Calm down. Breathe. Slowly.
It’s going to be okay.
In a week where nineteen teams ranked in the top 25 lost, almost all at the hands of unranked teams, it’s easy to freak out. We are after all in the midst of one of the craziest college basketball seasons that I can remember.
It’s easy to find the biggest cliff or skyscraper or soapbox and scream “THE WORLD IS FLAT!!!” like my boy Seth Davis has been doing.
Georgetown won AT Xavier. The world is flat.
— Seth Davis (@SethDavisHoops) January 20, 2016
Oklahoma State up 18 on Kansas final minutes. The world is flat.
— Seth Davis (@SethDavisHoops) January 20, 2016
Maybe Clemson should stay in Greenville. Tigers have beaten 2 straight ranked teams in Louisville and Duke. The world is flat.
— Seth Davis (@SethDavisHoops) January 14, 2016
How the heck am I supposed to complete the perfect bracket if literally any team could get upset at any moment?!?
And while you’re right in that I have a vested interest in you going and following Seth Davis on Twitter…seriously go follow him…he shares the sentiment of almost every college basketball fan this season. Essentially “nobody is good and nobody is bad so literally anything can happen!”
Calm down. Breathe. Slowly.
Turn to lesson two in your imaginary notebooks and get ready for another educational edition of the March Madness Survival Guide! I can sense your panic is so strong, this time I had to pull out some Latin.
“Post Hoc, Ergo Propter Hoc”
Since event Y followed event X, event Y must have been caused by event X.
What the what? It’s Latin for “After this, therefore because of this” and it’s logical fallacy that gets used all the time. Essentially all superstition is built on this fallacy. One time you’re watching a football game and your favorite team scores a touchdown while you were reaching into the refrigerator for a drink, and next thing you know you’re wearing your pants inside out with no shirt and a yamaka on while holding one arm in a refrigerator for the entirety of every game.
Essentially something was only good or true after a certain event happened, therefore that event must’ve caused the good or true thing to happen.
This gets used ALL THE TIME in sports. We come up with a theory of why things happened only after they’ve happened. Stop me if you’ve heard this one before: “The Warriors never had to play against a healthy team in their playoff run, so they really only won the championship because they got lucky playing injured teams.”
Well, Survivalists, here we are on January 20th, and event Y has already happened. Every team we rank in the top 25, every team that we rank in the top 10, every team that we rank number one, goes down shortly thereafter. But you’re smarter than the average fan, and that’s why you’re going to win Warren Buffet’s Billion Dollar Booty (I mean that in the pirate way not in the way you were thinking about it ya nasties).
It’s pretty simple for everyone else to grab up a lot of similar events and bundle them together into a “trend” and then say “these upsets are happening because it’s one of those years,” but then you’d be falling right into the hands of the post hoc fallacy.
So why then? How are these inferior teams beating these superior teams? In order to explain we’re going to have to decompose how teams win and lose basketball games.
Stay with me here, because I’m about to get kinda technical for a minute.
Imagine a test-tube scenario of a basketball game between Team A and Team B. They are of the exact same skill level as a team.
Think of all of the ways a basketball game can go. If both teams play at the exactly average level of their total potential, a basketball game between them will be essentially a tie, for reference sake we’ll say they both played at a 50 on a 1-100 scale of their abilities. But if Team A plays the best they can possibly play, at a 100 out of 100 on our made up scale of how well they played, and Team B plays at a 50, then Team A is going to win that game. Get it?
Now imagine the real world. Teams are not of equal ability and skill level, so therefore their scales are not all from 1-100. For example Kansas, who I still consider the best team in college basketball, might have a scale of 50-100 and a team like Oklahoma State who doesn’t have a great team this year might have a scale of 25-75. But on Tuesday night Oklahoma State beat Kansas 86-67. If you watched the game you know the Cowboys played as best they could (at a “75” in this example) and Kansas looked disinterested and lazy (a “50”).
But one step farther than that! Not every team has a range of 50 like I just used in my example. Every team has a different floor, and a different ceiling. Allow me to introduce the concept of variance.
Peter Keating and Jordan Brenner have been studying upsets in the NCAA Tournament for 11 years. They have been studying, collecting data, and analyzing trends. Every year they make correct predictions on their “Giant Killers” blog on ESPN, including VCU in 2011, Lehigh in 2012, last year Georgia State, and many many more.
Their basic conclusion is that teams can increase their chances of pulling upsets, of being “Giant Killers” by increasing their variance. Think of it in terms of the scale I made up earlier.
If you’re a team that has a scale of 45-55, you might win a lot of games in your mid-major conference, where teams that aren’t as naturally talented may be playing a lot of games in the thirties and forties, and your consistency is rewarded with a conference championship and a bid to the NCAA Tournament. Now you’re facing Kansas. The chances of you pulling an upset are incredibly small because you’d have to be playing almost flawlessly (55) and they’d have to be playing as bad as they possibly could (50).
But if you start to jack up the variance, you become a team that is extremely dangerous in a single elimination tournament (like the NCAA Tournament is). Suddenly you’re Shaka Smart and your style of play is incredibly high risk-high reward and your (again, fictional) scale is 20-90. That is why the 2011 VCU team lost regular season games to Northeastern by 11 and to Old Dominion by 11 at home, and then went on to beat Georgetown, Purdue, and Kansas in March.
You’re getting closer to understanding I can tell. Let me explain one level deeper and then we’ll bring this baby full circle I promise.
How do we determine the variance of a team?
If it was easy to assign a 1-100 range for every team in college basketball, so that we knew just how well they could play when they brought their A game and how horribly they could play, predicting games would be much easier. While we can’t do that, there are different factors to look for that can help us get a grasp on a team’s capabilities.
Keating and Brenner, through years and years of data analysis, have determined that a less talented team has their best chance of pulling an upset if they can do one or both of these things:
1. Generate more possessions
2. Maximize the value of your possessions
Duh, right? But in accepting those two ideas there are several statistics that begin to appear. How do you generate more possessions? By playing quickly, grabbing offensive rebounds, forcing turnovers, and not turning the ball over yourself. How do you maximize the value of your possessions? By shooting lots of threes. Sure it’s risky, but that’s kinda the point.
Shaka Smart’s VCU teams were the perfect test case for this. They could beat anybody (2011 Final Four run, Sweet Sixteens in 2012 and 2013) or lose to anybody (upset in 2014 and 2015 in overtime to lower seeded teams). Look at his team this year at Texas: they boast wins over North Carolina, Iowa State, and even this week against West Virginia. They also have losses to Texas Tech, TCU, and Washington.
The flipside of this coin also reigns true. If inferior mid-majors increase their chances of pulling upsets by jacking up their variance, then powerhouse teams also increase their chance of being upset by having a high variance.
So you want an explanation for why this year the world seems flat? You’ve found it: variance. Rule changes, including lowering the shot-clock from 35 to 30 seconds, have increased the overall pace of play for every team this season, thereby increasing possessions for every team.
And any fool with eyes can see that every single team is firing up more three pointers per game this season than ever before. This year, there are 40 teams that fire up 25 or more three pointers per game as a team. In 2013 there were seven. Even last season, there were only 14 teams that launched up 25 deep ones per contest. This year there are 60 teams that make 8.5 threes per game or greater, versus 20 a year ago.
Indulge me please on one more variance factor, this one being total speculation. Is it just me or have all of the big men in college basketball gotten way skinnier? Gone are the days of Sean May at UNC and Dexter Pittman at Texas. Heck, even fluffy Kennedy Meeks is looking ultra-trim now. Could this possibly be leading to more offensive rebounds? I’ve played basketball long enough to know that it’s much harder to offensive rebound when you’re being boxed out by a huge fat kid than by a tall skinny kid. Just food for thought (and for the fat kids out there).
Either way, it’s easy to see that every team has much higher variance this season. Do you really think the world is flat?? Do you think that Nebraska is really just as good as Michigan State? Or is it more believable that in 2016 both teams have a high enough variance that if Nebraska has a guy that gets hot (Shavon Shields 28 points) and they only commit 6 turnovers that they could beat the Spartans if they have an off night?
Here’s a counterexample for those who say the world is flat. Villanova was 159th in possessions per game last year, but this year ranks 320th. They are less reliant on three pointers than past years, and rank 19th in the country in fewest turnovers per game.Oh, and just to back up my point, they have a center who’s 6-11 and 245 pounds, and I don’t think anyone has called him skinny in a very long time.
The result? The Wildcats are 17-2, and most importantly haven’t been upset yet. Their only losses were to Oklahoma and Virginia, both ranked in the top 10 when they played them.
So don’t fall into the fallacy of “Post Hoc, Ergo Propter Hoc.” Instead, make note of the teams that have a high ceiling, which have a low floor, and which are consistent. I promise you, it will pay off when you are filling out your bracket come March.
That’s all for now Survivalists, stay safe out there, and enjoy one of the craziest years of college basketball in recent memory!!!
By Matt Craig
Please leave your thoughts and comments below!