Bracket Tips from the Analytics Department – Vol IV


What you see above is both me finding a reason to use a clip from The Wire and the reason why this article will read a bit different this year than the last three editions.  Between the thirty second shot clock, the offense friendly rules, and the prevalent use of analytics, one may not be able to look for the same signals to find champions as they have in prior years.  The game was changed.

We’ve reviewed the last six national champions looking for consistent threads. Here’s what we’ve found (all numbers per, and thanks as always to Ken for the use of the data):

* All had top 25 defenses
* All had top 50 offenses (so balance matters than being elite on one side of the ball.  Hi Tony Shaver and Joe Mihalich!)
* All had top 30 strength of schedules
* None relied on the three ball.  33.4% of Villanova’s points during the year came from deep, by far the highest of any champion.  They did take a lot of threes, but they didn’t actually hit a stunning number
* No team lost more than four games after February 1st (including conference championship), so it doesn’t look good for Oklahoma, does it?

That’s it.  Some champions handle the ball at an elite level, others rebound, others are brilliant shooting teams.  That fluctuates, but the above criteria has been 100% for the last 6 years.  Including even that 7 seed UConn team (keep your head up Kevin Ollie).

Keeping in mind that end of year analytics are just that – they reflect the team after the tourney is over, not before, we need to leave some wiggle room as teams strength of schedule will improve during the tourney, and this years National Champion will probably end the season with better numbers than they have today largely because they are about to go 6-0.  So here’s the criteria that we will look for to find this years potential champions.:

* Top 40 defense
* Top 70 offense
* Top 50 strength of schedule
* Less than 36% of their points come from three
* No more than four losses after February 1

Here’s your list of potential champions:

Virginia (1 seed – South)
Duke (2 seed – Midwest)
Purdue (2 seed – East)
North Carolina (2 seed – West)
Tennessee (3 seed – South)
Ohio St (5 seed – West)
West Virginia (5 seed – East)

The most notable thing about this list?  Only one number one seed is on it!  And that team is UVA, who just lost sixth man De’Andre Hunter for the tournament due to injury.  So the best advise I can give you when picking a winner this year is:  Don’t pick a one seed!

West Virginia is a very fringy entry on the above list.  They are exactly 40th in defensive efficiency right now and have four February and later losses.  While I believe they are a tough out, especially on the second game of the weekend since the short preparation time for a team that plays a complex defense is a huge challenge for their opponent, I’m scratching them off the likely champions list.  Given the injury to Hunter, I’m scratching the Wahoo’s as well, which makes Tennessee my favorite from the South.  This leaves us with a solid most likely Final Four:

Duke (2 seed – Midwest)
Purdue (2 seed – East)
North Carolina (2 seed – West)
Tennessee (3 seed – South)

Going a step further, Purdue seems to have faded slightly down the stretch, and Alan Boston doesn’t think Tennessee has the horses.  So if you buy that, you can pick the most blue blood tobacco road championship match-up possible:  Duke and UNC.

Other Notes:
San Diego St is a name that I found to incredibly just barely miss the list.  They scored 36.1% of their points from three, and had the 86th toughest Strength of Schedule which kept them off of it, but that’s darned close for a mid major to get.  They swept Nevada late in the season and picked up a home win against Gonzaga in the early going.  I’m a believer.  If you’re a conspiracy theorist, it’s no surprise then that the committee is making them play Michigan in the second round (the Wolverine’s barely missed this list) and UNC in the third round.  To make a serious run just to get to the final four, they’ll have to beat two Final Four worthy teams.  It’s a joke, and really too bad for the tourneys most likely Cinderella story.  (Also, if you like Cinderella stories, you should get this.)

Lastly, remember location, location, location!  Where they play the games matter.  Fans matter!  So if you’re filling out a bracket that doesn’t have the location of the game on it you’re doing yourself a disservice.  When Michigan State plays the winner of TCU/Arizona State, it will be played in Detroit!  Don’t overthink it.  Most of all, have fun.  It’s time for some madness!


Leave A Reply

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