You have to get lucky to win it all
We're starting with this first because it has never been more true: Never before in the 64-team era has the NCAA champion won a game in overtime ... in its first game.
Champions have won games in overtime before — 2008 Kansas did, 1997 Arizona did, and 1989 Michigan did1 — but surviving a scare like Connecticut did against Saint Joseph's is rare. The Huskies trailed by nine in the first half and five with five minutes to go, and needed an Amida Brimah three-point play in the final minute just to get to overtime.
Shabazz Napier and UConn's band of charity stripe benefactors took over in the extra period, with Napier scoring seven straight points to build a lead that would never fall under five points and the Huskies going 15-for-16 from the line. And with the benefit of hindsight, it seems obvious that the better team won that game.
But UConn was not the better team in regulation on that night; if it had been, it would have won that game in regulation. And the better team doesn't always win NCAA Tournament games, and the the best team doesn't always win it all, and six games against good to excellent teams over three weeks practically demands that a champion get lucky.
If we assume that a great team, like 2012 Kentucky, has a 90 percent chance of winning every NCAA Tournament game it plays, even the ones against other top-10 teams, that great team has a 53.4 percent chance of winning the NCAA Tournament, just slightly better than the chances of winning a coin flip. And even that Kentucky team, with the fourth-best Pythagorean rating in the KenPom era, didn't go into six games with a 90 percent chance of winning each one, and probably didn't even go into more than three with better than a 70 percent chance of winning. If we give Hypothetical Great State U. a 90 percent chance of winning its first three games, and a 70 percent chance of winning its final three, HGSU has "just" a 25 percent chance of winning a title.
Whether it's winning an overtime game in a 7-10 matchup or winning four straight games decided by five points or fewer in the NCAA Tournament after going 2-8 in games decided by five points or fewer before the NCAA Tournament (hi, Kentucky) or winning a No. 1 vs. No. 16 matchup despite playing poorly for much of the game (which applies to three of the four No. 1 seeds this year) or winning a regional final in overtime thanks to the Pac-12 Player of the Year taking too long to get a shot off, an element of good fortune runs through nearly every Final Four run in the history of the NCAA Tournament.
Considering the crapshoot that is March Madness, there's nothing wrong with that. Getting lucky is so much better than the alternative.
More specifically, you can't crap out
The alternative to getting lucky is something like "Being 2014 Wichita State."
Actually, that might be the opposite of getting lucky.
I watched most of every game Kentucky played in this NCAA Tournament, and I thought no team played better against Kentucky than the Shockers did. The numbers back me up, to an extent: Wichita State's 1.23 PPP against the Wildcats were second to Michigan's 1.27 PPP, but Michigan allowed 1.32 PPP to the 'Cats, more than American allowed to Wisconsin (1.29 PPP) in a 40-point loss; UConn doused white-hot Kentucky in the final, allowing just 0.90 PPP, but it only scored 1.00 PPP.
And both of those teams got to play Kentucky after Willie Cauley-Stein got hurt — though Cauley-Stein didn't do that much that showed up in the box score against the Shockers, his rim protection mattered. And neither of them had to deal with Kentucky posting an Effective Field Goal Percentage of 62 percent — which the Shockers dealt with with their own 65.2 eFG percentage.
And both Michigan and UConn had slightly more talent than Wichita State, and, at least, nulled one of the height advantages Kentucky had on Wichita State inside and outside: Michigan's big guards saw eye-to-eye with Kentucky's big guards, and Brimah and DeAndre Daniels kinda matched up to Kentucky's towering frontcourt.
Wichita State would have had to beat Kentucky and the three other teams Kentucky beat in close, dramatic games just to play UConn for the title. There is no way to oversell how fully screwed Wichita State — the first undefeated team to make the NCAA Tournament field in more than 20 years — truly was; it is as if they stepped to the craps table for the first time and rolled snake eyes.
Other teams crapped out, too. Against UConn, Florida played its first truly bad game in four months — posting its worst PPP of the season and allowing more than 1.00 PPP for the first time since Februrary — in the Final Four, and after building a 16-4 lead. Michigan State and Iowa State had to play UConn at Madison Storrs Garden, and the Cyclones were missing a player they could have really used against the Huskies. Villanova was the perfect No. 2 seed for UConn to draw. Every team that Kentucky played prior to UConn, save hopelessly outmatched Kansas State, crapped out — or, more accurately, got beaten on the river; Michigan would've had to beat two rugged SEC teams perfectly calibrated to take advantage of its defense to get to the Final Four, anyway.
If your team makes the NCAA Tournament, you are virtually guaranteed of one of three outcomes2: Your team will either win the NCAA Tournament, lose to a better team, or lose to an inferior team because of matchup problems or luck. And in an event with 67 losers and one winner, your team is virtually guaranteed of one of the latter two fates.
And it will either cash in or crap out. Bet that.
Matchups matter even more than we think
Kentucky, the NCAA Tournament's best offensive rebounding team, played the Big 12's second-worst defensive rebounding team, a team with no rotation players taller than 6'9", a team whose best defensive rebounder was 6'8", and a team that essentially played one post player at a time on its way to the Final Four.
Kentucky didn't win those games purely because of its offensive rebounding — and Wichita State, which was really good at defensive rebounding despite its height limitations, didn't lose because of the boards — but Kentucky might have had that advantage mitigated as the No. 8 seed in, say, the East Region, where No. 1 Virginia and No. 4 Michigan State, two teams in the top 20 nationally in defensive rebounding percentage, sat on the lines Wichita State and Louisville occupied.
Kentucky's matchups in the frontcourt played to one team's particular strength; that's rare. More common: Matchups that highlight a team's fatal weakness(es), which throw scares into great teams and fell good ones.
For Duke, Mercer's small guards were bound to be trouble; for Ohio State and Syracuse, running into a team smart enough to force the Buckeyes and Orange to win with jump shots was fatal. Kansas eventually lost to a lesser version of itself, but struggled with Eastern Kentucky, a small, quick team with guards that feasted on the Jayhawks' imitation point guard play. UCLA played a Florida team that knew how to play at its pace; Florida ran into a team with a better backcourt, which forced turnovers the Gators hadn't been committing in March.
Virginia succumbed to Michigan State, which could stretch its defense with shooting, and Villanova lost to a team of bombers led by the baddest bombardier breathing. Creighton, vexed by more athletic teams all season, ran into a stupendously athletic one, and had a total meltdown; Baylor, in turn, couldn't hang with a better-coached team that matched much of its athleticism. Arizona couldn't keep Wisconsin in check for 45 minutes.
And all those teams in the Midwest that lost to Kentucky? In their losses, they fell victim to the most obvious matchup nightmare there is in basketball: Being shorter and not as athletic as your opponent.
A truly great team could probably power through a bad matchup or two on the way to a title. But the truly great teams are few and far between, and those great teams might not have to deal with bad matchups simply because they don't have bad matchups, by virtue of being good at everything. In years like 2013-14, with no unequivocally great team — close but no cigar, Florida, and we'll get to why in a moment — it's more likely that a bad matchup will knock an excellent team out than that a team will march through March (and April) despite facing a spiky foe at some point.
Talent typically wins out
You could really have done well for yourself in your office bracket pool this year by picking the more talented team in each game. And that's usually true, I think, if we think of "talent" in terms of caliber and potential, and not production.
Look at the Final Four teams: Kentucky was unquestionably the most talented team in the Midwest; UConn was no worse than second behind Michigan State, or maybe third behind Iowa State, in the East; Wisconsin was third at worst, behind Arizona and maybe Baylor, out West; Florida was, arguably, fourth in talent in the South, but it only had to play UCLA, thanks to Kansas and Syracuse making early exits, and Kansas and Syracuse might well have earned higher seeds (or, uh, made the Sweet Sixteen) if not for injuries to their more talented players.
And while great depth of talent isn't a prerequisite for NCAA Tournament success, no team since 1986-87 Indiana has won the national championship without a first-round pick on the roster. At a minimum, teams have needed a great collegiate player, one who will absolutely be an NBA player at least briefly, at a minimum, to win a title. Whether that player is Shabazz Napier or Gorgui Dieng or Anthony Davis or Kemba Walker or Nolan Smith (yes, he was 2010 Duke's highest pick) or Mario Chalmers or Ty Lawson or Joakim Noah doesn't matter all that much; you just need one.
Florida did not have that player in its rotation this year, not unless you think Patric Young still has a shot at the first round, Kasey Hill, six feet tall in shoes, is a surefire first-rounder, or you mistakenly think Chris Walker, who played 18 minutes in the NCAA Tournament, was in Florida's rotation. And that was ultimately part of Florida's demise: Scottie Wilbekin, who will have to scrap like hell to make an NBA roster, was the guy playing the role Rose or Lawson would have for a similar team in the past, and he was playing against one of those championship teams' studs, Napier, in Florida's Final Four loss.
Could Florida have won it all, despite that limitation? I think so — and I still think Florida was the nation's best team this season, even without my orange and blue lenses on. But our last three lessons help explain why UConn, not Florida, won their Final Four game — and, more importantly, why UConn won it all.
Free throws may be the new inefficiency
Foul line failure was not the primary reason Florida lost to UConn. But the Gators didn't help themselves at the stripe: They missed two shots early that could have made that 16-4 lead an 18-4 lead, and Young missed two front ends of one-and-ones late in the first half that could have potentially swung four points in the game. And while UConn was still uncannily good from the line in that game, Brimah missed two free throws to give Florida the ball with a chance to tie in the second half. (Florida responded by promptly committed a turnover.)
That said: I've never seen a team control games from the free throw line like UConn did in this NCAA Tournament, and I'm convinced that UConn has, wittingly or unwittingly, stumbled on a new inefficiency in basketball.
The most interesting thing I read all March was this Dan Wetzel piece on UConn assistant Glen Miller "discovering" an obscure Steve Nash "drill" on YouTube in 2011 and trying it out with the Huskies. Except it's not video of a drill, per se — it's Nash trying to make as many free throws as he can in a minute as part of a challenge for dead-and-buried social network ibeatyou, best known for being a) backed by Jessica Alba's husband and Baron Davis, b) producing a strangely transfixing video of Alba staring at a camera for about 90 seconds.
Nash — in full bangs-and-Canadian flow glory — appears to be kind of leisurely shooting the free throws in his video, and makes all 21 of of the free throws he shoots. (What else would you expect from the NBA's career leader in free throw percentage?) But Miller and Kevin Ollie, Wetzel reports, turned the drill into something UConn does as a transitional exercise in the middle of practice — with heart rates elevated, mimicking the conditions of, oh, the last five minutes of an NCAA Tournament game.
We laypeople, who watch sports from couched and chairs and barstools, often fail to realize that the physical demands of high-level sports are so great that they drown out conscious thought3. A free throw in the final minutes of a close game is a slow-twitch act done without much physical capacity to think, regardless of crowd noise and so forth. So muscle memory, and specifically ingraining the stroke necessary to hit free throws without even thinking about it, is probably the as close as a key to free throw shooting as there is.
UConn has been largely4 great at the line since 2011, ranking 12th in free throw percentage in 2011, 33rd in 2013, and fourth this season, and I have to think that has something to do with Ollie training that into his Huskies.
It paid major dividends in this NCAA Tournament: Connecticut made an insane 87.8 percent of its free throws in the six games, setting the NCAA Tournament record, and was even better in the final five minutes of games and its overtime against Saint Joseph's, making 92.5 percent of its free throws in those final stretches.
UConn's last missed free throw with fewer than five minutes to play in the 2014 NCAA Tournament was the one Napier missed with 21 seconds to play against Iowa State, and the Huskies up five. UConn didn't miss a single free throw with fewer than five minutes to play and a score closer than that in the entire Tournament ... and Napier just drilled the next one, putting the Huskies up six, then watched as the Cyclones' DeAndre Kane missed both shots at the other end.
By tourney's end, UConn's foul shot mastery was so obvious that John Calipari, a devotee of fouling on defense to the point that he's one of the first names most hoops junkies think of when coming up with coaches who foul up three, said he didn't instruct his players to foul UConn because "They never miss." And though I'm really not inclined to bash Calipari with the Narrative Club over free throw shooting, especially given that his teams understand how to draw free throws, UConn went 10-for-10 in the final, while Kentucky's 13-for-24 performance left points that would have covered the six-point margin of victory at the line.
I imagine UConn will keep running the Nash Drill, but I'm not sure whether UConn will enjoy the advantages of exploiting this inefficiency in coming seasons. Given how referees went from calling everything before Christmas to their old, lax, and inconsistent form after New Year's this season, there might not be that many free throws to call; given how Florida and Kentucky adjusted to UConn in the Tournament, giving their guards wider berths and allowing UConn to draw just 23 free throws in the Final Four after shooting 22 against both Iowa State and Michigan State, I wouldn't be surprised to see other, smarter teams sag off the Huskies and make them win with jumpers rather than charity stripe donations.
The fruits of the inefficiency, though, included a new banner to hang. And banners hang forever.
Defense is more reliable — and thus important — than offense
Florida's renaissance under Billy Donovan in the last few years has been keyed by an emphasis on and understanding of defense that extends to a granular level. If you believe this 2013 ESPN The Magazine story, one I tried to unpack at Alligator Army when it dropped, the lightbulb moment for Donovan was realizing that holding teams under 0.90 points per possession was the recipe for Florida having success in the NCAA Tournament. And he's right: In the KenPom era, which dates to 2002-03, Florida is 10-0 when holding teams under 0.90 PPP in the NCAA Tournament.
But Donovan's always had really good offensive teams, too, which makes 0.90 PPP allowed more of a threshold for confidence than one for success. In fact, in the KenPom era, Florida is a staggering 20-0 when it holds its opponent under 1.00 PPP, and the 2005-06 national champions held all six teams they played in the 2006 NCAA Tournament under that threshold.
That's very, very difficult to do. The feat has only been matched once by 2007-08 Kansas, the best-rated team in KenPom history — and Kansas needed Mario Chalmers to make his miracle three to win that title, as you may recall. It's much easier to slip up: Florida did just that against UConn, allowing 1.10 PPP, and fell to 2-5 in NCAA Tournament games when allowing at least 1.00 or more PPP since 2007.
Meet the Bag Man
Meet the Bag Man
But a team can slip up earlier on and survive, usually. UConn allowed 1.12 points to Saint Joseph's, and 1.15 points to Iowa State, but still won both games, by eight and five points. Wisconsin gave up 1.18 PPP to Oregon, and still cruised. Michigan yielded 1.14 PPP to Texas, then 1.18 to Tennessee, and advanced.
It's when the slips become a trend that should really trouble us, and when things are really unsustainable that we should sound the alarms. Kentucky allowed three of its five worst PPP performances of the season against Michigan (worst at 1.26), Wichita State (second-worst at 1.23), and Wisconsin (fifth-worst at 1.19), and won all three of those games by a combined four points, because a) Kentucky was fantastic on offense in this Tournament, topping 1.16 PPP in four straight games, and 1.20 in three of them and b) Aaron Harrison had the single greatest NCAA Tournament by a shooter in the history of college basketball.
In the other two games among Kentucky's five worst defensive performances this season, Florida smacked the 'Cats by 10 points at Rupp and by 19 points in Gainesville.
Kentucky was due to return to terra firma offensively, and it did against UConn, mustering just 0.90 PPP while conceding 1.00 PPP — not abysmal, but too much. And that also meant UConn allowed just 0.90 PPP, hitting the band between 0.90 and 0.92 for the fourth time in the Tournament.
UConn combined that stinginess with alternately brilliant and decent offense: The Huskies topped 1.20 PPP against the Hawks and Cyclones, found the 1.10 PPP plateau against miserly Florida, and put up the sixth-best day (1.09 PPP) against Villanova, but were just okay against Michigan State (1.00 PPP) and Kentucky (1.00 PPP).
Championship teams can get away with just okay on offense, as long as their defense is better.
Go with guards
And UConn had actually whipped itself into a fine defensive team late in the year, but three games against Louisville helped obscure that fact. Louisville was such a bad matchup for UConn, in fact, that it probably damaged both teams: The 3-0 sweep of the Huskies did little to make up for Louisville's dismal non-conference strength of schedule, but a 2-1 split decision might have boosted the American's RPI enough to get Louisville a prayer of making the No. 3 seed line; the 0-3 brooming consigned UConn to the No. 7 line despite the nation's most recent win over the overall No. 1 seed, and set the Huskies' ceiling at a surprising run in the East in most minds, including mine.
Two of our previous lessons help explain why that happened. First, Louisville was a terror of a matchup for the Huskies, because the Cardinals had guards (Smith and Chris Jones) that could match the speed of Napier and Ryan Boatright, and they had Montrezl Harrell. Second, Louisville played better defense than UConn all season — in no small measure, this was because of its quicker guards — and held the Huskies under 1.00 PPP in all three meetings.
But Louisville's guards didn't get going in this NCAA Tournament. And that's a shame — our final lesson, which is simply that the the better backcourt will probably win any given game, is one that tells me Louisville should've been bound for more greatness than it ended up earning.
I spent some of my time in Dallas after Florida's loss to UConn thinking about UConn's national championships, which have basically bookended my interest in college basketball to date. And the best commonality I came up with was that UConn always has a great guard or two, whether it was Khalid El-Amin and Richard Hamilton in 1999, Ben Gordon in 2004, Kemba, Jeremy Lamb, and Napier in 2011, or Bazz and Boatright this year.
The thing that bugged me about that thread, though, was that I knew Ben Gordon wasn't a point guard. Once I got home, I looked up 2004 UConn's roster, and found out that I was trying to think of Taliek Brown. I also found out that Rashad Anderson, that team's third guard, scored 17.3 points per game in the NCAA Tournament, while setting school tourney records for threes in a game and tournament.
I was right: UConn always has a guard or two. Or three. And having multiple guards matters.
Having two players who can handle the ball on the floor at one time opens up so many things for an offense. You can feed the post on both sides of the lane, or have those guards take turns running the offense and playing off guard, or let one of those guards take possessions off on offense (as Florida did routinely with Wilbekin when Hill was in games this year) or defense (as, say, Syracuse would do with, say, Tyler Ennis, if it were ever not playing a zone). Playing "pure" point guard is a dying art, especially as the positional flexibility that LeBron James helped teach a generation of growing players becomes de rigeur in youth basketball, and every player thinks of himself as a combo guard, but more players knowing how to handle the ball and direct an offense is a good thing, and more talented players having multifaceted perimeter games makes those good players even better.
And having two "point guards" on defense might be even more valuable than having multiple ballhandlers on offense, because cat-quick players who can stay in front of the other guards, force turnovers, and limit perimeter shots with ball pressure are the most valuable pieces of a successful college basketball team, especially when deployed in tandem.
For Florida, Wilbekin and Hill had been those guys this season ... and then Napier and Boatright devoured them for two nightmarish hours. Kentucky had gotten to the NCAA Tournament final thanks in part to Andrew Harrison playing point guard extremely well for four straight games ... and the same thing happened to him.
The Harrisons being great also helped knock out other great guards, and greats to be. Wichita State was as good as it was this year because of Fred VanVleet, but Harrison's size nulled him; Louisville had a terrific tandem in Smith and Jones, but Smith was just off for the Cards' three games. Michigan, a team that usually played four guards with some PG skills, really only lost to Kentucky because Aaron Harrison was more brilliant at the right moment. And Wisconsin's guards, from Traevon Jackson to Ben Brust to Josh Gasser to Bronson Koenig, gave Kentucky plenty of trouble on a night when Frank Kaminsky couldn't give the Badgers much of anything.
When I cast about this year's bracket, I find good guards winning games. Arizona destroyed Gonzaga behind Nick Johnson and T.J. McConnell combining for nine steals. Iowa State only made the Sweet Sixteen because DeAndre Kane had a monster game, thoroughly outplaying North Carolina's Marcus Paige. Dayton's fleet of guards got the Flyers to the Elite Eight; Tennessee's sudden discovery of guard play nearly got the Vols that far, and got them just as many wins.
And those matchup problems mentioned above? They have everything to do with guards.
Quicker guards like Boatright are stressors, and the bigger (UCLA's Kyle Anderson) or stranger (Virginia's London Perrantes) ones are the things coaches have to scheme around. No one worries about the big man underneath anymore; you can just double him, or deny him, or cut off his touches by trammeling the guard who feeds him. Guards shoot over the zone, and control the tempo; not having a really good one leaves you at your opponent's mercy, and you can ask any Kansas fan how that feels.
UConn has four national titles in 16 years, and as many as any college basketball program has since John Wooden left UCLA. UConn always has guards.
I feel as though these facts may be related.
When in doubt, go with the guards.
All three — Michigan in the Kingdome, Arizona in the RCA Dome, and Kansas in the Alamodome — in dome stadiums that were not basketball-only facilities, if you care about weird coincidences like I do.
The fourth outcome is a no-show by the better team, one that happens independently of luck or matchup issues, and it's really, really rare. The one better team that arguably no-showed this year was New Mexico ... but the Lobos weren't that much better than Stanford in the first place, and still closed to within two in the final minute.
For an example of how this problem gets dealt with in a similar fashion in mixed martial arts, read Tim Marchman's brilliant piece on trainer/Svengali Greg Jackson for Deadspin.
UConn was also 270th nationally in free throw shooting in 2012. But Huskies other than Andre Drummond, who shot 30 percent from the line over 88 attempts two years before we thought Aaron Gordon was murdering basketball on every free throw, hit at a 72.2 percent clip, which would have been about 70th nationally.