Bluh Blup Gurgle Plah

I never intended to drop this article on Memorial Day but it lets me take the opportunity to remember the fallen and thank the family and friends of servicemen and women who have lost their lives while serving our country. Never forget those that made the ultimate sacrifice.

I’ve talked a lot about the NCAA criteria over the last two years but have only touched on the concept of bad losses. We saw our New England friend, “Ned”, come up with a Quality Win Index (QWI) that attempts to award teams with quality wins based on the Pablo Rankings. His index does take into account bad losses to a point but I wanted to look at it a little differently. I want to tie bad losses to the regional rankings, which is problematic as it doesn’t account for the depth of quality in regions, but at least it ties us in with the criteria a bit.

We know Ranked Wins is an NCAA criterion. We know that the data sheets provided show Ranked Wins as a won/loss record and includes a percentage. What we don’t know is how the RACs or the selection committee use this data. Our fear is that they only look at the percentage without looking at the teams that make up the record. Ranked Wins is further flawed by that pesky issue of the different quality of depth in each region and the fact that the larger regions are allowed to rank more than 8 teams. Initially, I decided to look at bad losses as simply the number of losses a team has to an opponent that is not regionally ranked. I jotted down all of the at-large bid teams from 2018 and added in the bubble teams. I then added in UT-Dallas because I still think the West RAC made a mistake by having La Verne ahead of them. I then noted their bad losses. DePauw had the honor of the most bad losses (5) of the at-large teams followed by Augsburg with 3. Of the bubble teams, Concordia (WI) and La Verne both had 4 losses. If you remember from my Selection Eve article where I try to determine the at-large bids, I really struggled with DePauw due to their bad losses.

The NCAA recently considered (and then rejected) a proposal to rank a maximum of 10 teams per region so that got me to thinking about how this would change the bad losses for each of these teams. In theory, I figured it would hurt the larger regions so let’s see if that plays out. In reality there wasn’t much change throughout any of the teams. Springfield went from 1 bad loss to 2. St. Thomas and Wesleyan picked up their first bad loss. Susquehanna (a bubble team) went from zero bad losses to 2. Stevens went from 2 bad losses to 3 and La Verne went from 4 bad losses to 5. Certainly interesting but nothing that indicates a pattern. I then recalculated Ranked Wins based only on teams ranked 10^th or better. This would give us an indication of what teams padded their Ranked Wins criterion with lesser teams. At-large teams that dropped two ranked wins or had their Ranked Win percentage go under 0.500 were Augsburg (4-5), Carnegie Mellon (5-9), Mary Washington (6-10), Franklin & Marshall (4-5), Chicago (7-4), Wellesley (4-5), Springfield (3-5). From the bubble teams, MIT dropped 2 wins to fall to 3-5. Of the eight teams mentioned, 6 are from the large regions. The NCAA is unsure if ranking more teams is a benefit and I’m thinking this proves that it is assuming that the selection committee doesn’t do deep dives into the Ranked Wins criterion.

What if we take this further and limit each region to ranking only 8 teams? With respect to bad losses, Springfield would now have 4. (Circling back, that means they have 4 losses to teams that were not regionally ranked 8^th or better.) On the bubble team side, Stevens would also pick up their 4^th bad loss. Overall, we have La Verne and DePauw with 5 bad losses and Springfield, Concordia (WI) and Stevens with 4. Now looking at the Ranked Win criterion, and keep in mind that I’m also disregarding the teams that were ranked the previous week but dropped out during the final ranking, we have some changes.  Carnegie Mellon goes from 7-9 down to 4-9. Franklin & Marshall goes from 5-5 down to 2-4. Mary Washington goes from 8-10 down to 4-10. Wesleyan goes from 6-5 down to 3-4. Wellesley goes from 6-5 down to 2-5. Springfield goes from 5-6 to 3-3. Of the bubble teams, Susquehanna drops a ton from 7-13 down to 3-11. MIT goes from 5-5 to 3-3. If you are scoring at home, every team just mentioned was from a large region, which, of course, makes sense.

I now want to look at bad losses again against the backdrop of the total number of Top 8 regional ranked teams a school played. Looking at the at-large and bubble teams, most played 8 or more Top 8 regionally ranked teams. Those that didn’t were Franklin & Marshall (6), Wesleyan (7), Wellesley (7), Springfield (6), MIT (6) and Hendrix (7). Franklin & Marshall was a shock NCAA at-large selection and their 5-5 Ranked Win criterion was a factor but if you looked at them with a 2-4 Ranked Win criterion factored in with 3 bad losses do they make it in? How about Springfield with 4 bad losses and only playing 6 Top 8 regional ranked teams (3-3)? I believe the Ranked Win criterion insists that the committee look at not only the percentage but also the quality of the ranked opponents and the opponents not regionally ranked where a loss was incurred. That, in a sense, is what we just did here.

One area I left out with regards to Ranked Wins is significant wins. These are the wins against elite teams. The teams that have a really good chance at winning the NCAA championship. I believe a team showing me they can beat an elite team is worth something but unless Ranked Wins is evaluated fully, this means nothing. I looked at the Top 10 teams going into selection based on the AVCA rankings. (I know those rankings are pretty trashy, but regardless of order they will typically indicate who the coaches think are elite.) The at-large eligible teams that took down a Top 10 team during the year were Cal Lutheran (twice), Trinity (twice), WUSTL (twice), Birmingham-Southern (twice), Ithaca, Hope, Carnegie Mellon and Concordia (WI). All of these teams made the NCAA Tournament except for Concordia (WI).

At the end of my analysis I have bad losses and I have significant wins for the at-large and bubble teams. I also have a modified ranked win/loss record as well as the total number of matches played against the Top 8 across the regions. If you remember from my Selection Eve article and my projected teams for the NCAA Tournament article, I didn’t believe Franklin & Marshall and Carthage deserved a bid based on the selection criteria. I also struggled with the order of Springfield and MIT in the New England Region ultimately selecting Springfield in a last minute change. For the record, I had both La Verne and UT-Dallas making the tournament but these were the last two teams I selected in my mock selection committee role. Going through the at-large teams again and I would probably struggle even longer over my selection of DePauw as they have 5 bad losses. What I come back with on them is that they had a horrible start to the year but have some of the stronger wins of the at-large teams. I’ll keep them. I would probably take some additional time with Carnegie Mellon with their modified 4-9 Ranked Wins criterion but their significant win (over Emory) probably keeps them in the tournament. Franklin & Marshall were not helped with my analysis and they certainly stay out of the tournament with a modified 2-4 Ranked Wins criterion, 3 bad losses and only 6 matches against the Top 8. I had Mary Washington ahead of Franklin & Marshall in the Mid-Atlantic while the RAC had the reverse so in a sense I never would have added Mary Washington since I wouldn’t have been allowed to consider them. Carthage wasn’t helped with my analysis and remains outside looking in. The New England Region is a mess as Wesleyan, Wellesley, Springfield and MIT were all hurt by my analysis. They simply didn’t play an NCAA Tournament at-large quality schedule. The 4 bad losses by Springfield really bother me but they did have a head-to-head over MIT so maybe I still give them that nod but in no way do I now select them for the tournament. Wesleyan and Wellesley are a problem for me, as well, but for now I’ll put them in. The rest of the teams remain in the tournament but I’m now faced with 4 open spots (Franklin & Marshall, Mary Washington, Carthage and Springfield). If I’m playing by the rules, I can’t consider blocked bubble schools like Susquehanna, Concordia (WI) and MIT. But, I’m the selection committee…the NATIONAL VOLLEYBALL SELECTION COMMITTEE…so I put on my NCAA Associate Director of Championship Jan Gentry hat, mute the line with the New England RAC Chair and make some changes. Mary Washington is placed ahead of Franklin & Marshall and despite their modified Ranked Wins 4-10 criterion I’m putting them back in. I move Concordia (WI) ahead of Carthage valuing their head-to-head victory over Carthage and their win over Chicago. I’ll be taking a look at Concordia now. Heck, I also move UT-Dallas ahead of La Verne as the Leopards 5 bad losses loom large in my mind (in this case I don’t care about La Verne’s head-to-head win over UT-Dallas…I know call me fickled).

Three spots with Mary Washington back in and I turn my focus to Concordia (WI). I really want to put them in and reward them for their victory over Chicago but they also have 4 bad losses and a low SOS. They stay out. I’m really looking at Bethel, Muskingum and UT-Dallas for these spots. Bethel and Muskingum are really close but with no bad losses I tip the scales to Bethel and they are in. I really like UT-Dallas and their 3 ranked wins over Trinity, Cal Lutheran and DePauw (3 at-large teams already in the tournament) so they are in. One spot left and I’m looking at Wartburg (too many bad losses), Muskingum, Carthage (yeah, sue me) and La Verne. There really isn’t a lot to separate those last three teams but with no bad losses and playing a tough schedule I’ll give the nod to Carthage so they are back in where they were although I never had them in but do now. Nearly 2,000 words to basically say that when considering bad losses (along with a lot of other things associated with Ranked Wins) Franklin & Marshall and Springfield are bounced for Bethel and UT-Dallas.

This article has sat in Word on my computer for nearly 2 weeks. This happens to a number of the things I write when my initial lofty goals for an article sort of slosh out through the margins until I’m left with something I’m not sure should be published. I intended to focus on bad losses but in the end that one statistic wasn’t as telling as I thought it would be. My thoughts and efforts drifted to other ideas and in the end I basically have another analysis of the NCAA at-large selections based on a deeper Ranked Wins criterion viewpoint. And when it’s all said and done does it really matter? Does it really matter that Franklin & Marshall and Springfield made the tournament over other teams? In the case of Franklin & Marshall I think it does because they were so obviously unqualified for the honor. In the case of Springfield I think it also matters because I think the selection rewards a region for not really trying. At-large selections are messages sent directly to the coaches that didn’t make it in the tournament as to what they need to do to make it the following year. What do these two messages tell me? In the case of Franklin & Marshall the message is bluh blup gurgle plah. In the case of Springfield it’s about having to do more to compensate for the bias inherit in the Ranked Wins criterion towards the larger regions. Honestly, I think both messages make about the same amount of sense.