March Madness Casuals Cheat Sheet: Circle These Teams For Your Bracket And Find Out Just How Much Better Auburn Is Than Everyone Else
Andy Lyons. Getty Images.March Madness is sneaking upon us NCAA men's basketball casuals. It's time for a crash course on what the hell is going on.
Let me be clear about one thing. I have no idea what the hell is going on. Sure, I've caught a few games here and there and watched Calipari walk onto the court in Kentucky like he was the main card of a boxing match, and caught some games here and there, but that's about it. And that's OK. We're going to let the data do the work for us. We can't all be Jon Rothstein or John Fanta. Or any other John variant. Nothing but respect, but absolute LUNATICS to be up on 364 teams year in and year out.
No matter. For those of us that have some semblance of a life, we're going to cheat. I've been hard at work these past few weeks assembling an insane amount of college basketball data for us casuals and I'm ready to take it for a drive. The end goal here is to find out who might be sneaky good / sneaky bad for our brackets. Without having to watch 43,429 games every night.
To accomplish this feat, I've created a poor man's advanced model that aims to outperform the fancy ones with its basic simplicity. My special secret model uses four key components in developing a team's true rank. Team win percentage. Team point differential. Opponent's win percentage. And opponent's average NET ranking. The "Special Score" is simply the combined z-score figures of all four components (opponent's average NET ranking reverse-coded). So, I guess the secret is out.
Why these four factors? Think about how you might ask yourself how good a team really is. First, you might ask how many games they won/lost. Then you might wonder how bad they beat their opponents. Next, you want to know what their opponent's win percentage is to see if the team you're evaluating played good teams. But just as records aren't everything alone for a team, same goes for their opponents. To control for opponents' records being a product of playing better/lesser competition themselves, you might want to know the average team's opponent average NET ranking.
That's the line of logic were going with here. In researching what all goes into the NET, it appears they removed scoring margin a few years back from their model. Seems like a mistake to me. Point differential is meaningful in sports. That's going back in. Then you have the Ratings Power Index. That takes into account win percentage, opponents' win percentage, and opponent's opponents' win percentage. I think instead of the latter, we just use opponent's average NET ranking as a better indicator for how good a team's opponent's really are.
Let's hit it.
Here are the top-64 teams as of Sunday morning according to the simplified Stathole rank with NET rankings through March 1st listed to the right for comparison.
Auburn is that team
ESPN losing the feed of the second half of Kentucky/Auburn was pretty sad. I didn't know the guys in the truck were probationary federal workers dealing with random hack job layoffs, but here we are. Either that or some fan ran out on the court with the Epstein list. But one thing is clear - Auburn is a tier above everyone else. The difference in their combined z-scores of the four factors (Special Score) and the second-best team (Alabama) is the same difference between Alabama and #8 ranked Purdue.
This is important info for us casuals. We all know about the chaos of the first couple rounds of the tournament usually simmers down for the big dogs to take over afterwards. Auburn is that big dog this year. I don't know how you can justify picking anyone else to win it all.
Alabama over Duke?
Surprised? Me too. Duke is all the talk of the town. Easy to see why too. They've been clobbering opponents. By far the biggest point differential (+639) of any team. This is especially weird to see Duke ranked so low since we added point differential to the model. But look at the two numbers on teams they've played. Opponent win percentage only 56% with an average NET ranking of 166. Pretty close to the worst opponents overall even if they've won some big games. Including against Auburn.
Meanwhile, Alabama faced much better opponents. Will be fun to see these two teams meet deep in the tournament. I'm siding Bama if they do.
Circle Kentucky as a deep tournament threat
I know they just got put in their place by Auburn - but did they really? Did YOU see it?
OK - yeah they did. But when you look at the average NET ranking of their opponents, it's much better (lower) than most teams around them. If they can escape whatever region Auburn gets in, they can make some noise. They even beat Duke too so who knows!
Purdue, Michigan, Texas A&M all with some value here too. Actually, pretty much every team with a Stathole rank of 8-14. Memphis jumps out a bit there too. Maybe an opening weekend darling play. We'll see.
Then you have your dud candidates. Iowa State / Texas Tech / Gonzaga all standing next to one another being prime examples. But I do want to draw attention to one team us casuals might want to really avoid. UConn. Pretty much rated in line with the big model in the mid-30s. Let's not all into the Dan Hurley trap. He can brag about having two rings all he wants this year. He'll be using the same brag next year as well.
Anyway, thought this would be a fun exercise for us casuals. No clue how any of this will work, but would be interested to hear from the NCAA hardos what they think of some of the teams with big differences in rankings. This is March. Get those brackets ready.