ICYMI: The Anatomy of an Atlantic 10 Winner

This piece was originally posted on TPOP 1/12/18, but I thought you might enjoy it if I dusted it off again for another look.  This will be particularly helpful for a piece I have in mind for next week.

Even though 100 teams (68 in the dance / 32 in the NIT) make the NCAA sanctioned March Madness events, it’s still the most difficult of any major American sport to make its respective post season. Sure, that’s a lot of teams, but over 350 teams compete in Division I. Less than a third of teams make it. With two wildcards, a third of MLB teams do. 38% of NFL teams do and half the NHL and NBA do. Bowl games in FBS football are basically a participation trophy.

As well as Duquesne’s season has gone so far, fans should keep this in mind when they start asking questions like “Can the Dukes make the NIT?” or “What do the Dukes need to do to make the NCAA tournament?” I had both these questions asked of me in the last two weeks, but the more relevant questions might be “What does a post season team from the Atlantic 10 look like?” I was asked that one too and I will do my best to answer it here. For me, that would not only help clarify what it takes this season to make it, but give folks a rough idea of what it would take in the future. In truth, Duquesne fans haven’t needed to worry about either much, so treat this as a guide to what it takes.

My answers to the other two questions, by the way: 1) win 13 or more games in the conference regular season including two of three against URI and St. Bonaventure, then make the semifinals of conference tournament. It’ll all make sense later, I promise. 2) Win the conference tournament.

I want to try to break down the qualification for the NCAA post season for a more casual fan than I would normally write for. If it seems like I’m over-explaining to some of our more savvy readers, it’s because I am. For good or bad, whether or not a team gets a bid to the either the Big Dance or the NIT revolves largely around a computer rating system known as the Ratings Percentage Index or more commonly the RPI. The RPI looks not just at wins and losses, but the quality of those wins and where those games are played on a home, road or neutral court. As the season progresses, the RPI also considers how well the opponents of a respective team play and weights wins against successful teams more highly than teams that are less successful.  The aggregate of how well scheduled teams perform is known as the strength of schedule or SOS.  In many cases teams with worse records but higher SOS can rate above teams with lower SOS.  The RPI does not consider the final score, so 20 point win gets the same value as a one point win. Games against Division II do not count. Also note that Ken Pomeroy’s numbers on Kenpom.com or Jeff Sagarin’s Index are independent rating systems and do not factor into the at large conversation, as far as I know.

The RPI is important for two reasons. First and foremost, teams with high RPIs are more likely to qualify for an at large bid to the NCAA tournament almost automatically. By and large, if a team is in the top 35 they’ll get an at large to the NCAA tournament from the A-10. There is only one exception to that rule in the last five years, St. Bonaventure. RPI also influences the more subjective parts of selection. If you ever hear the concept “NCAA resume,” that’s another way of saying how many good teams a school has beaten and how many bad losses have they avoided.  While there are more nuanced ways of breaking this down, a good resume has a lot of top 50 and top 100 RPI wins and minimizes losses to anyone rated 101 or higher. Losses to teams with an RPI 200 or higher can end up as a real resume killer.

Make sense now? If not, feel free to comment below and I’ll try to clarify. It’s simple once you get it, but intimidating if you don’t. I’ll do my best answer any questions you may have.

I looked at who the Atlantic 10 sent to the NCAA tournament over the last five years to help clarify what a team needs to do to keep playing in March. The league has sent 19 teams to the NCAA tournament and 9 teams to the NIT since 2013 (not including 2018). The average RPI of the NCAA teams from the A-10 is 28.6. La Salle had the highest at 46 when they played in and won their 2013 First Four contest. In the same sense, the NIT teams averaged a considerably higher 65.6 with only two teams with an RPI in the 80’s making the cut from the league. The seven others were mostly in the 50’s and 60’s. Only one team made the NCAA tournament from the Atlantic 10 with a SOS higher than 100 when 2015 Davidson squeaked over that line at 101. No NIT team had an SOS over 150.

I was going to do a nice paragraph breaking down the differences between NCAA and NIT resume, but the word counts getting up there already so you’re getting a chart with the average amount top 50 and 51- 100 wins as well as 101-200 and 200 + RPI losses from each group.


I don’t need to break these numbers down too much, but simply put it takes almost twice as much on the good side of the ledger to get an at large bid and about half as much of the bad stuff. When we look at the resume of that 2016 St Bonaventure team with the 30 RPI that missed the NCAA tournament, we notice three Top 50 wins, but also three 51-100 losses along with two 101-200 losses and one 200+ loss. Only one team from the A-10 made it with those types of losses, Temple in 2013, but that team had an extra top 50+ win and a better SOS. Only two Atlantic 10 teams are top 50. If the current Atlantic 10 becomes the new normal for the league, it will evolve into a one bid league. Not only won’t the top teams qualify, but they won’t have enough other top 50 teams to build their resume in conference.

I looked at some other factors in terms of the makeup of an NCAA team in the A-10, but they were largely unremarkable. NCAA teams are more veteran averaging 4.9 juniors and seniors among their eight most heavily used players compared to 4.6 for the NIT, though the NIT was more junior heavy. As one might also have guessed, the NCAA teams have dominated the all-conference teams. Most of them have multiple players on the first and second team. Only one team, Charlotte in 2013, made the NIT without any all conference players at all. Having a mature roster full of good players helps a team succeed in the A-10. Sorry to state the obvious, but recruiting class worship leads to fans sometimes exaggerating the immediate impact freshmen have. Although plenty of frosh contributed on excellent A-10 teams, they’ve generally had significant veteran support around them.

As promised I’d like to circle back to Duquesne in 2017-18 and why they need to win 14 more games to even have a shot at the NIT. Despite their hot streak, the Dukes RPI still sits at 202 meaning they have some significant ground to make up. They played one of the softest non conference schedules in the country and despite playing 3 league matchups, their SOS is still 333. The SOS will never recover no matter how many games the Dukes win and it mean the Dukes will need to make up crap games early with greater raw wins. I used RPIForecast’s RPI Wizard, a tool that allows you to manipulate a team’s wins and losses over the course of a season to help predict what their RPI could be under a set of circumstances different from what has or will happened. If I set every game as a win expect for at St Bonaventure and at VCU and have them beat Davidson in the Atlantic 10 tournament before losing in the semis, the Dukes RPI would end up around 73. Just good enough to think about the NIT. They’d have two top 50 wins and two 51-100 wins, both against Dayton. They’d also have 4 100+ RPI losses, which isn’t a complete deal breaker for the NIT, but doesn’t look great either. However, even after doing all that the Dukes would only be on the bubble.

In reality, the Dukes are probably CBI bound. They need a .500 record to qualify and if they win four more games they just need to pay to host a game. Not the most romantic post season berth, but certainly better than no basketball for seven and a half months. Keith Dambrot didn’t schedule to make a post season appearance, just like he didn’t set up a roster to make a post season appearance. If he could’ve used his crystal ball to see what his team would become, he may have replaced a couple of cupcakes with some middling schools or even a road game against an elite school. They’re playing like a top 100 team on the NIT bubble now, but they don’t look good in the computer system. Only a magnificent run will change that for them.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s