The Virginia Cavaliers' NCAA Tournament Hopes: Part I

It's hard to overstate the rollercoaster that has been 2012-13 Virginia men's basketball. Chronologically, the team beat Wisconsin, lost to Old Dominion, beat UNC, lost to Wake Forest, beat NC State, and lost to Georgia Tech. Up and down and up and down and up and down. The inconsistency makes it difficult to evaluate the success of the Hoos season, particularly as it pertains to the goal of any college basketball regular season: making the NCAA Tournament.[1]

As fellow Streaking the Lawn writer Brian Schwartz recently noted, the Hoos have some decidedly uncertain postseason credentials. ESPN alone has the team both on the verge of making the Tournament and with "much work left to do." The Hoos have ten games[2] to leave an impression on the Tournament Selection Committee when they take the floor tonight at JPJ.

In this two-part post, Brian and I will expand on the concepts he mentioned in the above article by detailing various scenarios that could play out between now and Selection Sunday. This portion, part I, will work backwards from a goal RPI to arrive at a breakdown of season outcomes with historical Tournament chances. Brian's Part II will elaborate on the likelihood of each final season result and the Hoos overall prospects of earning a spot in the Big Dance.

[The "Goal" RPI for Tournament Consideration]

Tracing the effect of the Tournament bubble on fellow ACC teams over the past five seasons gives a sense of the "resume" the Hoos need for serious consideration. The chart below compares lower-seed Tournament teams from the ACC to those who were banished to the NIT.[3]

The NCAA Tournament Selection Committee claims to rely on a number of factors in addition to the notorious RPI. Fans and pundits remain skeptical, insisting that Tournament hopes begin and end with the RPI ranking, and the above chart does nothing to soothe these fears. In each of these five seasons, no ACC team with a lower RPI was chosen over a team with a higher RPI.[4] Making the NCAA Tourney from the ACC appears to simply be a function of where the RPI cut-line falls.

Based on this data, any RPI better than 50 is safely in, even with a losing ACC record.[5] Any RPI worse than 65 is definitely out. We'll break the 50-65 bubble into three smaller bubbles containing varying success rates.

Teams ranked from 50-55 made it in four out of five times. If the Hoos end up in this range there is certainly some cause for concern, but since they're not the Hokies, I'd guess they're in. Things get complicated between 56 and 60. Only two of these six teams[6] made the Tournament, so if the Hoos finish here, wear your lucky socks, say ten Thomas Jeffersons and prepare for a coin flip. The lowest range from 61-65 is not promising. The Hoos would most likely be out, but could have a very small chance with some late season magic that inspires the Committee to pick them as a surging team.

[NCAA Tournament Chances based on Remaining Results]

This rough idea of the RPI necessary for a Tournament berth, coupled with the estimations from RPI Forecast previously mentioned by Brian, give us a sense of how the performance over the remaining ten-plus games will affect Tournament chances. Better yet, RPI Forecast now includes[7] potential conference tournament results to more accurately reflect the RPI that will be seen by the Selection Committee.[8]

A summary of the Hoos estimated chances of reaching the Tournament based on the results over the remainder of the season appear in the chart below:

** The mind-boggling misses denoted by the asterisks[9] reveal the major caveat to this analysis: the first chart was based entirely on historical RPI results. This isn't usually a problem, but in this case, the past doesn't really account for the bizarre RPI / record pairing the Hoos could easily have. Consider: if the Hoos finish 8-2 and lose in the first round of the ACC Tournament, they'll effectively[10] have an ACC record of 13-6 with an overall record of 23-9. Those records scream "IN." Their RPI in that scenario, though, is projected at 67, or a few spots below "serious danger zone."[11] Now you see what people mean when they say the Hoos resume is strange.

We could see a horribly[12] fascinating, epic tug-of-war between the Selection Committee and reason. If the record plays out according to the asterisks, we're looking at a monumental RPI debate; it would probably be the hottest topic in college basketball on Selection Sunday. Despite this caveat, the past is all we have to go on. Nothing suggests the Selection Committee wouldn't adhere to RPI as strictly as they have before. Just prepare yourselves for a potentially wild ride.

Now, head over to Part II, which elaborates on the results of the second chart ...

______________________________________________________________________

[1] Not to say that the season hasn't been a success otherwise. The fact that the NCAA Tournament is a legitimate discussion is a great sign for this young team. Certainly makes whoever wrote this look dumb.

[2] Not including the ACC Tournament, which we'll get to in a minute

[3] Could have re-titled this chart: How Tech Defines the Bubble. Seriously, that's just awful. After reading that fourth miss in a row I think some miniscule part of me even begins to feel a little bit bad. Must be getting soft in my old age.

[4] Go ahead, look for yourself. I'll wait.

[5] Although it's hard to say what would have happened to 7-9 GT in '09-'10 if their RPI had been in the mid-40s. That team is just a weird outlier. As we'll see below, the Hoos could be the opposite type of outlier this season.

[6] Seems strange that, in just these five seasons, six ACC teams fell into this small five RPI rank range.

[7] I came up with a separate method of predicting this and wrote it up before RPI Forecast was updated. The results were very close, but theirs is way more sophisticated. Just know that you were spared a good 2-3 paragraphs.

[8] Since, in the eyes of the RPI, a conference game is a conference game whether it happens in a tournament or the regular season.

[9] I used a separate "*" footnote here for two reasons: (1) I couldn't figure out how to correspond footnote numbers from the chart into the document, but more importantly, (2) it's a long and critical note that shouldn't be glossed over.

[10] because it would be including the additional ACC Tournament loss

[11] my "lower side of 50/50" is even a little generous

[12] for us