College basketball is back! And so is T-Rank: my effort to objectively rate all 351 college basketball teams, and project records for every single one of them. Indeed, T-Rank has already made predictions for all 5,000 or so college basketball games on the schedule.
As I explained in last year's preseason post:
The upshot is that there is a place for more data-driven, formulaic preseason ratings. The AP preseason poll is good because human beings are very good at the kind of fuzzy logic required to take in a bunch of disparate data and form a judgment, so it is worth paying attention to. But computer ratings like Pomeroy's, Dan Hanner's and Team Rankings -- each of which use different models -- can provide different perspectives, and maybe some surprising insights in the process.
On a "the more the merrier" principle, I created my own spreadsheet-driven preseason ratings, which I call the T-Rank (T is for "terrific," of course). The methodology takes a variety of data (stats for every returning D-1 player, team returning minutes, incoming recruits, transfers in and out, weighted historical program performance, program momentum and loss of exceptional players) and spits out KenPom-style efficiency projections and a pythagorean winning percentage for every team. T-Rank is undoubtedly inferior to Pomeroy's and Hanner's preseason ratings, but the important thing is that it's a lot of fun -- and it gives me plenty of fodder for writing preseason college basketball posts.
For more explanation on how the preseason T-Rank is created, you can see last year's post, or this more recent post on the nuts and bolts.
One important note: the version of the T-Rank I'm using to create these projections is actually a composite of my "pure" T-Rank, the Kenpom preseason ratings and Dan Hanner's preseason ratings published at SI. The pure T-Rank is available here, if you're interested.
For a look at the full 351-team rankings, click here. For this post, I'll focus on the top 25, and the Big Ten.
The T-Rank Top 10
Without further ado, let's take a look at the T-Rank's preseason top 10:
Most of the usual suspects -- Kansas, Kentucky, Duke, North Carolina, Arizona -- are unsurprisingly back in the fray this year. Maryland is the Big Ten's lone representative in the top 10, but as we'll see, the next tier is littered with Big Ten teams. (If you're wondering why Maryland is a little lower here than in the AP poll, etc., see my post "The Case Against Maryland.")
The T-Rank Top 25
Here is the rest of the top T-Rank 25:
As you can see, the Badgers check in at No. 19 -- right about where most of the human preseason polls have them. Michigan State (13), Indiana, (16), Michigan (18), and Purdue (25) also join the party. Two of the Badgers' non-conference opponents -- Oklahoma and Georgetown -- appear as well.
The Friday Facts -- Big Ten Projections
Speaking of the Big Ten, here's the full breakdown:
One interesting aspect of these projections is that although T-Rank projects Maryland as the best team, it projects Michigan State to win this conference. This is because Maryland has a murderous conference schedule, and Michigan State does not. The Big Ten's unbalanced schedule could be a decisive factor this year, in what figures to be a close battle among several high-quality teams. (I'll have a more detailed look at conference schedule strength in a future post.)
Overall, T-Rank projects seven or eight teams into the tournament -- with Ohio State and Iowa in the bubble zone -- and another three teams (Illinois, Northwestern and Minnesota) within striking distance.
Rounding out the bottom are Nebraska, Penn State and lowly, lowly Rutgers. Rutgers figures to be the worst high-major conference team, by far.
Note: these projections include a recalculation based on the late-breaking news that Northwestern's Vic Law will miss the season with a shoulder injury. This drops the Wildcats about 15 spots -- comfortably off the bubble, where they belong.
Week 1 Big Ten Games
Here's a chart of the top 15 games scheduled to played involving Big Ten teams in the first week of the season (Nov. 13 through Nov. 20). The games are sorted by Torvik Thrill Quotient, my formula to determine which games you, the average fan, will most enjoy.
|DATE||MATCHUP||T-RANK LINE||TTQ ▴|
|11-17||1 Kansas vs. 13 Michigan State||Kansas -4.9, 73-68 (66%)||87|
|11-20||19 Wisconsin vs. 23 Georgetown||Wisconsin -0.6, 66-65 (52%)||85|
|11-19||44 Iowa at 52 Marquette||Marquette -2.2, 68-65 (58%)||76|
|11-18||54 Illinois at 79 Providence||Providence -0.8, 68-67 (52%)||75|
|11-17||23 Georgetown at 9 Maryland||Maryland -7.2, 71-64 (74%)||75|
|11-20||29 Xavier at 18 Michigan||Michigan -5.8, 72-66 (70%)||74|
|11-19||68 Temple vs. 84 Minnesota||Temple -1.8, 70-68 (56%)||71|
|11-20||113 Penn State at 137 Duquesne||Duquesne -1.9, 73-71 (56%)||63|
|11-19||72 Creighton at 16 Indiana||Indiana -13.6, 83-69 (86%)||60|
|11-17||119 DePaul at 113 Penn State||Penn State -5.0, 72-67 (67%)||56|
|11-15||94 Vermont at 25 Purdue||Purdue -12.6, 73-61 (87%)||54|
|11-17||107 Nebraska at 6 Villanova||Villanova -18.0, 74-56 (94%)||53|
|11-15||105 North Dakota State at 54 Illinois||Illinois -9.1, 67-58 (81%)||51|
|11-20||114 Columbia at 62 Northwestern||Northwestern -8.3, 69-60 (78%)||48|
|11-13||142 North Florida at 54 Illinois||Illinois -13.1, 78-65 (86%)||48|
Michigan State leads the way with its Champions Classic matchup against Kansas, which figures to be a great early-season test for both teams. The Badgers' game against Georgetown in New York next Friday comes in as the second-most intriguing matchup, hopefully followed by a huge matchup with Duke on the following Sunday.
Iowa at Marquette presents an interesting early test for both teams -- and the middle of both conferences.
Finally, Maryland should get an early test from Georgetown. Hopefully Georgetown will beat the snot out of them and then come into their game against Wisconsin over-confident.