Alabama opened as a 7-point favorite over Ohio State, and on the morning of the game, the line is now 8.5. What does the GRT Statsy Preview Machine have to say about that?

Scoring Offense and Scoring Defense for both teams

Alabama’s Scoring Offense this year: 48.2
Ohio State’s Scoring Offense this year: 43.4
Alabama’s Scoring Defense this year: 19
Ohio State’s Scoring Defense this year: 22

From the perspective of Alabama

The Ohio State scoring defense of 22 is most similar to the following prior Alabama opponents:

• Texas A&M 21.7
• Georgia 20

And Alabama’s offense cares not one bit: The Tide scored 52 points against Texas A&M and 41 points against Georgia, an astounding 223% of what those teams did against it entire slate of competition this season. That makes the estimated points for Alabama against Ohio State 49.1. Sheesh.

But . . . the Ohio State scoring offense of 43.4 is most similar to the following prior Alabama opponents:

• Florida 39.8
• Mississippi 39.2

Florida got 46 against Alabama, and Mississippi got 48, so the defense can be vulnerable. Combined, that’s 119% of normal, and it puts the estimated points for Ohio State against Alabama at 51.7. Well.

Estimated score: Alabama 49.1, Ohio State 51.7

From the perspective of Ohio State

The Alabama scoring defense of 19 is most similar to the following prior Ohio State opponents:

• Clemson 20.2
• Indiana 20.3

Ohio State scored 49 points against Clemson and 42 against Indiana, which is an equally-astounding 225% of normal for those teams. It puts the estimated points for Ohio State against Alabama at 42.8.

The Alabama scoring offense of 48.2 is most similar to the following prior Ohio State opponents:

• Clemson 43.5
• Penn State 29.8

Ohio State allowed only 28 points to Clemson and 25 to Penn State, 72% of what those teams usually did this season. That Penn State comp is actually the second-best comp, but it’s really not even close. But because it’s the (second-)best we have, we’ll use it. Estimated points for Alabama against Ohio State: 34.7.

Estimated score: Ohio State 42.8, Alabama 34.7

SPM Final Estimates

Combining the results from both perspectives, here’s what we get:

SPM Final estimated score: Alabama 41.9, Ohio State 47.3

SPM Final estimated spread: Ohio State -5.4

Difference between the SPM and the Vegas opening spread: 12.4

It’s odd for this to happen, but according to the Machine, this is a Category 3 game, meaning it feels good about that prediction. But read on.

Eyeball adjustments

Yikes, that just feels wrong, doesn’t it? I don’t think I’ll use the word “shocked,” but I was really, really surprised by that. So I ran the thing again, this time with all games as comps, rather than just the two “best” comps. In fact, I didn’t stop there; I ran the thing for 3, 4, 5, 6, 7, 8, 9, 10, 11, and 12 comps. The Machine picked Alabama in every single one of those scenarios, most by right around the Vegas opening line of 7 points. I will note, for whatever it’s worth, that it also never got above the current line of 8.5.

So the question is, is there something wrong with those two “best” comps, or is it actually on to something?

The primary difference between those two comps and all of the others is that the Alabama defense got torched, giving up 48 points to Ole Miss early in the season but also giving up 46 points to Florida late. Nobody else got over 24 points the entire season. They held Notre Dame to only 14 in the most recent game. The Irish offense averages 33 points per game.

The question then is this: Is Ohio State’s offense like Florida’s (and Ole Miss’) or Notre Dame’s? Are they going to put up points at will, or are they going to find it much more difficult to put up points tonight? I don’t feel good about a prediction either way on that question. Maybe they just get their usual output?

Based on all of that, I’m going with Alabama 45, Ohio State 42.

Other predictions from other systems

As I said earlier, Vegas had Alabama as a 7-point favorite to open. With an over/under of 76, that’s a prediction of something like Alabama 42, Ohio State 35.

According to the SP+ rankings, Bill Connelly appears to like Alabama by 4.3.

Bottom line

The Statsy Preview Machine is waving the danger flag for Alabama by pointing to potent Ole Miss and Florida offenses actually being almost able to keep up with Alabama’s own potent offense. It likes Ohio State by 5 points. I have more confidence in Alabama’s defense than that, so I think the Tide win, but I will buy in to them not covering either the opening 7-point spread or the current 8.5-point spread.

What do y’all think?

In the midst of what feels like perpetual uncertainty, a surprise: the Vols picked up a transfer portal commitment from Virginia Tech quarterback Hendon Hooker. I live in the Blacksburg “metro”, and have spent nine of the last 15 years in southwest Virginia. I’m by no means an expert on the Hokies, just someone who watches them more closely and interacts more often with their fans than the average Tennessee fan. You can find good analysis on Hooker’s game out there – here’s a few thoughts on how he got there.

Virginia Tech’s stability at head coach under Frank Beamer manifested itself into great stability at quarterback. There’s a long line of multi-year starters in the post-Vick era: Bryan Randall, Sean Glennon, Tyrod Taylor, Logan Thomas, Michael Brewer. As Justin Fuente took over in 2016, however, the pattern breaks. The first year the Hokies went the juco route with Jerod Evans, who we saw in Bristol. They won their division and gave Clemson a scare in the title game, Fuente was ACC Coach of the Year, all was well. Evans actually set some VT passing records…and then declared for the draft, where he was not selected, and has yet to play a down in the NFL.

In 2017 the Hokies went with freshman Josh Jackson. They again finished the regular season 9-3 with losses to Clemson and a good Miami team, and fell to #14 Oklahoma State in the bowl game. All was well. Then Jackson broke his leg in the third game of the 2018 season, what ended up being a catastrophic 49-35 loss to Old Dominion, maybe the first sign of some real defensive trouble under Fuente. Jackson ultimately transferred to Maryland.

Hendon Hooker was a freshman in 2018, but when Jackson went down the job went to Kansas transfer Ryan Willis. The Hokies were game against Notre Dame in a loss, then suffered four straight defeats in which they gave up 49, 31, 52, and 38 points. Willis again had the job at the start of the 2019 season, which opened with a loss to Boston College, close wins over Old Dominion and Furman, and a 45-10 beatdown in Blacksburg at the hands of David Cutcliffe. At this point, as people were commenting on the heat of Jeremy Pruitt’s seat after Georgia State, BYU, and Florida early in year two, the sudden temperature of Fuente’s seat in year four made for a helpful comparison: Pruitt shouldn’t be in trouble after 15 games, this is what being in trouble actually looks like, etc.

Then Virginia Tech went to Hendon Hooker.

In his first start at Miami, the Hokies scored 42 points thanks to five turnovers and won by a touchdown. Hooker was 10-of-20 for 184 yards and three touchdowns, plus 76 yards on the ground. His numbers were similar in a win over Rhode Island, and got off to a good start against North Carolina before a familiar theme emerged: he got hurt in the first half, missing the rest of a six-overtime win over the Tar Heels. And he also missed Virginia Tech’s trip to South Bend the following week.

He returned to guide a three-game winning streak over Wake Forest, Georgia Tech, and Pittsburgh, never throwing for more than 260 yards but also never throwing an interception. His rushing totals became less productive – 7 for 10 against Georgia Tech, 20 for 27 against Pitt – but the Hokies were winning and their defense was playing well again.

In the last two games of 2019, the Hokies played wild affairs with Virginia and Kentucky. Hooker was 18-of-30 for 311 yards against the Cavs, but threw his first two interceptions of the year, the first on a hail mary at the end of the first half. The last one came with VT driving to take the lead in Virginia territory; he was also sacked three times in a row on the next drive. This is a theme you’ll hear on Hooker: if the defense knows you’re throwing it, he can struggle.

He was better on the ground against Kentucky in the bowl (12 for 50), but the Cats did a good job stopping him through the air (12-of-22 for 110).

Obviously, you take everything in 2020 with a grain of salt. Covid testing showed an unrelated medical issue that kept Hooker out of VT’s first two games of the year, wins over NC State and Duke. Oregon transfer Braxton Burmeister ran an efficient offense and stayed at the helm against North Carolina, where the Hokies lost 56-45 in a game both QBs played in. From there it was back to Hooker, who was sensational on the ground in a win over Boston College (18 for 164). He went for 98 more rushing yards against Wake Forest the next week, but threw three interceptions and the Hokies lost. The next week, however, he was 10-of-10 passing for 183 yards and ran for another 68 in a 42-35 win over Louisville.

Against Liberty, Hooker was again really good: 156 yards rushing, 20-of-27 for 217 through the air with three touchdowns and no interceptions. The Hokies lost a 38-35 shootout. Miami’s defense was better at taking all that away, beating Virginia Tech 25-24. And then the defense totally collapsed against Pittsburgh in a 47-14 defeat. Hooker again wasn’t bad statistically, but VT was never really in the game after halftime.

Then more weirdness: against Clemson, Hooker fumbled a snap on the opening drive and looked shaken, so Fuente went to Burmeister. He led a touchdown drive, so he stayed in the game. He was later removed for a scoop and score fumble, but the Hokies still trailed only 24-10 and had the ball in Clemson territory when Hooker returned. But he fumbled again, and appeared to have a medical issue on the sideline where he was shaking and “couldn’t get warm.” He did not play in the season finale against Virginia.

Had he thrown it enough to qualify, Hooker’s 2019 passer rating would’ve trailed only Trevor Lawrence among ACC quarterbacks. The Josh Dobbs comparisons I’ve seen are most fitting for pre-2016 Dobbs, the guy we wondered if they’d ever let him throw it deep down the field. That answer turned out to be an emphatic yes. Hooker’s dual threat skill set is something the 2020 Vols did not have on the roster, and he can certainly be efficient and mistake-free. It’s when defenses put him in third-and-long or he had to get them downfield in a hurry without his legs that Hooker struggled in particular. But there is no doubt potential and far more experience than anything the Vols will bring back next year.

After two weeks of right around 60% on all games, the GRT Statsy Preview Machine took a pie in the face this week, going only 7-12 (36.84%) overall, 1-5 (16.67%) in Category 2, and 0-4 ( let me see . . . 0%) in Category 3. For the season, the Machine is now 267-235 (53.19%) in Category 1, 100-83 (54.64%) in Category 2, and 54-43 (55.67%) in Category 3.

Because the Machine struck out on Category 3 games, it also struck out on both “Category 4” games. For the seven weeks we’ve been tracking Cat 4 games, they are 19-5 (79.16%).

The Machine had an identical result for all games using mid-week spreads: 7-12 (36.84%).

SP+, meanwhile, had a solid week, going 11-8 (58%) officially, using its own mid-week spreads. For the season, SP+ is 258-243-7 (52%). It had the same results against opening spreads (11-8, 57.89%) and is now 280-222 (55.78%) for the season against those spreads.

Below are the GRT Statsy Preview Machine’s picks for the 2020 college football bowl season. As always, if you’re wondering why we do this or what I mean when I refer to “confidence” and when I place game predictions into different categories, check out this post. Also, in case it’s not perfectly clear from the above results, spreads matter.

GRT SPM 2020 Bowl Picks

Ignore the Home and Away column headings again this week, as bowls are all neutral site games. I’ve accounted for them in the calculations, but haven’t denoted them in the table.

Bowl season features seven Category 3 games, those that the GRT Statsy Preview Machine likes particularly well. Of those, the following five make the Category 4 cut because they also agree with SP+:

• Iowa vs. Missouri (Iowa -13.5)
• Tulsa vs. Mississippi State (Tulsa -1.5)
• Marshall vs. Buffalo (Marshall +3.5)
• North Carolina State vs. Kentucky (North Carolina State +2.5)
• Texas A&M vs. North Carolina (North Carolina +5.5)

I’m really interested to see how the Statsy Preview Machine does in this weird season where there were very few, if any, data points for cross-conference play. I’m hoping that its focus on how a team does relative to what its opponents usually do does a good job of predicting how it will do against any opponent regardless of conference affiliation or schedule strength. We’ll see.

Well, a weird year gets weirder: the College Football Playoff field was announced yesterday, which means bowls start (checks calendar)…today. So jump in our free bowl pick ’em with confidence points at Fun Office Pools.

Today’s first game is at 2:30 PM ET, but as always we’ve made the earliest kick worth the fewest confidence points. So if you’re coming to this late, you’ll only be one point behind.

Good luck!

We’ve talked a lot about blue-chip recruiting ratio around here. The idea is that if a team wants to have legitimate national championship hopes, its recruiting class has to consist of at least 50% 5- and 4-star recruits.

That’s true, but it’s also true that football is a game of relativity, and as much as it matters that you hit certain standards, it matters more how you compare to your competition. As Will recently pointed out, a blue-chip ratio of 50% is probably not a high enough target for the Vols due to their schedule regularly featuring teams with even higher blue-chip ratios. When Alabama and Georgia are regularly hovering around 80%, the milestone of hitting 50% doesn’t matter nearly as much as it otherwise would.

The whole concept of “gap” in college football has been on my mind a lot lately. Everybody loves lists, and we have a tendency to look at them and draw conclusions based solely on rankings while completely ignoring the ratings that lead to those rankings. In doing so, we miss the gap. Being in second place doesn’t mean you’re in contention for first if you’re miles behind.

National Signing Day Part I on Wednesday got me thinking about recruiting gaps, and that led me to wonder just how much food was on the table and how much was being hoarded by the heavies in this league.

Using 247Sports’ data (including both signees and commits and current as of the morning of 12.17.20), there are 111 blue-chip recruits and 167 others. Over 81% of the blue-chips went to only six teams, less than half of the league.

That’s a huge disparity between the top six teams and the rest of the conference, but the disparity is also quite pronounced among those six. For instance, the sixth team on the list had only half as many blue-chips as the first. Alabama, whose kingdom extends from the castle campus to the recruiting trails, has a full 18% of all of the SEC blue-chip recruits.

Here’s a visual:

And here is the same information in another form:

One of Tennessee’s main challenges, of course, is not only that it is the sixth team on the list. And it’s not only that plus the fact that most of those other five teams have at least 50% more blue-chips than do the Vols. It’s both of those things, plus the fact that they have to play three (and sometimes four) of those teams every season. They could do well and never know it because so many of their regular opponents are doing even better.

All of that leads to a question about on-field strategy. If the coaching staff is committed to coaching techniques and game plans that rest on an assumption that you have more talent than your opponent, they’re essentially conceding at least three losses every year, at least until they can catch some of those other teams on the recruiting trail. The rub is that it’s hard to catch those other teams when they are also winning more often and turning coaches over less often than you are. And if you not only lose to more-talented teams but to less-talented ones as well, you have a real problem. You need wins to improve your blue-chip ratio, and you need a better blue-chip ratio to get more wins.

There’s both a chicken and egg shortage on Rocky Top. Tennessee doesn’t need both an elite blue-chip ratio and all the wins our hearts desire right now, but it does need to figure out how to bootstrap itself into one or the other quickly. Being sixth on a short list of luminaries is not an insignificant accomplishment, but in this case, it hides just how long the journey is to the desired destination.

What’s the solution? That, of course, is the million dollar question and will have to wait for another day and another post.

Tennessee opened as a 12.5-point underdog to Texas A&M this week, and the line has since moved to 14. With a gun to its head, the GRT Statsy Preview Machine is predicting that the Aggies will cover, but it’s also screaming that both teams are inconsistent enough to be wary of the whole thing.

Scoring Offense and Scoring Defense for both teams

From the perspective of Tennessee

Tennessee’s Scoring Offense this year: 22.4
Texas A&M’s Scoring Offense this year: 31.4
Tennessee’s Scoring Defense this year: 29.7
Texas A&M’s Scoring Defense this year: 22.1

The Texas A&M scoring defense of 22.1 is most similar to the following prior Tennessee opponents:

• Auburn 23.7
• Georgia 19.9

Tennessee scored 17 points against Auburn and 21 points against Georgia. Combined, that’s 87% of what those teams usually give up, making the estimated points for Tennessee against Texas A&M 19.2.

The Texas A&M scoring offense of 31.4 is most similar to the following prior Tennessee opponents:

• Georgia 33.2
• Missouri 26.1

Georgia scored 44 points against Tennessee, but Missouri got only 12. So, against the best comps, Tennessee allowed 94% of what those teams usually get. That makes the estimated points for Texas A&M against Tennessee 29.5.

Estimated score: Tennessee 19.2, Texas A&M 29.5

From the perspective of Texas A&M

The Tennessee scoring defense of 29.7 is most similar to the following prior Texas A&M opponents:

• Mississippi State 27.9
• Florida 26.3

Texas A&M scored 28 points against Mississippi State but 41 points against Florida, which, taken together, is 127% of what those teams usually give up. That makes the estimated points for Texas A&M against Tennessee 37.7.

The Tennessee scoring offense of 22.4 is most similar to the following prior Texas A&M opponents:

• South Carolina 23.5
• Arkansas 25.7

South Carolina scored only 3 points against Texas A&M. Arkansas got 31. Combined, that’s 69% of what those teams usually get, which makes the estimated points for Tennessee against Texas A&M 15.5.

Estimated score: Texas A&M 37.7, Tennessee 15.5

SPM Final Estimates

Combining the results from both perspectives, here’s what we get:

SPM Final estimated score: Tennessee 17.4, Texas A&M 33.6

SPM Final estimated spread: Texas A&M -16.2

Difference between the SPM and the Vegas opening spread: 3.7

That makes this a Category 1 game, meaning the Machine is not particularly self-confident.

Eyeball adjustments

I think it’s wise to be wary of the Machine’s results this week. For one, using Georgia and Missouri as scoring offense comps from Tennessee’s perspective is troubling because Georgia got a lot more than they should have and Missouri got a lot less. Plus, all four of the comps from A&M’s perspective have similar problems, especially South Carolina and Arkansas as scoring defense comps. This demonstrates quite nicely that both Tennessee and Texas A&M have been inconsistent this season.

So, I decided to run the Machine again using every game of the season as comps for both teams. The result? Tennessee 17.6, Texas A&M 31.7, a spread of -14.1.

So . . . what we’re dealing with here is a situation where both teams are inconsistent, but one is clearly still better over a larger sample size. That means that I’m going with Texas A&M 31, Tennessee 17 (A&M -14), but also hopeful for a wildly different result in Tennessee’s favor if the Vols play well and the Aggies play poorly.

Other predictions from other systems

With the Vols 12.5-point underdogs and an over/under of 51.5, Vegas is predicting a score of something like Texas A&M 32, Tennessee 19.

Bill Connelly’s SP+ likes Texas A&M by 12.5 (Texas A&M 33, Tennessee 20, Vanderbilt 17) and gives the Vols a 24% chance of winning.

Bottom line

The Statsy Preview Machine has done its best with the data it has, but I think it has overshot the mark a bit due to using only two comps each from inconsistent teams. However, we both think that A&M covers the opening spread. SP+ spit out the same prediction as the opening line and also has A&M against the spread, but only after rounding.

• Vegas: Texas A&M 32, Tennessee 19 (Texas A&M -12.5)
• SP+: Texas A&M 33, Tennessee 20 (Texas A&M -13, covers)
• GRT’s Statsy Preview Machine: Texas A&M 33.6, Tennessee 17.4 (Texas A&M -16.2, covers)
• Me: Texas A&M 31, Tennessee 17 (Texas A&M -14, covers) (but on upset alert!)

Season results on Tennessee games

• Week 4: The SPM and I were both right (albeit overconfident) in saying the Vols would cover against South Carolina.
• Week 5: The SPM and I were both wrong in thinking the Vols would not cover and would win by only 6 points. They won by 23.
• Week 6: The SPM was right, but I was wrong. The SPM predicted Georgia to cover a 14.5-point spread with a score of Georgia 29, Tennessee 12, while my eyeball adjustment revised that to say the Bulldogs would not cover with a score of Georgia 24, Tennessee 20. Georgia covered with a score of Georgia 44, Tennessee 21.
• Week 7: Both the SPM and I were wrong in thinking the Vols would cover 6.5 points as a favorite over the Kentucky Wildcats. The Vols lost 34-7. Oof.
• Week 8: Both the SPM and I were right in predicting that Alabama would cover the 20-point spread, the machine calling it 47-22 and me calling it 47-24. Alabama won 48-17.
• Week 10: The SPM was right and I was wrong about the Arkansas game. The Machine predicted the Vols would not cover a 1-point spread against Arkansas, and I predicted a 1-point win. Arkansas won 24-13.
• Week 12: Both the SPM and I were right in predicting that Auburn would cover the 11.5-point spread. Auburn won by 13, 30-17.
• Week 14: Both the SPM and I were wrong in predicting that Florida would cover the 17-point spread. Florida won by 12.
• Week 15: The SPM and I were both wrong in predicting that Tennessee wouldn’t cover as 15-point favorites over Vanderbilt. The Vols won 42-17.

Season-long results for all games can be found here.

What do y’all think?