Tuesday, December 27, 2016

Bracket Projection: 12/27

The very first Crisler Spider-Man Bracket is here! Big thanks to to my known friend and trusted agent Ryan Teich, who threw together some great VBA code that allows me to actually do these projections at a fast enough pace that will make them a regular installment on this blog for the rest of the season. A few quick notes before we get to the actual bracket:

Some seeds are weird - You may wonder why, for example, Indiana and Minnesota are both 7 seeds, when common sense tells you that Indiana is much better. The answer to almost any question about weird seeds is RPI. Some teams (like Indiana) played a few high profile games in between walkovers against the dregs of Division I. This leads to a very low RPI that will certainly rise during conference season. Big Ten teams tend to schedule this way (look no further than Ann Arbor), and it has hurt their seeding over the years. Maybe next year they'll learn their lesson. Until then, you'll see a lot of low seeds for Big Ten teams, despite the projection putting eight teams in from the conference.

The regions are just for fun - Brackets are evaluated based purely on seeding, and how close my seed predictions are to the actual thing. I have them presented in regions just for the fun of it, so we can use our imaginations to see what a real bracket might look like. There's no consistency for the committee in choosing which regions get play-in games, and sometimes they aren't even 11 seed play-ins, so anything could happen. Regardless, you can still have fun imagining a Michigan-Notre Dame rematch with the winner playing Kentucky.

Here is the first bracket. You can compare it to Bracket Matrix for a sanity check. I'll be contacting them shortly to see if I can get my picks included in their database. Most of my seeds are pretty close to their averages; some are a bit off. The bold city names at the top of each column are the locations of the Sweet 16 and Elite Eight.

1 Villanova Louisville Baylor Gonzaga
2 North Carolina Kentucky Duke Kansas
3 UCLA Creighton Butler Virginia
4 West Virginia Arizona Florida Xavier
5 USC St. Mary's Oregon Purdue
6 Wisconsin South Carolina Florida State Cincinnati
7 Minnesota Notre Dame Arkansas Indiana
8 Virginia Tech Maryland Miami FL Clemson
9 TCU Oklahoma St. Seton Hall SMU
10 Iowa St. Michigan Northwestern Texas A&M
11 Pittsburgh Wake Forest Wichita St. UT Arlington
11 VCU Illinois
12 Monmouth Middle Tennessee Dayton UNC Wilmington
13 Chattanooga Tennessee St. Valparaiso Fort Wayne
14 Vermont Winthrop Akron San Diego St.
15 Princeton Florida Gulf Coast Lehigh Texas Southern
16 Wagner North Carolina Central Sam Houston St. Long Beach St.
16 North Dakota Cal St. Bakersfield

1 seeds? Three of my four original 1 seed predictions are currently sitting on the 2 line. I expect Baylor to fall off after a few conference losses. Louisville has a better RPI than Duke right now, so they take a 1 seed that will be settled by the grueling ACC schedule. As for Gonzaga, they had been off my radar for most of the year, but they're currently undefeated and a significant favorite in every game except one the rest of the way, so they're actually a decent bet to hold on to that 1 seed the rest of the season. Finally, Villanova is a 1 seed in every single prediction I've seen, and currently have the best chance of being the #1 overall seed come March.

Big Ten? As mentioned earlier, the Big Ten is pretty under-seeded right now. It would be quite unusual for the highest Big Ten seed to be a 5 (the last time it happened was 2004, when a very down Big Ten only sent three teams to the tournament). I expect someone from the Wisconsin-Indiana-Purdue tier to climb their way up as they take control of the conference. Minnesota, Maryland, Michigan, Northwestern, and Illinois will battle it out with fellow bubble teams Ohio State and Michigan State, who are currently on the outside looking in (not just because I'm a homer I promise). Just about any combination of those 7 bubble teams could make the tournament and I wouldn't be all that surprised.

Michigan? That's right, I just called Michigan a bubble team...for now. I truly expect them to work their way out of that label, possibly pretty quickly. Simply going through the motions of a Big Ten schedule will raise their RPI (which currently sits at 76), and Kenpom still has them as the fourth best team in the conference with a projected record of 10-8. Speaking of which...

10-8 and 11-7 are almost equally as likely. I would expect the difference between those records to lead to a difference in seeding, depending on the quality of that extra win or loss. Michigan starts off conference play relatively lightly, with three home games where they're reasonable favorites (Penn State, Maryland, Nebraska), and two road toss-ups (Iowa and Illinois). Going 4-1 or better over that stretch will be very important before their game in Madison, where the'll be a significant underdog. I'll certainly be watching them against the Hawkeyes on Saturday, hoping they can get off on the right foot in conference play.

Mondays seem like a good day for me to post these projections, so I'll stick with that, except in cases of extreme circumstance. Be sure to keep following along on Twitter for smaller updates throughout the week as the mild chaos of January college basketball begins.

Friday, December 2, 2016

Bracketology Update: 12/2

As expected, the ACC picked up their first win in the ACC-Big Ten Challenge since 2008, taking 9 out of 14 games from the Big Ten, including the most disappointing result, a 73-70 Virginia Tech comeback win over Michigan, the ramifications of which will be discussed in this post. First, we look at the one seeds, which are exactly the same.

1 seeds: Still Villanova, Duke, Kentucky, and Kansas. The only notable thing that happened between these four teams was Duke's 9-point home win over Michigan State. The game was tied at halftime, but Duke's win probability never dipped below 75%. They did temporarily drop out of the #1 overall spot in Kenpom, being replaced by conference foe Virginia, but since then have reclaimed it and therefore can stay at as a 1 seed in my book. Meanwhile, let's check in with the conference after every team played a significant nonconference game this week:

Big Ten: It wasn't a particularly good week for the conference, but neither was it unexpected. The ACC is simply a better conference this year, and many Big Ten teams fell in close games to overmatched opponents. Let's take this opportunity to catch up with every team in the conference, since I usually omit a few teams that don't need to be talked about. Here's where they stand:

Wisconsin - 2 seed: Wisconsin won convincingly in a 17-point victory over Syracuse, who is now starting to look not quite as good as we originally thought. Regardless, the Badgers are still in very good shape at the top of the conference.

Indiana - 3 seed: Indiana got their second huge win of the season in a 76-67 home victory over North Carolina. They now have two top 10 wins and one 100+ loss, so they will be a difficult team to judge going forward. It appears #CHAOSTEAM is as alive in basketball as it is in football.

Purdue - 4 seed: The Boilermakers lost on the road to a very good Louisville team, which doesn't affect their outlook that much. However, they now sit at 0-2 against the top 50, so they will look towards a December 17th matchup against Notre Dame to find that elusive nonconference big win.

Ohio State - 5 seed: Similar to Purdue, the Buckeyes lost to heavy favorite Virginia but don't fall in seeding. They actually had the lead for much of the game, and improved their Kenpom ranking despite the loss.

Michigan State - 5 seed: As mentioned earlier, the Spartans lost to Duke, and with that, the miserable Michigan State athletics year 2016 is just about over. At 4-4 overall and 1-4 vs. the top 50, they look to finally regroup these next few weeks against much weaker competition (although they will have to do so without superstar freshman Miles Bridges).

Michigan - 6 seed: Much like the 2K Classic victory was not cause for overexcitement, the loss to Virginia Tech is not cause for doom and misery. It does, however, cause me to reset seed expectations just a bit, but not to lower than they were at the beginning of the season.

Maryland - 9 seed: After declaring them "mysterious" last week, I figured I should check out Maryland in their game against Pittsburgh. It went...not well. The Terrapins couldn't get much of anything anywhere on the court, losing by 14, and now need a win over Oklahoma State this week otherwise things start to look dicey.

Northwestern - OUT (barely): They beat Wake Forest, and The Dream is still alive and well! They were expected to win that game, so it doesn't move them into the tournament just yet, but it certainly doesn't hurt their chances.

Iowa - OUT: Still losing: this week lost to Notre Dame in a game that wasn't particularly competitive. This week they face Iowa State and things continue to look grim.

Minnesota - OUT: Despite losing to Florida State, the Golden Gophers have actually been steadily climbing the Kenpom rankings throughout the season and currently have the same projected final record as Maryland. I'm still skeptical, but I'll certainly be keeping an eye out. This week they play Vanderbilt in Sioux Falls, South Dakota, for some reason.

Illinois - OUT: A home win against North Carolina State is nice, but the Illini are still going to end up with way too many losses to be in tournament consideration. If they can somehow win this week against VCU in Miami, maybe we'll start talking.

Nebraska - OUT: Lost to Clemson this week, now have lost three in a row, and in-state rival Creighton comes to town this week, so things still are not looking too good for the Cornhuskers.

Penn State - OUT: One of the five Big Ten wins actually came from the second-worst team in the conference when Penn State beat Georgia Tech, a team even worse than them. Congratulations, but you're still not coming close to making the tournament.

Rutgers - OUT: They are very bad. Not as horrifically bad as they were last year, but awful nonetheless. For some reason they had to play Miami (FL) in the ACC-Big Ten Challenge, and it went about as poorly as you'd expect. Unless they pull some crazy upset later in the season, this may very well be the last time I talk about Rutgers basketball this season. The end.

Michigan Update: The graphs didn't lie, they just picked a smaller bar than we would have liked. Virginia Tech is a bubble team, and losing at home to them doesn't look good, but I think in the end Michigan will end up with enough quality wins in the Big Ten to be just fine. Here's what the graphs have to say about that:
We finally have a bar in the nonconference graph that's above 50%, and that's at 10-3. If you remember my preseason post, that was originally the number of wins with the highest probability, at that time around 30% likelihood. Just like last week, the UCLA game appears to be a loss, and the rest appear to be wins. 11-7 is still the most likely Big Ten record, and the graph looks quite similar to how it looked two weeks ago, which is still better than it looked at the beginning of the season.

As for this week, Michigan plays Texas on Tuesday (your friendly neighborhood webslinger will unfortunately not be in attendance), and this game is pretty close to a must win if they want to stay off the bubble this year. Texas is the first team in the 50-100 range the Wolverines have faced, and it's very important to show the committee that you can take care of business against teams of that tier. Even after the VT loss, Kenpom still has Michigan as an 81% chance to beat the Longhorns. Let's hope they can bounce back, and when we come back next week, we'll be looking at their last tough nonconference game on the road against UCLA.