In building my (mostly incorrect) bracket this year, there were two key missteps in my process. Firstly, I talked myself into a flawed contender as the champion. Michigan State was an offensive juggernaut, a team that ranked first offensively by Ken Pomeroy's metrics. Even factoring in the four games that Denzel Valentine missed, they graded out at the top in offensive efficiency. But their offensive prowess masked some of their subtle weaknesses (more on this later). Hindsight is 20/20, and looking back at past precedents, Sparty does not profile as a team that was capable of winning six consecutive tourney games.
My second misstep is more of a character flaw than anything. I wanted to be the smartest guy in the room with my bracket. And putting Wichita State in the championship game would certainly validate my analytics-based approach if they indeed got there. In reality, it is foolish to use analytics as a means of predicting March Madness perfectly. The tournament remains an arbitrary exercise, one that would need to be played thousands of times over to prove the efficacy of analytics. But I ignored all of that and decided to put Wichita State in the championship game, along with some other unlikely upsets that would not bear fruit (St. Joes over Oregon, anyone?). So I lost both halves of my championship game by 2:30 PM on the third day of the tournament. Now, what else have we learned from the first two rounds of games?
Size Doesn't Matter, Rim Protection Does
This was not hard to notice watching the UNI v. Texas game, or the aforementioned Wichita State v. Miami game. Both favorites were equipped with a formidable interior defender, Prince Ibeh and Tonye Jekiri, respectively. While neither player recorded gaudy block totals, their impact was undeniable. Able to challenge any interior shot, their presence alone caused the opposing offense to stagnate. Northern Iowa's offense was stuck idling during the 20 minutes Ibeh was on the floor.
Ibeh's positive effect was exemplified over the final 5 minutes, where Northern Iowa managed just two made field goals, one of which was from 42 feet away. Their other possessions ended in missed shots, turnovers, or FTAs. Ibeh had 3 blocks despite being played sparingly, and Northern Iowa made only 13 two-point FGs, below their season average.
Against the Shockers, Jekiri played 27 minutes, disrupting Wichita State's offense the entire time. Though he finished without a block, film study and numbers affirm his positive impact. Saddled with foul trouble for part of the second half, Wichita State's comeback occurred while he sat helplessly on the bench. He was subbed back in with 9:25 to go, with Miami having just regained a 44-43 lead. From there, Wichita State did not make a two point basket until the 2:43 mark. Miami was able to escape with the victory.
Watching the game, one can see how the size of the 7'0'' Nigerian disrupted the Shockers offense. Throughout his entire career, Fred Van Vleet has been adept in the paint. But against Jekiri, he often drove only to retreat or pass out to the perimeter. Wichita State missed 12 shots from within the paint, including 7 layups.
It is often an oversimplification, but these two games helped prove a commentator's favorite notion: The mid-major guards were not ready for the size of the major-conference team.
These Coaches Know Nothing
As fun as tourney games are for the crunch-time suspense, it is all too often a byproduct of shoddy coaching. But the problems extend far beyond late-game management. It would appear that many of these coaches do not understand the basic rules of basketball. Why else would they not use a foul to give on the final possession of the first half?
Or maybe they graduated from the Andy Reid school of clock management. That would explain Gregg Marshall's curious (read: dumb) decision to not go for the two-for-one towards the end of the first half, despite being down by 13 against Miami.
And when it comes to yanking players due to foul trouble, many coaches are far too ready to sit a star. Even if he has only 3 fouls and there are 11 minutes left in the second half. Mid-major and power-conference coaches alike waste no time in sitting a key contributor once he is presumed to be in 'foul trouble.'
Jim Boeheim provided an model of what to do with a star in so-called 'foul-trouble,' when Malachi Richardson picked up his third foul with 16:30 left in the second half.
Rather than pull his primary ball-handler and hope his backup could preserve a 5 point lead, Boeheim let Richardson play. And with Malachi facilitating the offense, Syracuse quickly built on their advantage. By the under-12 official timeout, it was 47-33. They extended their lead even more. By the under-8 timeout, the score was 55-33 and the game was decided. Ironically, Richardson actually fouled out of the game, recorded his 5th with 1:54 left and Syracuse up 16.
Boeheim's decision should ring in the ears of foul-wary coaches everywhere. The idea is fairly elementary. Suppose a game is tied with 4:57 left when one team's best player gets his 4th foul. If he is pulled right then and there, and sat until the 2:50 mark or so, the team is guaranteeing itself at least 2:07 without its best player. And factor in the uncertainty on when he can check back in, or the need to burn a timeout just to execute that substitution. And he can still pick up his 5th foul during his abbreviated crunch time.
Keep him in the game however, and the team does not need to worry about substituting or burning a timeout. They also keep themselves in the running for the best possible outcome, which is that the star is able to play all of the remaining 4:57. And it isn't particularly difficult to hide a foul-ridden star on defense if he is a perimeter player.
The overly conservative approach to foul trouble has been my biggest gripe with coaches throughout the tourney. But this risk-averse strategy is hardly limited to managing foul-happy stars. Many college coaches seem to make decisions not to pursue the best outcome, but to avoid the worst. This is the wrong way to coach in any sport. Just ask Mike McCarthy.
Slow Teams are Winning, at Least for Now
Of the remaining field, several teams play more like the Jazz than the Warriors. Virginia played at the slowest pace in all of CBB this season, per Ken Pomeroy. Wisconsin was 7th slowest. UNC plays at the fastest pace of any Sweet 16 team, and even they rank only 48th in KenPom's Adjusted Tempo (Poss/40, adjusted for Opponent). The mean rank in AdjT of the Sweet 16 is 211th (out of 351 teams).
Is this surprising? Somewhat. High-seeds that play at slow paces are more susceptible to upsets, and conversely underdogs that slow it down tend to knock off more favorites than their fast-paced counterparts. As an underdog, it almost always helps to turn the game into a short contest, because of the increased variance in outcomes.
That's part of why Stephen F. Austin over West Virginia (219th in AdjT vs. 95th) was a better upset pick than Iona over Iowa State (44th in AdjT vs. 56th). Over a larger sample size, random events that typically help the underdog win are diminished. Turnovers mean less when there are 75 possessions than they do when there are 62. A hot stretch of 3 point shooting matters a lot more in a 66-64 game than it does in a 91-87 game.
Perhaps this will prove to be the round where Virginia's ploddingly slow pace ends up costing them. They are matched up against the second-fastest team left in the bracket, Iowa State. It will be an interesting dynamic to see which team imposes their pace on the other.
Many talking heads will make the point that "Iowa State can only win if it makes Virginia play fast," but I would dissuade Iowa State from needlessly increasing the tempo. A slow pace only adds value to every shot that Monte Morris and Georges Niang can hit.
Michigan State was Never Ready to Play from Behind
It's time to talk about the monumental upset of the tournament. According to FiveThirtyEight, it was the 3rd-biggest upset in tourney history. Norfolk State's narrow win over Missouri in 2012, and Coppin State's embarrassment of South Carolina in 1997 stand as the lone upsets more improbable than the Blue Raiders' win over Sparty.
Michigan State was a favorite by both the computers and fans. On ESPN's Tournament Challenge, 61.8% of brackets had them in the Final 4, and 22.3% of brackets predicted a second championship for Tom Izzo. I was one of them.
Only Kansas was a more popular champion, with an even 25.0% of brackets going chalk, and putting the overall 1 seed as the champ. FiveThirtyEight's ELO model also had the Spartans as the second-most likely team to win it all, with only Kansas ranking ahead.
This iteration of Michigan State was fairly dependent on the three, but not to a concerning extent. Threes accounted for 35.6% of their total FG attempts. They were well above the ~29% average for Division 1 as a whole, but their volume was matched by their efficiency (for those interested, D1's most three-dependent team was Akron, who shot 50.6% of their FGA's from distance).
Michigan State made 43.5% of their threes, an incredible figure considering they only possess two (maybe three) high volume shooters from deep. Adept shooting was the trademark of their top-ranked offense, with everything running through NPOY Candidate Denzel Valentine. The homegrown senior had a USG% of 28.4, and an absurd AST% of 45.8. He was even 2nd on their team in DREB%, finishing behind senior power forward Matt Costello. Valentine was an undeniable do-it-all star this year. Counting stats, advanced metrics, and the eye test all corroborate his immense value.
Relegated to the backseat on this team were the two constants on has come to associate with Michigan State. Capable defense and inspired coaching. Per KenPom, State had the 51st most efficient defense in D1.
Looking up and down the roster, it's hard to pinpoint a defensive stopper. Deyonta Davis (6'10'' with 7'2'' wingspan) is a nascent rim protector, and played just 18.6 MPG. Neither Valentine nor Forbes are elite defenders, though Valentine's 6'10'' wingspan allows him create havoc in passing lanes.
Much of State's defensive competence this season was born out of their ability to limit turnovers on offense and avoid fouling. They relied on composure and execution to play defense, rather than athleticism and pressure. That is not a good thing when it comes time to make up a deficit.
Even as Bryn Forbes and Valentine sank threes and Matt Costello bullied his way into low post scores, Middle Tennessee never relinquished their lead. Michigan State forced just 10 turnovers, scoring 12 points off them. Late in the game, they rarely pressured the MTSU ball handlers. The Spartans did not resemble a senior-led team poised for a deep run. Instead, they looked like the 2014 Duke team who faced Mercer: Thin on defense, and unable to create turnovers and extra possessions.
Lacking a defensive stopper and inexperienced in trap schemes, Michigan State had to play defense without fouling for the duration of the shot clock down the stretch. After drawing within 1 at 77-76 with 3:34 left, Michigan State did not score again until a desperation jumper by Valentine with 31 seconds left.
By then, the outcome had already been decided. MTSU had survived crunch time against the most efficient offense in the country, with surprisingly little resistance. When time started to run low, Michigan State's defense could not squeeze out the few more possessions their offense so badly needed.
Who left in the field of 16 is like the Spartans? Is there another team that excels with the lead, but struggles to make up deficits? In a tournament marked by parity and upsets, a flawed contender will surely reveal itself. But no upset will have the resounding effect of Middle Tennessee's wire-to-wire win over Michigan State.