The NCAA Tournament in Stats

So I'll just get come right out and say it: I got my ass handed to me this year.

To recap a few of my Roundball Roundtable picks:

Top team to fall: Xavier to Portland State, first round.

Xavier made the Elite Eight.

(Kudos to Jonathan Lowe and Jean Neuberger for taking Cleveland State over Wake Forest and Ross Lancaster for taking Western Kentucky over Illinois.)

This year's Davidson? Arizona State.

Waxed by Syracuse in the second round.

(Extra credit for me for saying the world was about to find out how great ASU sophomore James Harden is. All he did was completely mail it in with 19 points combined on 3-of-18 shooting.)

Final Four: Louisville, Memphis, Pittsburgh, Arizona State.

What do they call it in baseball when you strike out four times in a game? Oh yes. The Golden Sombrero.

Vamanos, amigos!

So now that we've established my credentials at predicting the future, let's leave the Final Four prognosticating to somebody else. Rather, I'm going to put my true talents to test in figuring out the solution to a very important question:

What the hell just happened here?

I've broken down all the stats I could find, looking for not only some sort of explanation of what has happened so far in the tournament, but also looking for a leg up on next year's competition.

The major headline: it's all about the offense.

Thus far, there have been 60 games, not including the play-in. Teams that came in averaging more points per game are 42-18. Teams that came in with a better scoring defense went just 28-32 and teams with the better field goal percentage defense went just 29-31.

But offense isn't just scoring. It's also how you score. Are you overly reliant on dribbling to create a shot, or are you more of a passing team? And if you are a passing team, how good are you?

Teams with a higher assist total are 39-20 and teams with a higher assist-to-turnover went 41-17. That's not to say that the team the turns over the ball the least wins, because those teams are only 27-29. It's more of how many turnovers you have in relation to how many assists you have.

And of course there's what you do if you miss a shot. Teams with the better rebounding margin are 43-17, the single most dominant statistic of them all.

Those stats are all well and fine, but they're not winning your bracket pool. Even if you took the rebounding margin all the way through, you'd still have 17 "X" marks on your sheet. Unless you're in a pool with a bunch of morons, that's not good enough.

So I dug deeper.

Using the rebounding margin totals and adding in scoring margin as an additional variable, you get a 31-8 success rate. Add in more assists per game and it's 22-4. Now there's a winning percentage you can hang your hat on.

Slice it a different way and add assists per game to points per game and you're at 31-10. Add in field goal percentage and you're at a very nice 23-6.

One thing I found out was that different statistics changed in importance as the tournament progressed. For instance, teams with a higher free throw percentage went 21-17 during the first two rounds. But during the Elite Eight games on Saturday and Sunday, the team with the higher free throw percentage went a perfect 4-0.

Conversely, PPG, as I mentioned, was a great harbinger of success in the first round at 23-9. But those teams went just 1-3 in the Elite Eight.

Really, the only stat that carried its success through every level so far is rebounding margin: 21-11 in the first round, 12-4 in the second, 7-1 in the Sweet 16, and 3-1 in the Elite Eight. (If that carries forward, it's good news for Michigan State and North Carolina.)

In order to help make sense of it all, here are the top indicators by round (broken into single category and combo stats).

First Round Single Category

Points per game: 23-9
Assist-to-turnover ratio: 22-9
Rebounding margin: 21-11
Scoring margin: 21-11

First Round Combos (Two)

Points per game + Points per game allowed: 9-2
Scoring margin + Rebounding margin: 17-6
Points per game + Assists per game: 17-6

First Round Combos (Three)

Assists per game + Points per game + Field goal percentage: 11-3
Assists per game + Scoring margin + Rebounding margin: 11-3
Points per game + Field goal percentage + Three-point field goal percentage: 11-4

Second Round Single Category

RPI: 14-2 (I know it's not a stat, but I tracked it with the intent of mocking the performance of higher-rated RPI with a look toward discrediting the whole system next season. Whoops.)

Field goal differential (Difference between own field goal percentage and field goal percentage allowed): 14-2

Points per game: 13-3
Scoring margin: 13-3

Second Round Combo (Two)

Scoring margin + Rebounding margin: 9-0
Assists per game + Points per game: 9-0
Assists per game + Assist-to-turnover ratio: 11-3

Second Round Combo (Three)

Assists per game + Points per game + Field goal percentage: 8-0
Points per game + Field goal percentage + Three-point field goal percentage: 8-0
Assists per game + Scoring margin + Rebounding margin: 7-0

Sweet 16 Single Category

Rebounding margin: 7-1
RPI: 6-2
Assist-to-turnover ratio: 6-2
Assists per game: 6-2

Sweet 16 Combo (Two)

Assists per game + Assist-to-turnover ratio: 5-1
Scoring margin + Rebounding margin: 4-1
Assists per game + Field goal percentage: 4-1

Sweet 16 Combo (Three)

Assists per game + Scoring margin + Rebounding margin: 3-0
Assists per game + Points per game + Field goal percentage: 3-1
Assists per game + Field goal percentage + Three-point field goal percentage: 2-1

Elite Eight

Free throw percentage: 4-0
Rebounding margin: 3-1

Those were the only two stats worth mentioning. None of the single stats or combos went better than even, and most (15 of the remaining 28) went worse than .500, including points per game allowed + field goal percentage allowed + three-point field goal percentage allowed at 0-3 and three-point field goal percentage allowed as a stand-alone at 0-4.

Now for one last question: will this or won't this help me fill out my bracket next year?

Yes.

Now whether or not it will be any better than this year's is another matter entirely.

Leave a Comment

Featured Site