Welcome to the official Blogger So Dear bubble breakdown. In this piece, I will summarize the bubble landscape, but more importantly for this audience, I will handicap where Wake Forest stands in the larger bubble picture (hint: if you read the headline, you probably know the answer). We have a lot to get through, so let’s get started.
Back on Tuesday, I began to analyze the bubble picture by classifying the conferences. That remains the best place to start. Here’s that table, updated:
Most of these designations are fairly self-explanatory. The Missouri Valley is right in the middle, due to the uncertainty regarding Illinois State. The MAAC, Sun Belt, and Ivy are not quite in the “Definite One-Bid” category, due to the slight bubble chances of Monmouth, UT-Arlington, and Princeton, respectively.
Undoubtedly, the nine conferences on the far left will claim the vast majority of the 36 at-large selections. Here’s the table of teams that are overwhelmingly likely to claim a slot, otherwise known as “locks”:
For each conference, I’ve listed the automatic bid winner on the far left. Next, I count the number of locks each league has, excluding the auto bid winner. These are then summed, currently totaling 27. Note that we don’t yet know the winner of the Big 10, SEC, or AAC, but none of those championship games will change the math: all six teams involved are locks. The main situation to watch is the Atlantic 10, where Rhode Island, a non-lock, has a chance to win the automatic bid.
With 27 slots accounted for above, there are nine slots available for bubble teams (36-27). Note that if VCU defeats Rhode Island, they would move out of the lock count and into the auto bid pool, which drops the lock total to 26 and raises the bubble count to 10 (36-26).
From a seeding perspective, we can also address where these 10 teams should fall in the S-Curve the committee uses to rank the teams. The above table should account for all of seeds 1-9 (27 locks + 10 auto bids = 37 teams). Meanwhile, there are about 22 auto-bid winners that will be seeded behind the bubble teams, accounting for seeds 12-16 (remember there will be six 16-seeds because of the First Four). In short, the bubble teams will essentially be seed lines 10 and 11, with possibly 1-2 exceptions on the 9-seed or 12-seed lines.
So, we’ve established how many slots are available for bubble teams (9 or 10), and where those spots should fall (primarily seed lines 10-11). Who’s on the bubble? In that same Tuesday piece, I compiled an overly-inclusive list, based on the Matrix, the RPI, and my own resume analysis. Here is that table, updated to include notable mid-majors that failed to capture their league’s automatic bid:
This table follows the same format as the previous one. By my count, there are 28 bubble teams (again, being considerably over-inclusive just to be safe), fighting for nine slots.
(Note that I’m temporarily double-counting Rhode Island; if they win today, this table becomes 27 teams for nine slots. If they lose today, this table becomes 28 teams for 10 slots.)
So, as we start to transition to thinking about Wake Forest’s chances, there are a couple key numbers to keep in mind. First, Wake Forest needs to rank in the top nine or 10 of this group in order to make the NCAA Tournament. Additionally, Wake Forest will need to rank in the top five or six of this group in order to avoid the First Four.
Still with me? I think that’s enough procedural background. Let’s get to the fun part.
Where does Wake Forest rank against these 27 bubble teams?
Below, I’ve compiled a comprehensive bubble spreadsheet. Before you get overwhelmed by the sheer quantity of colors and numbers, let me break it down a little bit.
First, the table starts with five different computer rankings, which are then averaged. In this section, I tried to blend together a wide variety of metrics. Kenpom and BPI represent the “efficiency-based” side of the spectrum. On the other end, despite it’s many flaws, I of course had to include the RPI. I also used ESPN’s Strength of Record (SoR) metric, which “is a measure of team accomplishment based on how difficult a team's W-L record is to achieve.” Historically, this SoR metric has correlated remarkably well with the Selection Committee’s decisions. Finally, to help bridge the gap, I added in the always reliable Sagarin rankings.
Beyond those metrics, I included a number of factors that often surface in bubble discussions. Working from left to right, the table includes each team’s number of RPI Top 50 wins, RPI Top 100 wins, and respective win percentages. It includes their number of bad losses (RPI > 100), and their overall worst RPI loss.
Next, I include each team’s overall record, broken down into wins, losses, and win percentage.
Of course, after wins and losses, it’s important to consider the difficulty. So the next four columns represent attempts to measure Strength of Schedule. The first two are from the RPI, while the latter two are from Kenpom. “NC” stands for non-conference.
After the schedule strength, I include pertinent conference information as well. Each conference is ranked according to the RPI and Kenpom. Then I list what place in the league the bubble team finished, as well as their conference win percentage.
Finally, the last two columns are simply a count of the number of road and neutral site wins each bubble team earned.
In every column, the spreadsheet is formatted such that:
green = good
red = bad
Now, without further ado, I present the 28 bubble teams, in alphabetical order:
Ok, I have to admit, that’s pretty hard to make heads or tails of. Let’s sort by the average metric ranking:
Much better. So what can we draw from this composite? Number one, the Deacs rank number one! And no, I did not “rig” the spreadsheet; I selected all the metrics before I compiled any statistics. Furthermore, pick any combination of metrics, and you’ll end up with the same result:
Among these 28 bubble teams, Wake Forest ranks in the top 5 in each of the five metrics (RPI, Kenpom, BPI, SoR, and Sagarin).
None of the other bubble teams can say that. Outside of Xavier, none are even particularly close.
There has been quite a bit of speculation regarding the relative weight the Selection Committee will place on the RPI, in comparison to metrics like Kenpom. What this analysis proves is that, no matter how far along the committee is in that transition, Wake Forest is in great shape.
Scanning Wake’s row, you can quickly spot our relative weaknesses. Obviously we do not have a good win percentage against the RPI Top 50 (3-10), and we did finish 10th in our conference. Other than that? Wake is pretty solid across the board. We should get major points for our strength of schedule — both overall and in the non-conference — and conference affiliation. Not only does Wake not have any bad losses, but the Deacs have the best worst loss of any team in the table. Wake also has tied for the most road wins among the Power 5 bubble teams.
Big Picture for Wake Forest
Does Wake truly have the best resume of these 28? That’s debatable, and it depends on how you choose to weigh the factors. Beauty is in the eye of the beholder, and this is a subjective process. However, the key point I want to emphasize here draws on the initial analysis:
Either nine or 10 teams from this table will make the NCAA Tournament
So, it really doesn’t matter if Wake is number one. The main key is to be in the top 9/10. And quite simply, as subjective as this process is, I do not see a realistic scenario where nine or 10 teams from that table are selected over Wake Forest, and that’s why I feel comfortable with the emphatic headline.
The secondary goal is to be in the top 5/6 of that table, which will allow us to avoid the First Four. Personally, I absolutely believe we’re in the top five of these 28. Truth be told, I think we could stack up well against some of the teams in the “lock” category as well. We really have the resume of an 8 or 9-seed, considering that the metrics unanimously place us in the 30s. Unfortunately, the conventional wisdom regarding Wake’s standing has lagged behind the metrics for much of the year. I won’t be surprised at all if we are sent to Dayton, even though I don’t think we should be.
Other notable takeaways:
-Which bubble team might be the best bet to pull some “upsets” in the tournament? The answer appears to be Marquette, as they rank number one in each of the more predictive statistics (Kenpom, BPI, Sagarin). They also have plenty of experience defeating high-quality teams, leading the group with seven RPI Top 50 wins.
-Despite the late-season struggles, Xavier’s resume is quite strong across the board. They should safely make the field.
-Vanderbilt’s RPI strength of schedule certainly jumps out, and should lead them to set a new record for losses by an at-large (15). They also lead the table with 11 RPI Top 100 wins.
-Rhode Island’s resume is stronger than I expected. They are the only team in the table with a winning record against the RPI Top 100, for example. Plus their metrics are consistently solid. I tend to think the A10 is getting three bids, regardless of the championship game outcome. A VCU win would still be very helpful to our chances of avoiding the First Four.
-The top 10 in my table matches 8/10 currently in the Bracket Matrix. The two discrepancies are Clemson and Indiana instead of USC and Providence. This is no surprise to those who follow Kenpom’s “luck” metric. The predictive metrics tend to like teams with a host of close losses more than humans do, and Clemson and Indiana certainly fit the bill. On the flip side, note that USC and Providence are boosted by their RPIs and SoRs primarily, as they both have records better than their efficiency margins would indicate.
I fully expect the Wisdom of the Crowds to beat my crude spreadsheet here. This analysis is more geared to developing an objective jumping off point for bubble analysis. In this case, I think it just shows that Clemson and Indiana were probably closer to the tournament than their conference records would suggest. Flip one or two of their nail-biting losses and the narrative is much different. The inverse holds for USC and Providence as well.
-For me, the most interesting case of the day is Syracuse. They stand out, both good and bad, all over the place. The Orange rank 84th in the RPI, 31st in BPI. They have two (2!) wins outside the Carrier Dome, but finished above .500 in the country’s top ranked conference. They have six RPI top 50 wins, but five sub-100 losses. If I had to guess what the ESPN talking heads will be yelling about around 7:00 tonight, Syracuse is my answer. If they are indeed left on the outside looking in, expect to hear the analysts criticize Wake’s inclusion, given the head-to-head result and their superior conference record.
-Best example of a team still in consideration on the Matrix, that we really shouldn’t worry about? California. Outside of total wins and conference record (by one game), we trump them across the board.
-Shoutout to Texas Tech for compiling the worst non-conference schedule in the entire country.
-Back on February 4th, I remember thinking to myself that Georgia Tech is incredibly bad for a team in position for an at-large bid. This analysis confirms my suspicion.
And there you have it. If you find a more comprehensive Wake Forest bubble breakdown on the internet, I’d love to read it, so go ahead and put a link in the comments. I hope you all enjoy Selection Sunday. Go Deacs!