It often differs from other ranking models.

NCAA NET: This is the model the NCAA uses for seeding. 

AP Top 25: The traditional poll made up of journalist votes. This poll is the most cited ranking.

Coaches Poll: The traditional poll made up of coach votes. To put it nicely, coaches are not great at ranking teams and tend to copy other polls and cluster based on recency. It’s worth noting that Mike Young does not vote in this poll.

Goodman: A single-person poll that uses no preseason or anchoring effect. It changes from day to day based on what you have done, not the name on your jersey. 

Rothstein: A single-person poll that blends preseason and actual game results.

Katz Power 36: A single-person poll that blends preseason and actual game results.

CBS: A single-person poll that blends preseason and actual game results.

Bracketology: A single-person NCAA tournament prediction. This is not a ranking. It is projecting the NCAA field based on priors and actual game results. It updates on an irregular schedule, so often can be out of date.

Kenpom: The gold standard in statistical computer models and Hokie alum. Kenpom relies on offensive and defensive efficiency to rank teams.

Massey Composite: A composite average of 25 computer models. We are lobbying to get EVTB included.

College Basketball Reference: A simple rating system based on point differential and strength of schedule. This really simple model tends to find 90% of the same result as the mega models above. 

RPI: The nearly forgotten dinosaur of rankings. We include it, because it is often cited, but pretty much everyone (NCAA doesn’t even use it anymore) agrees there are much better options out there.

Reddit: This is a weekly fan poll. We often cite journalists and computers, but why not the most important part of the game… FANS! This gives us a chance to hear their voice.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.