Predictions are a lot easier to make in hindsight. Almost nobody got Nebraska’s record right, not even the .500-ish, computer-model projections that drew so much attention over the summer. Even those ended up being optimistic when the Huskers finished 4-8.
The crushing, all-out collapses are often hard to see coming, but the positive “surprise” seasons are often a little easier to predict.
I’ve been playing around with Pythagorean expectation since we launched Hail Varsity in 2012. It’s a formula originally created by Bill James for baseball that is based on the idea that scoring differential is a better indicator of quality than wins and losses. Based on that premise, you can get a sense of which teams were better or worse than their records may have suggested the previous season.
This is not a predictive measure. It’s more just part of the complete picture that has to include schedule, coaching changes, returning starters and all of that other summer stuff. But I like to look at it each year anyway. It does much better highlighting teams that could be poised for a breakout than those due for disappointment, and if you had looked at the Pythagorean expectations from 2016 . . . well . . . just look at the list of teams that were at least 1.5 wins below their projected win totals last season.
Pythagorean expectation had a good year.
1. Notre Dame (Wins below expectation in 2016: 2.74, 2017 wins: 9): The Irish’s four-win season last year wasn’t nearly as bad as it seemed, and Notre Dame has proven the point this year, winning nine games and remaining in playoff contention until last week’s loss to Stanford.
2. Auburn (2.49, 10): Bizarrely mocked after a completely reasonable loss at Clemson early in the year (and a less reasonable second-half collapse against LSU mid-season), the Tigers are headed to Atlanta this week to face Georgia for an SEC title and trip to the Playoff.
3. Iowa State (2.12, 7): So the Cyclones didn’t come out of nowhere? Pythagorean expectation said Iowa State was two games better than its 3-9 record last season, and October’s darlings (with wins over Oklahoma and TCU this year) have delivered on that secretly promising year one. And here’s the thing: The Cyclones scoring differential this season says they should have 8.4 wins, not the 7 they have, so there still might be more on the table for Iowa State in 2018.
4. Michigan (2.01, 8): Here’s one that didn’t work. The numbers said last year’s Michigan was a 12-win team that only actually won 10 games. Given all the Wolverines had to replace, you knew this typically solid indicator probably wouldn’t apply in 2017 and it hasn’t. Michigan finished the regular season 8-4, which in my mind is still pretty good given the circumstances.
5. Michigan State (2.00, 9): Hi, Mark Dantonio. The numbers knew last year was a little fluky. I should have paid more attention to that, plus your sterling coaching record, when I made my Big Ten picks over the summer.
6. Missouri (1.98, 7): Barry Odom’s Tigers started 1-5 and he had to beat back calls for his job. Since then Missouri has won six straight and he just got a contract extension.
7. UCF (1.69, 11): Ah, yes, Nebraskans’ second-favorite team at the moment. The numbers said last year’s Knights (6-7) were closer to an eight-win team. This year? Well, you’re all well acquainted with UCF’s record.
8. LSU (1.61, 9): The Tigers got some preseason buzz, but it took them a while to deliver on it. Since losing to Troy, LSU edged out Florida, beat Auburn, lost to Alabama and won every other game comfortably to sit at 9-3, one win better than last year.
9. Utah State (1.58, 6): Few have probably noticed Utah State’s improvement, but the Aggies were a three-win team a year ago. This year they’re going bowling.
10. Miami (1.55, 10): The Hurricanes’ nine-win 2016 season was good, and, so far, the 2017 season has been even better. Miami faces Clemson this weekend for the ACC title and perhaps a Playoff spot.
11. Fresno State (1.52, 9): The Bulldogs bottomed out last year, winning one game so the fact that the numbers said they were more like a 2.5-win team didn’t matter much to anyone outside of Fresno. But, it certainly didn’t hurt new coach Jeff Tedford when he arrived and had a squad that probably wasn’t quite as bad as it looked in 2016. He has Fresno State at 9-3 and playing Boise State for the Mountain West title.
Overall that group of 11 teams has increased its collective win total by 39 games over last season, and 10-of-11 have improved win totals individually (Michigan being outlier). That’s not uncommon production over the last three years. In 2016 11-of-11 teams that were 1.5 wins below projection the year before improved their win totals, and it was 10-of-12 in 2015. Since 2007, 86-of-111 teams 1.5 wins below projection have improved their win totals the following year (76.6 percent).
What spot are the Huskers in entering 2018? I’d call it starting from scratch. In 2015 Nebraska was 1.76 wins below projection at 6-7, so 2016’s turnaround wasn’t out of the blue. At the end of last year, however, the Huskers were 1.70 wins above projection at 9-4 (only four teams “overachieved” to a greater degree in 2016), indicating Nebraska was closer to a seven-win team.
As for this year Pythagorean expectation says Nebraska should’ve won 3.76 games, close enough to the actual four that, when combined with a coaching change, I’ll just throw it out when thinking about next season.
I won’t be doing that, however, with Baylor, Texas, Purdue or Duke in 2018, Power 5 teams that all look a little better right now than their records suggest.