The 2023 NCAA Tournament has deleted turnovers from its hard drive
plus, three-point shooting makes an abbreviated comeback
This won’t be the longest post in the world, but I know several have expressed interest in uncovering major/minor trends from the world’s most beloved small sample. There’s a few fun narratives from this year’s Tournament worth extracting, though. Some of them reflect the 2022-23 season at large; others are fascinating curiosities that we’ll look forward to seeing if they matter or not in 2023-24.
First and foremost, though, we’ve got to discuss some very huge trends that tell the story of the Tournament itself.
Say goodbye to turnovers
With just 15.8% of possessions ending in turnovers this March, this is the least turnover-heavy NCAA Tournament in history. Last year’s finished up at 16.4%, which was low but not the lowest (2014, 16%). There’s a lot we can figure from this, but some of the key observations once diving further into the data are as follows.
None of this year’s Final Four teams were elite at turnover prevention in the regular season. Miami finished 60th in offensive TO%, which is indeed very good, but 19 teams that made the Tournament were superior to them in this stat. Miami wasn’t even a top-quarter team in offensive TO%. Meanwhile, UConn actually possessed a below average TO%.
However, they were all better than the national average at forcing turnovers. The lowest was 170th-placed Miami, who still forced turnovers on 18.1% of opponent possessions.
Uniquely, none of this year’s Final Four teams can force a turnover. The highest TO% forced by any of them is Miami, who sits at 16.4%. That’s 26th-best among this year’s 68-team field.
Offensive rebounds are less important than ever…unless you want to go far in March
At just a 28.7% OREB%, this is the second-least friendly March on record for the offensive boards in NCAA Tournament history…after last March, which closed at 28.8%. Teams rely less on the offensive boards than ever before and more on shooting efficiency, which tracks with basketball’s general retreat to the perimeter versus the paint.
Still, offensive rebounding has played a huge role in this year’s Tournament. All four teams left have rebounded 30% or more of their missed shots, and it would be very fair to say that Florida Atlantic likely wouldn’t be in the Final Four had they not posted a 39.7% OREB%. They’ve had the worst turnover problem of any team left in the field and haven’t shot well. Without crashing the boards, you might be talking about a different team coming out of this year’s South Region. Shot volume still wins games!
The regular season’s Four Factor importance is a little different from March, but overall provides the same road to success
From November to February, the distillation of efficiency to offensive success via the Four Factors is roughly as follows:
eFG%: 50% of the pie
TO%: 25%
OREB%: 20%
FT Rate: 5%
If you win eFG% and TO% together, it’s produced a win rate of roughly 81% over the last three years. In the 2023 NCAA Tournament, teams that win both eFG% and TO% are 26-3. Shooting in general still matters; teams that win eFG% in general are 65-16.
But the other Four Factors can cover up for a bad night. Ask one of our Final Four participants, in fact. Miami got outshot heavily by Drake, but won the turnover battle, beat Drake in the OREB% contest, and went to the free throw line much more often. Without winning those other Four Factors, they’d probably be sitting at home right now and watching Houston or Texas steal their spotlight.
There’s been seven games where a team has lost eFG% but won TO%/OREB%/FTR; the three-factor winner is 5-2. On nights where you’re not hitting shots, simply not turning the ball over and out-efforting your way to a great night on the boards can cover up for a lot.
The 2023 NCAA Tournament: sloooooow
This is perhaps the biggest surprise of all. After entering last year’s Final Four at a hair under 69 possessions per game, the 2023 NCAA Tournament enters this year’s at a shocking 66.4. Possessions are lasting 0.6 seconds longer on average, which could help explain a decent bit of the ugly missed shots we’ve seen.
But…well, I’m fascinated to see where this leads. I personally don’t have a great explanation, other than the games have generally been close, teams play slower in close games to bleed the clock, and officials have called even fewer fouls than last year. Still, if it were to hold, this would be a two-possession dropoff from the regular season and would be fairly easily the slowest NCAA Tournament since the shot clock dropped from 35 to 30 seconds before 2015-16.
While this tournament hasn’t been beautiful, it’s been more efficient by about 2.5 points per 100 possessions than last year’s disaster show, and shooting has quietly rebounded a bit after a very slow start. Is this a sign of things to come, or is this a blip on the radar?
Here’s some other, more minor observations.
After an awful opening weekend, shooting has more or less gotten to where everyone hoped it would be. In the Rounds of 64 and 32, teams shot 30.8% from deep, a shocking dropoff from the regular season and something we already explored on this site. In the Sweet Sixteen and Elite Eight, at completely different venues, teams combined to shoot 34.9% from deep, which was above the regular season average of 34%. I think we can stop complaining about the ball now, hopefully.
Shooting wins again, except for San Diego State. Three of the Final Four participants have better eFG% numbers than the 2002-2022 average of 53.1% on the season. The fourth is San Diego State, who entered this Tournament at 50.1% and has fallen to 49.6% across their four games. They’re the first team with an eFG% of 50% or worse to make the Final Four since 2017 South Carolina and just the third in the last decade (2014 Kentucky). Lesson learned: when your opponents shoot 17% from three, you are well and truly capable of anything.
A few pre-Round of 64 principles held up. Such as:
High seeds with negative turnover margins don’t go far in March. Of 89 1-4 seeds with a negative turnover margin, just 10 have made the Final Four. Only two of those have happened since 2009. Against an expected KenPom win average of 1.92, they’ve won 1.69. This would’ve ruled out five teams before the Tournament began from going deep: Purdue, Alabama, Arizona, Xavier, and Indiana. That group of five combined to win five games in March, with none of them even making the Elite Eight.
Shot Volume still matters. Remember this post? While a lot of the specific examples did not hold up, the overall theme sure did. There’s now been 29 1-4 seeds with an Offensive Shot Volume of 112 or lower. Just one (2019 Texas Tech) has made the Final Four. Six came this year, and all six failed, with only Kansas State and Texas even making the second weekend.
Strength of schedule - particularly of the non-conference variety - remains an overrated talking point. The non-conference KenPom SOS of this year’s Final Four teams: 210th, 18th, 251st, 246th. Only San Diego State played a truly difficult slate. Plus, the top five conferences combined to produce one Final Four team. Let’s maybe back off of various schedule-based narratives.
And one, sadly, did not.
The strongest predictor of March success was, surprisingly, conference success. My continued railing against college basketball media taking February far too seriously did not manage to work out this year. As such:
Top 15 teams in November/December: average seed 3.8, average wins 1.4
Top 15 teams in January/February: average seed 3.7, average wins 1.6
Alas. Some of this was influenced by injury, and Memphis was one missed timeout call away from a potential deep March run of their own, which would’ve influenced the numbers by a lot. This still would’ve told you to fade teams like Texas A&M, Duke, and TCU, but you would’ve missed out on Creighton.