For a few week span in late September and early October, Tennessee fans talked a lot about Butch Jones recruiting well but underperforming in big games. And, to no one's surprise, rivals were quick to chime in about how Tennessee was always overrated in preseason and never got it done on game day. Even non-rivals got in on the action, with bloggers from Michigan, Northwestern, Florida State, and who knows who else questioning Tennessee's hiring of assistants, late game strategies, and general in-game coaching.
After the Vols closed the year with five straight wins, those voices got a little quieter, but there remains a lingering perception that the Vols have been underachieving. But is it true? Has Tennessee underachieved under Butch Jones? Earlier this season, we took a look at how Jones' Tennessee teams have fared compared to expectations at kickoff. With the regular season in the books, it's time to take another look. In the last three years, how well have the Vols met expectations? Here's the data, organized by kickoff point spread.
|Point spread||# of games||Est. win %||Expected record||Actual record||Difference||Notable games|
|-23 to -50||4||98%||3.92-0.08||4-0||+0.08 wins||N/A|
|-17 to -22||3||93%||2.79-0.21||3-0||+0.21 wins||South Carolina (-17, 27-24 W)|
|-11 to -16||5||83%||4.15-0.85||5-0||+0.85 wins||Kentucky (-11.5, 50-16 W); Vanderbilt (-15, 24-17 W)|
|-8 to -10.5||1||73%||0.73-0.27||1-0||+0.27 wins||Kentucky (-10.5; 52-21 W)|
|-3 to -7.5||6||63%||3.78-2.22||4-2||+0.22 wins||Iowa (-3, 45-28 W); Missouri (-5, 29-21 L); Arkansas (-5.5, 24-20 L)|
|+2.5 to -2.5||5||50%||2.50-2.50||1-4||-1.50 wins||Vanderbilt (-2.5, 14-10 L); Florida (pk, 10-9 L); Oklahoma (+2.5, 31-24 L); Florida (-1, 28-27 L); Georgia (+2.5, 38-31 W)|
|+7.5 to +3||3||37%||1.11-1.89||2-1||+0.89 wins||South Carolina (+6, 45-42 W); South Carolina (+7, 23-21 W)|
|+8 to +10.5||0||27%||0.00-0.00||0-0||N/A||N/A|
|+11 to +16||5||17%||0.85-4.15||0-5||-0.85 wins||Missouri (+11, 31-3 L); Georgia (+13.5, 34-31 L); Alabama (+14.5, 19-14 L)|
|+17 to +22||3||7%||0.21-2.79||0-3||-0.21 wins||Georgia (+19, 35-32 L)|
|+23 to +50||2||2%||0.04-1.96||0-2||-0.04 wins|
What can we learn from this? First of all, let's add it all up: in Jones' three years at Tennessee, the Vols have underperformed relative to pregame expectations by a mere eight-hundredths of a game. That's well within the margin for error, so any accusations of global underachievement are unsupported by the numbers. Compared to pregame point spreads, Tennessee has won pretty much exactly as much as they were supposed to.
When you look at the games as organized by point spread, a pattern does begin to emerge. Jones' Tennessee teams have been entirely by-the-book in games that were expected to be one-sided, going 13-0 as a double-digit favorite (more than a game above expectations) and going 0-10 as a double-digit underdog (more than a game below expectations). Tennessee would be well-served to see the former continue, but the latter may be interesting only for its value in predicting other things. Perhaps an elite coach, given ten tries as a double-digit underdog, wins one or two games, and perhaps Jones' failure to do so suggests (defeasibly) he isn't an elite coach. But with the way he is replenishing Tennessee's talent level, the Vols should see fewer and fewer games as a double-digit underdog. With Alabama coming to Knoxville next year, there's a good chance Tennessee doesn't find themselves in this scenario again until at least 2017.
Beyond that, two things stick out. On the plus side, Tennessee has done better than usual as a modest underdog, scoring upset victories over South Carolina in 2013 and 2014 to take two out of their (admittedly tiny sample of) three games as an underdog of between 3 and 10 points. On the minus side, the Vols are just 1-4 in games with a line of under a field goal in either direction. All five such games have been one-score contests, but Tennessee has fallen short more often than they've come out on top. Five games remain a very small sample, but the eye test and the stats alike see major issues with the Vols protecting a late lead, be it on offense or defense, so Tennessee's mark in the perceived tossup games continues to warrant further consideration.
Of course, all of this data covers three years. What of the claims that Tennessee underachieved this year? Joel has already examined how Team 119 fared compared to our internal expectations, but we can also look back at how the Vols performed relative to outside expectations. Back in July, we took a look at Vegas' expectations for the Vols, and what we found was not the nine or ten-win hype that seems to be remembered. Certainly, there were people projecting Tennessee to go 9-3 or better, but their over/under on regular season wins was eight, and a look at expected win total based on preseason point spreads pegged the regular season in roughly the same place, projecting 8.07 wins and 3.93 losses. And falling seven-hundredths of a game short can hardly be considered underachievement. Again, it is well within the margin for error. Perhaps the voices expecting 9-3 or 10-2 were loud, but looking back at the most objective expectations we'll find, 9-3 or 10-2 can hardly be considered the consensus. The consensus was 8-4, and the result was 8-4.
Of course, regular season win projections don't take Tennessee in isolation. They are inherently reliant on projections of Tennessee's opponents as well. And Tennessee's opponents certainly did not all perform as expected. Most of them did, with seven of Tennessee's ten games against FBS-caliber competition (sorry, North Texas, but you don't count) closing with lines within a field goal or so of the offseason projections. The exceptions were Alabama, which moved from Tennessee +9.5 to +14.5, Missouri, which moved from Tennessee +2.5 to -6, and South Carolina, which moved from Tennessee -9 to -17.
So while most of the lines closed about where they were projected, the precipitous falls of Missouri and South Carolina did make Tennessee's schedule a bit easier than expected. If you rerun the numbers for 2015 alone, you'll find that Tennessee was expected to win about 8.4 games this season, after adjusting for changes in perception from the offseason to the regular season*. They won eight. So on the whole, it's fair to call this season a mild underachievement. But the emphasis must be on the "mild"--8-4 was the dominant preseason projection, and if you redid the preseason projections based on today's perceptions of UT's opponents, 8-4 would still be the dominant preseason projection. The Vols could've done better, but they're still as close as they can possibly be to where they were projected.
In Butch Jones three years here, Tennessee has been about where they were expected to be. They've done a bit better than expected as sizable favorites and modest underdogs and a bit worse than expected as sizable underdogs and in tossup games. But on the whole, they've been what we thought they were. Does this sanction any definitive claims about Jones' quality as a game coach? Not really. Oddsmakers don't ignore coaching when they set the lines, and perhaps Jones' critics will tell you that Tennessee's expectations were so low only because of his reputation as a poor game coach. But we do know that he's not taking a team that's supposed to be mediocre and winning big with them, and we do know that he's not taking a team that's supposed to be great and turning in mediocre results. After year three, everything is still on schedule. The tasks for years four and beyond? Keep scheduling steps forward, and keep staying on schedule.
*These do still use kickoff lines and not projections based on December 2015 point spread power rankings, so they don't necessarily take into account all the changes in perception between July and now. If you want to argue that Oklahoma would be favored by more than 2.5 if they came into Knoxville today, go right ahead. Just be sure to note that Georgia probably wouldn't be favored by 2.5 today either. As I see it, it evens out.