After following the bubble closely for both Sendek and Gottfried, I mostly enjoy the bubble discussions we have around here. The only time that it is not really enjoyable is when the “20 wins” or other such nonsense rears its ugly head…and that’s not been a problem this year. (General apathy has been the biggest problem this year.) The overall purpose of this entry is to outline where the bubble teams stand and whether or not they need something special out of the ACCT. I’ll also briefly outline NCAAT seeding for those teams that have moved comfortably above the bubble.
RPI TREND GRAPHS
But before I get to the bubble details, I wanted to back up and discuss what I’ve been doing this year with the NCAAT classifications of “IN”, “BUBBLE” and “OUT” and why I thought that it was necessary to sometimes ignore my simple bubble definition based on an RPI rank of 41-75.
If you’re going to track NCAAT status throughout the conference schedule, you need to need to acknowledge that the RPI calculation has little/no predictive ability. The RPI calculation is simply a reflection of the past with factors combining a team’s winning percentage and strength of schedule. (Work with me here because I’m simplifying the details.)
The RPI ranking isn’t a horrible system, but sometimes RPI overrates some teams because their strength of schedule is “carrying” a team (especially early in the conference schedule). If a team is truly over rated by the RPI calculations, then its winning percentage will drop as the season unfolds which slowly drags down their RPI ranking. This year, we have Clemson and Pitt as prime examples of how an ACC record sometimes points out a weak team faster than RPI ranking:
Surprisingly enough, we saw GT and Syracuse provide the converse of Pitt/Clemson:
Syracuse had a pretty pitiful OOC schedule and pretty poor results against that schedule. Combine those two things with a road loss at BC to start the ACC schedule, and I wrote Syracuse off on New Year’s Eve. I kept expecting them to start losing ACC games, but they turned their season around (at least at home). GT’s OOC schedule was even worse than SYR, but they started the ACC schedule with a bang by beating UNC in Atlanta…and they kept winning (though mostly at home).
Now GT and SYR may not end up making the NCAAT, but neither turned out to be nearly as bad as you would have thought from their RPI ranking at the conclusion of the OOC schedule. Just two more examples of a mismatch between ACC record and RPI ranking having value when analyzing the ACC teams.
In contrast to Clemson and Pitt, Miami’s turnaround couldn’t have been predicted by anything that I know of:
I don’t follow Miami at all, so I don’t know if someone returned from an injury or if Larrañaga juggled his lineup or exactly what happened. But when they started winning, it showed up on the trend graph and we tracked their progress towards locking in an NCAAT bid. As for as WF goes, how many people would have bet on WF finishing the last week of the regular season going 2-0 against L’ville at home and then on the road in B’burg?
So in summary, I want to identify teams that are trending in either direction. I also want to identify teams that need to pick up the pace AND to identify them as soon as possible. Neither RPI ranking nor ACC record have any predictive ability in and of themselves (remember that L’ville started out 0-2 in ACC play). When RPI ranking and ACC record “agree”, then I feel pretty confident that we’ve got a handle on just where a team sits in the conference hierarchy and with respect to the NCAAT bubble. But a mismatch between RPI rank and ACC schedule is reason enough for me to keep an eye on those teams to see which way they’re really heading.
Sorry (not really) for the long-winded discussion on how/why I sorted out the ACC. I figure that anyone still reading about college basketball (court action not carrousel speculation) is more interested in the fine details than most. So I thought that I would outline my thinking and see if anyone could find holes in my approach.
But that’s enough of that for now. Let’s see where everyone finished up at the conclusion of the regular season:
After the tie-breaking, we’re left with the following matchups in the ACCT:
The ACC match-ups are more important this year than most previous years because the outer edge of the bubble runs down to Tuesday’s teams. This means that Wednesday’s games could be an opportunity for a bubble team to get one of those valuable big wins.
RPI SUMMARY TABLE
(I have no idea what to do with teams where Pomeroy values them much more than RPI. I’m just noting it for now.)
Let’s change things up and start somewhere other than the RPI Rankings and Top 50 wins. It turns out the most total losses to ever receive an at-large bid is 14. There is no reason to think that the Selection Committee has a hard number for maximum number of losses. So that doesn’t mean that there is no chance for those teams already sitting with 14+ losses…just that the odds are probably not in their favor. The Selection Committee has to dig deep enough to unearth a total of 68 teams…no matter how much some of them may stink. But total losses are just one more thing working against Clemson and GT.
Add the ACCT results to the RPI Summary Table and you have the large majority of information that the NCAAT Selection Committee will be using from now through Selection Sunday. But let’s add the Dance Card before we start our discussion:
DANCE CARD SUMMARY
WF’s great week moved them up on the Dance Card and they’re now one spot ahead of VT. A couple of MVC teams snuck into the Dance Card summary and we’ll get to them after we finish summarizing the ACC
If you need an introduction to the Dance Card, catch up here first. Now we have the ACC Standings, ACCT bracket with potential matchups, RPI rankings, Top quality wins, Pomeroy’s rankings (for the first time in one of my entries), and the Dance Card calculations before the conference tournaments start. So what does all of that mean for the ACC teams below the top six?
I’ve changed my mind since last week…I think Miami is IN even though they went 0-2 last week. Compared to VT and WF, they have a slightly better resume in many respects since Miami has more Top 25 wins and a better road win than either team. Top 50 wins from CBS:
First the bad news: Horrid OOC SOS and little success outside of B’burg. We’ve seen the first negative send teams to the NIT, but not with an RPI ranking as good as VT’s. It’s always good to have big wins away from home, but there is no reason to think that it’s essential. If we use the good Sendek years as a guide (and I pretty much always do), VT is probably in the NCAAT. But their position with respect to the Bubble makes me a little nervous. Top 50 wins from CBS:
VT’s best road win is against a bubble team, so it’s hard to know how much weight that win carries with the Selection Committee. A key difference between VT and our Sendek Standard is that State’s big win frequently came in the ACCT…ie technically a neutral court (though the crowd normally cheers for the underdog). So I think that VT is probably IN, but I wouldn’t take out a mortgage to bet on it.
WF managed to get back in the NCAAT discussion in a positive light with a great week. The national press (ie CBS and ESPN) hadn’t updated their discussion on WF as I was writing this, so I don’t have much beyond this table to go on. Even with an early loss in the ACCT, WF is going to be in the discussion for one of the last few at-large bids. I don’t think we’ve ever seen an ACC team with a 0.500 conference record and an RPI ranking in the Top50 sent to the NIT. I hope that this isn’t that year.
Looking at the ACCT bracket, we should have a WF/VT match-up on Thursday. So if those two do play on Thursday, the winner is definitely in and the loser will be a little nervous on Sunday evening.
The NCAAT has the official RPI rankings up through Sat’s games which is not good news for the ‘Cuse. Their six Top 50 wins (all at home) are great, but their current RPI rank of #80, is worse than any team ever to get an at-large bid. An early loss in the ACCT could only make that worse. On top of that, they have a number of bad losses (some really bad) which can’t help their chances.
A win on Thursday against Miami would help, but doesn’t feel like enough. I’m betting that they need a win against UNC on Friday. (But note that I rarely lean towards boundless optimism.)
GT also has some really good wins, but their RPI ranking is approaching triple digits. Four top 25 wins are great, but 3-5 against teams 51-100 and two losses to 101+ are not good. If they can beat ND on Friday, then we’ll reevaluate.
Their best win in 2017 was WF at home. Enough said.
I listed the Pomeroy rankings in the RPI Table, but I have no idea what to do with them as far as decision making goes. So instead, I’ll just make fun of them. Well not really…but sort of.
In the Dance Card entry, I noted that Wichita State may have gotten one of the last at-large bids last year because of their unbelievably high Pomeroy rating. If Pomeroy helped them last year, then they’re sitting pretty good again this year with a Pomeroy ranking of #11. Let’s look at a summary of Scary Wheat’s season thus far from CBS (more stuff here):
My Summary….Wichita St played virtually no one of interest and didn’t do too well against those few good teams that they played. Their Top 25 win was against conference foe Illinois St and the two teams split their regular season meetings with each team winning at home. If I was voting on at-large bids, it wouldn’t be too hard for another bubble team to roar past the Shockers.
Margin of victory (MOV) pretty much has to be the only explanation for their high Pomeroy ranking. Assuming I didn’t fat-finger the info from CBS, here’s their MOV broken down by their opponents’ RPI rank:
People frequently belittle RPI for giving too much credit for Strength of Schedule (SOS). As illustrated above with ACC teams, that complaint can occasionally be justified. But I think that Wich St shows that MOV can also skew rankings far beyond what the team actually deserves. Does anyone really think that Wich St is the 11th best team in the country? Yea, I didn’t think so.
Now I have no doubt that Ken Pomeroy can spin me in circles in a detailed discussion on statistical analysis. But I’ll happily defend my position that good teams prove that they deserve that title by beating other good teams. The best that I can say about Wichita St is that they are one of the tallest midgets.
Now if the Selection Committee is only left with sorting through midgets for those last few at-large bids, then Pomeroy (or Sagarin, BPI, etc) might be as good a way as any to sort through the last few teams. But I would hate to see the Selection Committee switch to using a MOV based system entirely. It’s a real accomplishment when you beat good teams. But I don’t see beating up the weaker sisters of the basketball world as much of an accomplishment.
Side Note – If Greg Marshall can schedule an OOC schedule that isn’t truly embarrassing, then there is simply no excuse for a team from a power conference doing substantially worse.
Wichita St and Illinois St are leading the MVC and both are close to the burst line on the Dance Card. So we’ll see what we see on Selection Sunday and try to figure out if Pomeroy played a role in at-large bids to the MVC.
PLAYING FOR HIGH NCAAT SEED
As we’ve discussed in the past, about 75% of the NCAAT teams are seeded within +1 seed that you would calculate based solely on RPI. We’ve also seen that finishing first or second in the regular season conference standings generally mean more with seeding that conference tournament results. But there were enough stumbles among the Top Six down the stretch, that we’re not left with many ACC teams truly vying for a #1 seed.
If UNC wins the ACCT, then they’re clearly a #1 seed. Even with a loss in the ACCT finals, then they’re still in the conversation. Louisville is probably going to end up as #2 or #3 seed. They might be in the conversation with an ACCT title, but would probably need a lot help from around the country.
For the rest, divide RPI by four and round up, look at Lunardi at ESPN, and Palm at CBS if you really care about the seeds for the ACC. Sorry, but I can’t work up enough enthusiasm to read tea leaves for that…I’ll just look at what the Selection Committee decides on Sunday evening.
Clearly the ACC will end up with the most teams in the NCAAT. As long as we ignore State’s struggles, it’s been an interesting year. But for the second year in a row, State starts the ACCT on Tuesday at noon. At least around here, BWW has traditional wings for half-price on Tuesday so there’s at least that to look forward to during the first half of State’s game.
OK, I’m heading out for my real job now. You kids behave while I’m gone and I’ll respond to questions and ridicule later.