Tuesday, March 25, 2014

How to Reconcile Tennessee's "Disappointing" Basketball Season with Its Tournament Run

Many UT basketball fans argue that UT had a "disappointing" regular season, and they say, therefore, that Coach Cuonzo Martin (CCM) deserves the blame for this disappointment (even if he deserves respect for his Tourney run).  Furthermore, they argue that Tennessee's "improvement" on the court matches CCM's growth as a coach.  In other words, they believe that Tennessee played poorly for much of the season and they blame CCM for that poor play.

I wonder, however, if he really does deserve so much criticism for the "disappointing" season.  If you judge the season simply on wins and losses, you wouldn't be unreasonable to call the season a disappointment.  It certainly was.

I don't want to argue here so much whether or not the season was a disappointment in terms of wins and losses, but, rather, I'd like to discuss why Tennessee lost 12 times (and some of them to bad teams).  We need to have this discussion because--if you believe the media--CCM found himself on the verge of being fired.  He certainly bore the brunt of the fans' blame for the losses.

But what if Cuonzo Martin coaching acumen wasn't to blame for the losses? What if Cuonzo Martin was just as good of a coach then as he is believed to be now? Even more startling, what if Tennessee actually played quite strong, well-coached basketball all season long? What would in that case have caused all the disappointment?

What if I were to tell you it was simply bad luck?

In fact, such statisticians such as Ken Pomeroy argue just that: Tennessee was one of the unluckiest teams in all of college basketball. 

After all, Kenpom--for those who follow such sabrmetrics fellows--has UT ranked 6th in the nation right now, statistics which weigh the entire season, including all the "disappointments."

How can that be, these CCM critics argue? Everyone knows that UT underperformed this regular season. What is more, he believes that Tennessee's 11 seeded birth into the NCAA is quite low for a team of UT's caliber.

But, again, how in the world can a team that supposedly so underperformed during the season that 30,000 fans signed a petition to get the guy fired be under-seeded at the 11 spot?

Sabermetrics gives us the answer: luck.

In fact, in a recent blog post, Pomeroy argues that "[t]he deceptively limited season of this sport is why a team that's probably one of the country's 25 best can be seeded like it's not in the top 40." In other words, if you flip a coin 1,000,0000 times, you could reasonably expect the coin to split its landings on heads or tails in half.  However, if you flip it only 35 times, you might end up with tails (or heads) 30 times even though it stats tell us it should be closer to 18 or 17 split. 

Therefore, sabermetrics/Kenpom.com thinkers argue that the two TxAM losses (where Jarnell Stokes's missed free throw to win the game and a TxAM player who can't shoot knocked down a three down 2 different times to win) and the  Missouri loss were just unlucky, statistical noise.  Some might even throw in the Wichita game in the mix.  After all, when UT started making a run, Jordan McCrae gets a ticky tack technical for excessive celebration and etc.

 In short, the luck just didn't go UT's way this year.  If UT played an NBA length season, this type of thinking goes, the luck would balance out and UT's record would look more like it should.  In short, UT lost some close ones to some bad teams which resulted in disappointment and calls for the coach's job due to his inability to "get the job done." 

Well, let's take a closer look at the losses:

1) @Xavier (NCAA Team): 4 points.
2) UTEP (neutral): 8 points
3) @WSU (NCAA Team #1 seed): 9 points
4) NC State (NCAA Team) : 10 points.
5) A&M: 1 point
6) @UK (NCAA Team): 8 points
7) @UF (NCAA Team #1 seed): 26 points
8) @VU: 4 points
9) UF (NCAA Team, #1 seed): 9 points
10) @Missou: 5 points
11) A&M: 2points  (OT)
12) UF (NCAA Team #1 seed): 7 points

So 3 losses to the number 1 team in the nation, two of them were "Games."  One loss to a another #1 seed in Wichita St. and another loss to a Kentucky Sweet 16 team.

2 losses on prayers (A&M) and another that UT just missed a final shot on (Missou).

Every other loss, UT  had a chance to win and for one reason or other just didn't. 

Another way to think of it is like this: in the tournament, Wichita lost to UK by missing that last second 3.  If he'd hit it, they tie (or win; I can't remember).  If he knocks it down, is WSU a better team?  Is CGM a better coach?

TxAM is another example: CCM draws up a set play on the final in-bound play after the time out.  It works perfectly: Stokes gets fouled for a shot at walk off free throw win.  Stokes misses.  Bad three point shooter knocks down a three and UT loses.  If Stokes  hits the free throw, CCM gets credit for a great play and strong coaching.  Instead, fans wanted him fired.  Was he really a better coach one way or the other? 

That's the question that sabermetrics forces us to ask ourselves.

In sum, then, the numbers suggest that CCM and UT weren't really far off from expectations, but they were extremely unlucky.  It should be no surprise then that kenpom.com rates UT as one of the top ten unluckiest teams in the nation. In fact, only Alabama, Oklahoma ST, and Iowa (of the big time programs) were unluckier.  No surprise again that Iowa was UT's toughest opponent in the NCAA tourney because both Iowa and UT were vastly under-seeded.

So what's at stake here? Well, our perception of what we see often deceives us. What looks like a disappointing season due to seeming incompetence due to poor coaching was really just a result of a skewed sample size.

In short, UT almost fired a strong coach because of a mis-perception.