 |
| CBS' post-bracket reveal show may be good, BUT ESPN'S GOT JAY BILAS! from ESPN's Youtube channel. |
Late on Selection Sunday every year – after the bracket has
been announced, after the kids have been put to bed – I like to turn on a
couple of the big sports networks’ “bracketology” shows, where experts and
former players interview coaches, talk about whether they agree with the
selection committee’s decisions, and make their predictions on who will win
each match-up in the tourney.
Now, I know these folks are working on coming up with info
nuggets essentially on the fly since the bracket isn’t revealed until 5-6pm
central time, and these shows typically start shortly after. It’s tough to basically
wing it on live TV on these shows that have a lot of moving parts.
But to a large extent, college basketball followers have a
pretty good idea of who ~60 of the 68 tournament teams will be at least a day
in advance. And I am always blown away by the number of match-up predictions
where the analysis offered up is almost literally something like, “I like
Alabama, BUT GONZAGA HAS DREW TIMME, I’m going with the Zags.”
And I’m like, “……so, is this a good match-up for Timme? Why is that? Does Alabama have a weakness that Timme can exploit? Does Gonzaga do something with their offense that makes Timme particularly notable? Is Gonzaga favored simply because a certain person exists?”
In recent years, these broadcasts have added more
voices like CBS Sports’ Matt Norlander, who are capable of adding a lot more substance
to their diagnoses of “this team better than that team” as these shows’ run
times progress. But every year, I am struck by how seemingly smart analysts who
are paid to be ready to talk about college basketball…just aren’t all that
ready to give cogent analyses at this time!
 |
| PICTURED: Matt Norlander's chin, suit and laptop. From USA Today. |
This is even more of a problem for the mid-major,
double-digit seed teams who are often barely mentioned as these broadcasters
absent-mindedly swipe high seeds into the 2nd round without giving much
insight into, like, why Morehead State could threaten Illinois (or why they
probably won’t!). Part of the reason for this is that, there’s just too many
Division-I teams for even analysts paid to follow the sport to have a quickly
accessible working knowledge of how teams match up with one another in a
context they’ve never met in before.
But this brings me to the title of this year’s article: mid-major
teams – especially ones that experts think have a chance of being really good
before a season starts – are getting fewer and fewer opportunities to show they
can beat high-major teams early in the college basketball season. Sports
journalist Joe Sheehan noted this last year when he tweeted, “Every March we
get reminded that so very much of the perceived difference between the top and
middle tier of college basketball is one cohort never plays the other away from
home.” I would maybe even remove the “away from home” part of that quote, but
that’s definitely true, too.
The Clemsons and the Texas A&Ms of the world don’t want
the potential red flag on their resume of a loss to a team like Charleson or
James Madison – two mid-majors who were very good last year and presumed to be
very good again this year. So they load up their non-conference schedules with
games against other high majors that they think are of a similar caliber to
themselves and, like, a few tune-ups against Prairie View A&M and Houston
Christian.
That’s how you end up at tournament time with teams like JMU
and Grand Canyon with a combined 4 opportunities to play anybody from the most
powerful conferences all year to this point.
So it comes to now, to tournament time, and even folks who
follow the sport closely for a living have to do quite a bit of digging to
figure out just how much we should value that JMU win over Michigan State back
in November. And you end up with experts, on Sunday, immediately post-bracket
reveal, bleating out, “McNeese has that gaudy record, BUT GONZAGA’S GOT RYAN
NEMBHARD” or whatever.
 |
| OK, but like...Ryan Nembhard is actually pretty good. From Gonzaga University Athletics. |
The catch-22 they run into now is, should these good mid-majors
knock off a team or 2 in the tournament and raise their profile further, it may
help their recruiting and bring the school a nice share of those tournament
revenue bucks, but they run the risk of being even more unattractive for high-major
teams to add to their schedules next year.
I guess my plea here is simple: Give these guys a chance, Clemsons
and Texas A&Ms and – yes – the Gophers of the world. It’ll make for better
and more varied viewing opportunities for the fan in those early stages of the
season, and it’ll give us all more to talk about and break down as we get ready
to enjoy the best 3 weeks of the year in March and April.
Now, onto the stuff you came here for!
---
First, as I mention every year, I am, much like the outro
bumper says at the end of any episode of Brooklyn Nine-Nine, “Not a doctor.”
What I am is a guy who likes stats and gets a kick out of updating the Warehouse,
my collection of what is now a decade-plus worth of data points from NCAA
tournament teams. Running statistical analyses on this data has uncovered some tasty
food for thought that may help you compete in the bracket pools you enter this
week. Let's dig into what was revealed in this year's model:
1. Who Everybody’s Got vs. Who’s Elite (KenPom Top 10):
In each of the last 3 years, the top-2 most popular picks for
national champ in brackets submitted to ESPN have accounted for 35-40% of all brackets’
champions. That’s the case again this year, with UConn (24%) and Houston
(14%) taking the “national champion”
spot on 38% of America’s brackets.
In the smaller bracket pools like many of you play in, those
numbers might be even more concentrated, as ESPN’s percentages are likely diluted
somewhat by users being able to submit up to 25 brackets – and users likely leveraging
that real estate to pick a lot of different champs in those brackets.
 |
| Donovan Clingan, tall person. From Getty/The Athletic. |
Combined, the four 1-seeds account for 57% of champion picks
– and since 1979, 26 of 44 national champs (59% of them) have come from that
seed line. So you would expect people to pick 1-seeds as champs a lot, and you
would expect those one-seeds to be truly elite.
So…are they this year? Well, over the past 11 years, there
have been 18 teams that entered the tournament with top-10 ratings in both
offense and defense at the irreplaceable data site KenPom.com. Fourteen of those
18 teams (78%) won at least three games in the tournament.
There is one team who meets this criteria in this
year’s bracket.
And it’s a 4-seed. It’s Auburn.
This is crazy. Auburn won the SEC conference tournament,
they are ridiculously under-seeded, and in order to join the group of elite
teams to win 3 games in the tourney, they’ll likely have to play defending
runners-up San Diego State in the 2nd round…AND defending
champs UConn in the Sweet 16. UConn, by the way, missed out on this
double-top-10 elite distinction this year……by TWO ranking spots (their defensive
efficiency is 12th-best in the nation).
2. A Note on Your Defending Champs and #1 Overall Seed:
Most of the top statistical metrics sites have UConn
as, at worst, the 2nd-best team overall in the country this year. They
are once again really good! But they face strong historical headwinds in their quest
to repeat as champs.
Since UCLA won 7 straight titles in the late 60s and early
70s, only two teams have successfully defended a championship: Duke in 1992 and
Florida in 2007. That’s not to say they can’t do it! But this data just points
to 1. The parity in the modern D-I basketball landscape, and 2. Just how much
of a gauntlet the one-and-done tourney format is.
3. Nailing the Final Four:
So how exactly do you win a bracket pool? You do 2 things:
A. You nail as much of the Final Four as you can.
B. You gain a small advantage early on by picking the right
upsets.
If you’re in a larger bracket pool, often there is value in “zagging”
where everyone else zigs in picking their Final Four and Champs. If UConn and
Houston are upset, but you have anyone else as your champion, you’re
basically guaranteed to finish in the top half of your bracket pool just based
on that happening. (If you follow this strategy, just remember that it’s also
possible that UConn and Houston play in the final. Which, que sera sera.)
Typically, you're looking for balanced teams who have no
trouble scoring points in bunches but who still play D well enough to be able
to get some stops when they need them (which roughly equates to being top 45-ish
or better in KenPom’s defensive metrics).
Last year’s Final Four was something of a statistical anomaly,
with a 4-seed, two 5-seeds and a 9-seed making the national semis. But two of
the 4 semifinalists – UConn and 9-seed (!!!) Florida Atlantic – still met these
metrics criteria.
 |
| These guys did NOT meet the criteria. From the Palm Beach Post. |
The teams who can say they hit these criteria in each region
of the bracket this year are:
EAST: 1. UConn and 4. Auburn are the only
teams who hit the mark here (#2 Iowa State has the T-1st best
defense in the country – tied with Houston – but their offensive marks, while
not terrible, still lag behind several other teams in this quadrant).
WEST: 2. Arizona – who is another team who missed
being a double-top-10 elite team, and was even closer than UConn was (!!!) with
their 11th-ranked defense – is the only team in this region to hit
the criteria (#1 North Carolina juuuust misses the mark on the offensive
side; #10 Nevada and #11 New Mexico also just missed hitting the criteria
on offense).
SOUTH: 1. Houston and 4. Duke are the best
teams by the metrics here. #2 Marquette juuuuust misses being among the
true elite offenses. #6 Texas Tech
came close in both measures but are dealing with injury concerns, as you’ll
read below. In a reversal of what we normally see for #5 Wisconsin,
their offense meets the criteria but their defense just misses out.
MIDWEST: 1. Purdue, 3. Creighton and 5. Gonzaga
all meet the criteria in both categories. #2 Tennessee has a great
defense but their O just falls short. #4 Kansas is kind of a mess (again,
see the injuries section below.)
4. The Lil’ Conference Tournament Fun Fact:
So, above we have 8 different teams that have “the goods” to
make a nice long run into April. But 2 of these 8 have a historical obstacle to
navigate due to poor performance in their conference tournaments last week.
No team has ever won the national championship after losing
its first game in their conference tournament. Now, that doesn’t mean Creighton
(3-seed) and Duke (4) can’t make a good lil’ run in their regions. But
think twice before clicking “submit” on a bracket that has either of these two
teams as your champ.
5. Who NOT to Pick to Make Deep Runs:
So, there’s a look at the teams most likely to make deep
runs in each region. Who might we expect to NOT live up to expectations and
potentially crash out early?
One thing I’ve pinpointed the past couple years are lopsided
teams – that is, teams that are really good at one of offense or defense…but
are pretty bad at the other (sub-100th ranking in D-I metrics). Over
the past ~12 years, there have been 57 teams who fit this criteria, including
some pretty high seeds, like #3 Baylor last year, who were beaten by #6 Creighton
in the 2nd round.
These teams average about 1 tournament win per team; last
year, the 14 teams who met this criteria managed 10 total victories. Four of
those were secured by Miami, who made the Final Four, becoming just the 3rd
of these 57 teams to do so in the past 11 years (none have made the National
Championship game).
This year, there are only 4 such lopsided teams.
But the teams that skew toward good O/bad D are big ones: Kentucky
(3-seed), Alabama (4), Florida Atlantic (8 – they were a MUCH
better team defensively last year).
 |
| Reed Sheppard is great, but his team is a bunch of turnstiles on D. From Bleacher Report. |
Another few teams - including 3-seed Illinois, #7 Florida,
and #7 Dayton – all come verrry close to meeting these criteria with
their sub-84th ranked defenses.
And there’s only one bad O/good D team this year: Duquesne
(11-seed).
6. The impact of Floor Generals:
I almost dropped the Floor Generals stat from the article
this year – for the first time since I’ve started compiling the data warehouse,
only 1 out of the 4 Final Four teams had a “floor general” last year (it was a
weird year, see here). But in a typical tournament year, if you've got a steady
presence as your point guard/primary ball handler, your team may be in for a
longer stay at the tourney.
Over the past 11 years, if your team has a "floor
general" who averages at least 3.8 assists per game, you win about a
half-game more than teams who don't. And 35 of the past 44 Final Four teams (this
stat was more impressive when it was 34 out of 40 last year, haha) have had a
floor general toting the rock up the court. So having a capable ball handler
helps both the mid-level seeds get a win or 2, and the top teams achieve long
runs.
I usually bring this up to pinpoint which teams come into
the tourney without a floor general, because usually, your 1-, 2- and
3-seeds possess a player that fits this criteria.
This year's pack of notable floor general-less teams
includes: Arizona, Illinois, Auburn (what in the world do we do with these
guys!), San Diego State, Clemson, Dayton, Nebraska, Florida Atlantic, TCU, Mississippi
State, Boise State, Drake, Oregon, NC State, and James Madison.
7. This Year’s Cinderellas:
The other way you win your bracket pool is a little dash of
Fairy Godmother-style luck in picking which double-digit seeds can poke the
right holes in everyone else’ brackets except yours. Here are the double-digit
seeds that my model is pointing to as stronger than the typical team for their
seed line. Some of these have tastier 1st-round match-ups than
others.
11. New Mexico, coached by former Gophers head man
Richard Pitino and with former Gopher Jamal Mashburn Jr. as one of their stars –
is one of the highest-ranked 11-seeds I’ve ever seen in my model. Their
conference, the Mountain West, received 6 bids this year, all at relatively low
seeds except for San Diego State’s 5. And before last night’s Colorado State
win over Virginia, Mountain West teams other than SDSU had gone 0-11 in the tournament
the past 5 years. So there’s some trepidation in moving MW teams forward. But
New Mexico, at least statistically, is a stud of an 11.
 |
| Richard Pitino has defeated this many ACC teams in the NCAA tourney before. From Yahoo. |
(I’ll also note Oregon here, who – while not blowing
me away in my model – just got injured pieces of their rotation back in the last
few weeks. They blitzed through the Pac-12 conference tournament to clinch
their tourney bid, beating Arizona by 8 and Colorado by 7 in the process.
Something to note.)
12. Three different 12-seeds were absolute juggernauts in
their mid-major conferences this year: Grand Canyon, James Madison,
and McNeese State. And all 3 perform better than average for 12-seeds in
my model. GCU beat San Diego State early this season and hung tough at South
Carolina. James Madison kicked off their season with a 3-point win at Michigan
State. And McNeese State won at both VCU and Michigan by 11 apiece, AND beat the
4th 12-seed in the bracket, UAB, by 21 points early in the season.
13. Samford and Yale projected as slightly
above average 13-seeds; of the two, Samford definitely has the more appealing
match-up against injury-plagued Kansas.
14. Akron isn’t quite as good as some of the
Mid-American Conference teams we’ve seen make the tourney in the past 5 or so
years, but they’re still above average for a 14-seed.
15. Western Kentucky has a coach with tournament experience,
a fast-paced style that Marquette could have some problems with if their top-tier
point guard Tyler Kolek isn’t completely over his late-February injury.
8. Injuries + Suspensions:
The late-season injury/left-the-team bug is often something
that the tournament selection committee takes more into account than the
advanced metrics can, given teams may not have played many games without
players that get hurt late in the regular season or in their conference
tournaments. For these injuries, advanced analytics haven’t yet adjusted to
fully “price in” these absences or their potential effect on their teams'
journeys through the tournament.
These are worth knowing about and monitoring, as some of the
players on this list are key contributors for their teams, and their absences
could tip the scales in favor of their opponents if these players either aren’t
on their teams’ rosters or can't get cleared in time for their tip-offs. Here's
a list of the injuries and other departures you should know about this year,
and keep in mind: injuries to point guards and big men often hurt the most in
March. (Number denotes team's seed)
2. Marquette – Tyler Kolek (All-American level point guard,
15ppg/5rpg/7.6apg) ON THE PROBABLE SIDE OF QUESTIONABLE for 1st
round of tournament, coming back from an oblique injury.
| Marquette is being opaque about Tyler Kolek's oblique. How unique! From the AP. |
Also – Oso Ighodaro (leading rebounder, 14ppg/7rpg/3apg)
QUESTIONABLE, nursing a knee injury.
3. Baylor – Langston Love (key depth wing, 11ppg/3rpg) QUESTIONABLE
for start of tournament; managing an
ankle injury.
4. Duke – Caleb Foster (6th man,
7.7ppg/2.4rpg/2apg) OUT indefinitely with a foot injury.
4. Kansas – Kevin McCullar Jr (wing & leading scorer, 18ppg/6rpg/4apg)
OUT for tournament with a bone bruise injury he suffered 6 weeks ago and tried
to come back and play with 2 weeks ago.
Also – Hunter Dickinson (star center, 18ppg/11rpg) QUESTIONABLE
for 1st round, coming back from a dislocated shoulder.
5. St. Mary’s – Joshua Jefferson (starting forward, 10ppg/6.5rpg)
OUT for the tourney with left knee injury.
6. Texas Tech – Warren Washington (starting big man and best
defender, 10ppg/7.4rpg) QUESTIONABLE for round 1 with a foot injury.
6. South Carolina – Myles Stute (starting wing, 8.5ppg/3rpg)
ON THE PROBABLE SIDE OF QUESTIONABLE for 1st round with a hip
injury.
7. Florida – Micah Handlogten (starting center, 5ppg/7rpg)
OUT for the year, broke his leg in a fall during the SEC tournament.
9. Northwestern – Ty Berry (starting wing, 11.6ppg/4rpg) OUT
for the year with a torn meniscus.
Also – Matthew Nicholson (one of their two bigs, 5ppg/4rpg)
OUT for the tourney with leg injury.
10. Colorado – Julian Hammond III (6th man,
7.5ppg/2.5rpg) likely OUT with a knee injury.
10. Nevada – Hunter McIntosh (6th man wing, 6ppg)
QUESTIONABLE for 1st round with knee soreness.
13. Vermont – Matt Veretto (starting big man, 9ppg/3rpg) OUT
indefinitely with a shoulder injury.
9. Round 1 Best Bets:
Here in Minnesota, bills to legalize sports gambling in some
form or fashion continue to be bandied about the state legislature, but nothing
has yet been passed into law. So for the 3rd year in a row, I offer
up some betting lines for you to consider should you find yourself in a betting
state or situation, but for your humble author, these remain more of a “Boy,
would I!” than a money-making venture (also, in the last two years, these
picks have been 9-7 combined, so…profit, ROI, things of that nature. Make of
that what you will.)
Thursday
BYU -9.5 over Duquesne
Oregon +1 over South Carolina
McNeese State +6.5 over Gonzaga
St. Peter’s +21.5 over Tennessee
Friday
Texas A&M pick’em over Nebraska
Houston -24 over Longwood
TCU -4 over Utah State
Grand Canyon +5.5 over St. Mary’s
------------
That’ll do it for this year’s installment, everyone. Have a
great 1st weekend of the tournament, and may Kevin Harlan narrate your every
good dream from now until tourney’s end.
No comments:
Post a Comment