August 26, 2014

The Diff: Ohio State’s March Madness odds

As a reminder, The Diff is your weekly Wednesday look into the world of sports statistics. This week’s edition is already my 10th such post at WFNY. In last week’s edition, I shared some updates about the Cavs, Indians, NFL Draft and March Madness. Now, with the bracket set, I’m back again this week to discuss what it all means for the Ohio State Buckeyes.

The Diff

Some of you may know that I have a bit of a history with NCAA tournament projections. No, it doesn’t only go back to as far as my version of The Diff two weeks ago about NCAA tournament math. Nor does it merely go back to my days writing at my own blog called The Sports Report and featuring lots of college basketball content. In actuality, it goes back to as early as March 2007 and David Lee Morgan Jr.’s article about me in the Akron Beacon Journal. To make a long story short, however, for being a statistically minded person, I’m actually ended up being pretty awful at bracket predictions.

That obviously doesn’t inspire a ton of confidence in this post then, for you the reader. It wasn’t really meant to. I think I just meant it as a very large qualifier: For as efficient as these statistics might be, the tournament every year is still chaotic; for as good as Nate Silver or Ken Pomeroy might be, and thus I strive to be as well, crazy things can — and usually do — happen. No matter how much math you have at your disposal, this is still a game.

Bracket predictions are a strange, strange science. Personally, I kind of prefer the science of bracketology better — the practice of figuring out which teams will make the 68-slot field and why. I had tons of fun over the past few weeks dissecting the bubble race and my aggregate view turned out to be not so bad after all. Going back to this game prediction business, it’s just not necessarily my cup of tea.

Despite all that nay-saying, I think I can still add to the perspective of the Ohio State Buckeye fan this week with some words of wisdom about their odds moving forward. Along with showcasing the statistical odds from Nate Silver and Ken Pomeroy, of course, I’ll also share what the American populate at ESPN.com is predicting. Additionally, I’ve also got a massive database of the history of the 64-team-style tournament dating back to 1985. This helps to provide some historical data that can be useful for placing wacky predictions in a bit more of context.

————————————————————————————————————————————

South Region: 2 Ohio State Buckeyes vs. 15 Iona Gaels

This season marks Ohio State’s fourth consecutive as a top-2 seed in the NCAA tournament. In the 29-year history of the current bracket style, only four other programs have pulled off such a run: Duke (two 6-year runs, including currently and 15 of the last 17 years); Kansas (two 4-year runs, including currently); Kentucky (one 4-year run); and Memphis (one 4-year run).

Before I get too carried away with historical data, let’s just first get started with OSU’s odds against Iona, per the four outlets I’m using today:

# Team Pomeroy Silver America Historical
2 Ohio State 89.8% 94.1% 98.1% 94.6%
15 Iona 10.2% 5.9% 1.9% 5.4%

 

Normally speaking, I probably shouldn’t spend a whole lot of time on this matchup. The Buckeyes are huge favorites to win. It’s just worth noting the following: 15-seeds have won six of the 112 total matchups dating back to 1985′s bracket expansion. Of those six upsets, two occurred in 2012 (Norfolk State over Missouri and Lehigh over Duke).

Later on, I’ll share a bit more about why I think, despite the much hyped parity in 2013, this is unlikely to happen for Iona. It’s just incorrect historically to not mention this recent trend and give credence to the mathematical odds of an upset possibly occurring.

————————————————————————————————————————————

Third Round Game: 7 Notre Dame Fighting Irish or 10 Iowa State Cyclones

After likely dispatching the MAAC champions, the Buckeyes will next face of against Notre Dame or Iowa State on Sunday at UD Arena. TD had me pretty concerned from the moment the brackets were introduced about the possibility of a OSU-ISU matchup. Apparently, based on what he saw against his beloved Jayhawks, these Cyclones could be a scare for the Buckeyes. TD then wrote even more about the Buckeyes’ odds just on Monday.

So, using the same protocol as before, let’s take a look at who the Buckeyes are likely to face:

# Team Pomeroy Silver America Historical
7 Notre Dame 52.5% 60.0% 62.6% 59.8%
10 Iowa State 47.5% 40.0% 37.4% 40.2%

 

It figures to be a pretty close matchup, especially per Ken Pomeroy’s log5 odds. America loves Notre Dame a bit more than they probably should — of course — while Silver’s complicated odds are right on the money with the historical averages between 7-10 seeds.

Overall, I think it should be a fun preliminary game to watch. According to Pomeroy, these are two of the top 12 efficient offenses in the country. Notre Dame tends to pound it in the paint (led by offensive rebounding machine Jack Cooley), while Iowa State lives and dies on the three-point line (four different players have 127+ three-point attempts).

Alas, only one of these four teams then will advance to the Sweet 16 next weekend. So, given the likelihood of either Notre Dame or Iowa State winning, what are the odds OSU runs away with this grouping in Dayton?

# Team Pomeroy Silver America Historical
2 Ohio State 66.6% 73.2% 86.1% 64.3%
7 Notre Dame 16.7% 16.7% 9.0% 17.0%
10 Iowa State 14.2% 8.5% 4.1% 18.8%
15 Iona 2.5% 1.5% 0.8% 0.0%

 

Generally speaking, compared to the historical odds, this year’s mathematical odds are slightly more favorable for the Buckeyes. And lest we get carried away in the concept of them being an abnormally strong No. 2 seed, let’s not forget that Ohio State was likely to be closer to a 4 or 5 seed before the Big Ten tournament.

Another fascinating point: The closeness of Sweet 16 rates for #7 seeds and #10 seeds, historically. Taking this oddity a step further and removing the six instances where the 15-seed won in the first round (in case you didn’t know, they’ve never won two games), 2-seeds have a 73.8% winning percentage against 7-seeds and a 58.5% winning percentage against 10-seeds.

So it certainly should be an interesting and fun weekend for Buckeye fans. I wouldn’t necessarily expect a blowout either way — I think both teams are competitive enough, and have been all season, that it shouldn’t be a walk in the park. But for now, it’s still pretty statistically clear that OSU is the favorite from this group to advance and see the second weekend in Los Angeles.

————————————————————————————————————————————

Best of the West: 1 Gonzaga, 3 New Mexico, 4 Kansas State and 5 Wisconsin

Ever since Sunday night’s unveiling of the bracket, it’s been the hip thing to say that the West Region is one of the easiest in recent memory and that the Buckeyes are quite fortunate. Gonzaga is a weak No. 1, according to this line of thinking, while New Mexico and Kansas State also aren’t the most formidable of a 3-4 punch.

To a certain extent, yes, it’s historically rare for the top of a bracket to be composed in this way. This is only the 6th time in 29 years (thus, 116 regions) of the 64-team bracket style that two of the top-3 seeds in a given region were not from BCS conference.

But yet, for any “hip” line of thinking in 2013 culture, there’s also a counter movement. And in the case of the West Region, there’s the idea that maybe this actually is a sneaky difficult region to win. Here are two quick reasons why:

Point #1: I actually was reluctant to included 4 Kansas State above as “Best of the West.” In fact, according to the odds that I’ll get to later, they actually are the seventh-most likely team to win the region. And that’s not just because they’re not so great.

Point #2: Watch out for both 5 Wisconsin and 8 Pittsburgh. Both these teams are loved by the stats companies — Pomeroy has them ranked No. 9 and No. 7 in the country, respectively, in his efficiency rankings. And that’s overall, just to be clear.

Let it be known I didn’t even mention the strength of Gonzaga (No. 1 team in the country, No. 4 in Pomeroy standings and 14 straight wins) nor Ohio State (No. 7 team in the country, No. 5 in Pomeroy standings and 8 straight wins) in those two points.

Keeping up with the OSU theme, however, let’s move forward next to seeing the likely Sweet 16 partner on their side of the bracket. Here are those regional semifinal odds for the 3-14, 6-11 grouping:

# Team Pomeroy Silver America Historical
3 New Mexico 47.7% 49.7% 62.4% 50.9%
6 Arizona 33.4% 38.1% 31.1% 33.9%
11 Belmont 15.4% 9.7% 4.8% 13.4%
14 Harvard 3.6% 2.5% 1.7% 1.8%

 

America is showing its true homer-ism here with the love for New Mexico here. It’s kind of surprising, at first glance, to know that only about half of No. 3 seeds advance to the Sweet 16. Belmont is a sneaky sleeper, but Arizona isn’t too bad of a team either.

Now, continuing forward one more time, the winner of the 2-15, 7-10, 3-14, 6-11 half of the West region will then move on to face the other half’s winner in the Elite 8. So who are those likely teams?

# Team Pomeroy Silver America Historical
1 Gonzaga 42.5% 52.4% 49.9% 70.5%
4 Kansas State 5.8% 11.6% 12.9% 14.3%
5 Wisconsin 22.7% 19.0% 29.4% 7.1%
8 Pittsburgh 20.0% 10.1% 3.6% 6.3%
9 Wichita State 2.7% 3.5% 0.9% 0.9%
12 Ole Miss 3.9% 2.0% 2.7% 0.9%
13 Boise State/La Salle 2.4% 1.4% 0.2% 0.0%
16 Southern 0.0% 0.0% 0.4% 0.0%

 

Wow. This is all kind of neat to continue to absorb. Just slightly over 70% of No. 1 seeds have advanced to the Elite 8 dating back to 1985. Then, here in 2013, Wisconsin is receiving tons of love from both Ken Pomeroy and Nate Silver. As I mentioned above, Pittsburgh also should be receiving more love from America.

Overall though, it’s top-heavy. The Elite 8 matchup for Ohio State would be a fierce one — as I’ve now shown who the likely advancing teams are in the West Region and why it might be a bit better than you first expected.

————————————————————————————————————————————

Tournament’s Best Teams: Where does OSU rank in championship odds?

Stretching all the way through the Elite 8 with possible matchups, I’ve likely exhausted what I can do for analyzing Ohio State’s odds as much as I can on the Wednesday before the 64 teams even begin play. Thus, I’ve only got one more question to answer for now: What are OSU’s overall odds of cutting down the nets?

The storyline seems right: Besides maybe Louisville or Miami, there might not be a hotter team in the country. The Buckeyes have won eight straight, including six against tournament teams, and are a No. 2 seed again this season. If not now, when will they ever finally win it all in April?

Per an average of Ken Pomeroy’s log5 odds and Nate Silver’s statistical odds, here are the top 12 teams in the tournament sorted by their odds of making it to the Final 4:

Rk Re Sd Team Rd32 Swt16 Elite8 Final4 Finals Champ
1 E 1 Indiana 97.8% 82.9% 63.8% 50.7% 32.1% 17.9%
2 MW 1 Louisville 99.1% 82.7% 69.2% 50.3% 33.6% 21.5%
3 S 3 Florida 95.6% 80.1% 63.0% 42.6% 26.7% 16.9%
4 W 1 Gonzaga 97.2% 67.2% 47.5% 30.7% 15.1% 6.8%
5 S 1 Kansas 96.7% 75.4% 49.0% 24.9% 11.7% 5.9%
6 W 2 Ohio St. 92.0% 69.9% 46.1% 24.1% 12.2% 5.6%
7 MW 2 Duke 94.1% 66.8% 42.2% 18.4% 9.9% 5.1%
8 E 2 Miami FL 88.9% 61.1% 39.3% 14.1% 5.8% 2.0%
9 E 4 Syracuse 89.1% 55.9% 20.4% 13.0% 5.6% 2.0%
10 S 4 Michigan 88.6% 61.6% 28.6% 12.0% 5.1% 2.3%
11 W 5 Wisconsin 73.2% 49.3% 20.9% 11.2% 4.8% 1.9%
12 MW 3 Mich. St. 81.5% 56.2% 28.2% 10.9% 5.0%

2.1%

 

You can see my point again from earlier about Wisconsin being a sneaky team; they’re the only sub-4 seed in this listing. Also notable: The computers love, love, love Florida. Early on this season, the Gators were on pace for the most statistically dominant year in modern college basketball history. They then slid a bit toward the end of the regular season and notably only have one win over a tournament team away from Gainesville (Middle Tennessee, who no longer is in the tournament after their loss Tuesday night in the First Four).

The Buckeyes then rank No. 6 in Final Four odds, No. 5 in odds to make the finals game and No. 6 in overall championship odds. These all are purely mathematical.

According to America’s picks on ESPN, the Buckeyes are the clear West favorite, finishing at No. 3 in Final Four odds, then unlikely to seal the deal, placing No. 6 in the next two categories.

Overall, they’re clearly one of the strongest teams in the land. In closing, however, it’s tough to really consider this team one of the best in the land. They were quite mediocre for most of the season leading up to the big losses against Illinois and Wisconsin. They’ve been a different team since then — but does that mean they have enough for a long tournament run? We’ll see shortly.

————————————————————————————————————————————

Miscellaneous Tournament Thoughts

I can’t give up that easily. Without further transitions, here are two other random blurbs:

— Mid-major regular season champions: I teased about this earlier and I’ll expand on those thoughts now. In essence, I argued that although there are no dominant teams a la 2012 Kentucky in this year’s bracket, there’s unlikely to be an upset by a 16-seed or 15-seed this year. And that’s because several unexpected mid-major regular season champions fell in their conference tournaments.

In total, I counted up 10 automatic qualifiers for the National Invitational Tournament this season. This includes a few teams that were projected to be relatively decent seeds just two weeks ago, such as Louisiana Tech, Stephen F. Austin and Stony Brook. Instead of these more qualified teams, several underdog mid-major teams swept into the fray.

When you combine this movement, along with the ineligibility of a decent Connecticut squad, plus the new 68-team format, you’re going to have a relatively weaker field than the 64-team style of old. Overall, I think the low-end mid-majors were relatively weak this year, and especially of late. It will be up to the powers in the A-10, MWC and others to prove any worth for the smaller conferences.

— Akron being a 12-seed: I was surprised about this. I hadn’t seen a single bracketology projection with Akron as a No. 12 seed since Jerry Palm’s on March 8. That was right before Akron’s second loss in a week to a mediocre MAC squad in Kent State at home. After that game, brackets had the Zips consistently as a 13 or 14, but then UA received another boost with the losses of Stephen F. Austin and Louisiana Tech.

So, in the end, somehow with the necessary last-minute restructuring in the never-quite-so-clear bracket, Akron ended up as a No. 12 seed. And with the worst possible matchup imaginable: against former UA assistant Shaka Smart and the havoc-oriented VCU Rams. I would’ve much rather preferred Akron to be a No. 13 seed playing Michigan, a team that struggles with height. This one could get ugly Thursday night in Auburn Hills. Yuck.

  • Natedawg86

    so you’re saying there’s a chance…

  • mgbode

    math loves my bracket! all final-4 picks are listed in your 12 and 6/8 elite-8 picks (with the others getting positive mentions in the article). and here I thought I was being silly staying away from the 1 seeds (only 1 makes even makes the elite-8 and they lose there).

  • Ezzie Goldish

    You might enjoy this ranking system as well Jacob: http://www2.isye.gatech.edu/~jsokol/lrmc/

    I have used it for the tourney for a few years – won and lost on a tiebreaker my brother-in-law’s pool consecutive years. Interestingly Silver and Pomeroy’s overall brackets come in similar to theirs.

  • richard

    no point in even watching now

  • mgbode

    but they have Middle Tenn State #13 and they didn’t even make the 64?

    kidding, another data point is always good to have. I’m surprised that those rankings don’t seem to account much for the W/L vs. top50 teams despite listing them in the table.

  • Ezzie Goldish

    Yeah, I know, that loss made me ‘worried’. :) Though St. Mary’s was #19, so pretty tight…

    Yeah that’s interesting.

  • http://twitter.com/osuadamr Adam Rosen

    Nailed the Akron/VCU matchup… that was very ugly. And a New Mexico loss last night and a poor performance by the Zags shows just how weak this bracket is. I will disagree with you about the Buckeyes seeding before the Big Ten tournament. I think we were a 4 seed AT WORST if we lost the quarter final matchup. If we lost to Michigan State in the semis, we still had a strong chance at the 3. Winning it all and we got the 2, in a solid bracket. OSU, while very frustrating to watch at times this year, closed the season strong. LETS GO BUCKS!