College Football Analysis by Dr. Bob
Note: The lines quoted are the consensus line at the time each game was released to my subscribers on my Best Bets release page. The lines have moved so make sure to pay attention to the line constraints at the end of each analysis for the current rating of each game.
I’ve released 3 Best Bets so far:
(231) ***Miami-Florida (-1.5) 3-Stars at -2.5 or less, 2-Stars at -3.
(251) **Boise State (+4) 2-Stars at +3 or more.
(221-222) *UNDER (60.5) Illinois-LA Tech 1-Star UNDER 59 or higher
(227) **Virginia Tech (+3) 2-Stars at +3 or more
i
*UNDER (60 ½) – Louisiana Tech (-6) 27 Illinois 24
Fri Dec-26-2014 at 10:00 AM Pacific Rotation: 221 Over/Under 60.5
Louisiana Tech’s games averaged 62.2 total points per game in regulation and Illinois’ games averaged 60.5 total points per game, which are both higher than the national average of 55 points per game. However, there is good value towards the Under in this game for a few different reasons, from Louisiana Tech’s variance in red zone scoring average to the significantly slower pace that Illinois is running their offense in the second half of the season with a run-oriented offensive approach replacing the pass-heavy attack of the first 6 games of the season. It all adds up to solid value on the Under.
The Illinois offense will be run by senior Reilly O’Toole, who was a backup for 3 ½ years before getting his chance to start when Wes Lunt went down with an injury in the middle of the season. Lunt is a better passer but he wasn’t nearly as effective when he came back from a month long absence in week 12 against Iowa and it was O’Toole that led the Illini to wins the following two weeks to secure a bowl bid and secure his spot as the starter for this game. O’Toole’s compensated passing numbers (5.3 yards per pass play against teams that would allow 5.5 yppp to an average QB) aren’t that much worse than Lunt’s compensated numbers (Lunt faced worst pass defenses) but O’Toole isn’t as careful with the football and has thrown 7 interceptions in just 167 pass attempts (4.2% compared to Lunt’s 3 interceptions on 233 passes, 1.3%). O’Toole has thrown 16 interceptions on 337 career pass attempts (4.7%), so his higher than normal interception rate this season (2.9% in the national average) is most likely not the result of variance and he’s certainly in danger of throwing multiple picks against a ball-hawking Louisiana Tech secondary that leads the nation in interceptions. O’Toole does add a running element to the position, as he ran for 355 yards on 62 runs this season (not including sacks, which I count as passing plays), including 147 yards on 21 runs in the win over Northwestern that earned the Illini a spot in a bowl game. O’Toole’s running should improve the overall rushing numbers despite injuries to two starting offensive linemen. The other affect of having O’Toole at quarterback is more runs and fewer passes, which has led to more average time of possession and fewer plays per minute for the Illini since week 7 due to the clock stopping less often. Illinois averaged 2.7 plays per minute of possession the first 6 games of the season when they were averaging 23.3 rushing plays and 44.0 passing plays per game and their games averaged a total of 148.6 plays from scrimmage (not including kneel downs and spikes). In 6 games from week 7 on, with mostly O’Toole at quarterback, the Illini averaged 2.4 plays per minute of possession while averaging 34.2 runs and 31.7 pass plays and those games totaled just 137.6 plays from scrimmage. The fewer number of plays expected in this game with O’Toole at quarterback is not factored into the total on this game, which is part of the reason we have line value on the Under. For the season the Illinois attack averaged 5.4 yards per play against teams that would allow 5.4 yppl to an average team and they rate the same with O’Toole at quarterback but with more projected turnovers.
The Louisiana Tech defense was consistently good this season, allowing 5.0 yards per play to a schedule of opposing offensive units that would combine to average 5.7 yppl against an average defensive team. The Bulldogs were particularly good defending the run (4.3 yards per rushing play allowed to teams that would combine to average 5.2 yprp), which matches up well with an Illinois attack that runs it more often than they throw it with O’Toole behind center. Louisiana Tech will be without suspended starters DL Aaron Brown, LB Terrell Pinson, and LB Tony Johnson but those 3 also missed the Bulldogs’ game against an explosive Marshall attack and the defense played their best game of the season (relative to the strength of the opposing offense) in holding the Thundering Herd to just 5.4 yards per play and 26 points. I didn’t think Brown or Johnson would be missed since neither registered very many impact plays but I thought Pinson’s absence would hurt the pass defense since he’s proven to be very good in coverage (3 interceptions and 11 total passes defended). However, the coaching staff started a 5th defensive back against Marshall and the pass defense was even better and the Bulldogs gave up just 4.3 yards per rushing play to one of the best running teams in the nation. So, I certainly have no reason to think the absence of the 3 suspended defenders will hurt the defense given how well that unit performed in the CUSA Championship game and there is actually reason to think it might make the Bulldogs’ stop unit even better – although I made no adjustment either way. Louisiana Tech’s defense has a 0.7 yards per play advantage over the Illinois offense and the math projects the Illini to gain just 337 yards at 4.9 yppl in this game with O’Toole projected to throw 1.45 interceptions against a Bulldogs’ defense that leads the nation with 25 interceptions in 13 games.
Louisiana Tech’s offense is led by QB Cody Sokol, who transferred from Iowa and had a solid season throwing the football. Sokol averaged 7.1 yards per pass play while facing teams that would combine to allow 6.9 yppp to an average quarterback. The rushing attack features Kenneth Dixon, who ran for 1236 yards at 5.2 ypr, but overall the Bulldogs were well below average running the ball this season, as they averaged only 4.8 yards per rushing play as a team despite facing opponents that would combine to allow 5.3 yprp to an average offensive team. Overall, Louisiana Tech’s 37.5 points per game is very misleading given that the offense averaged their 6.0 yards per play against teams that would combine to allow 6.0 yppl to an average FBS attack. The Bulldogs also inflated their scoring average with 76 points in one game against Rice and they had a red zone efficiency that is too high to maintain. The Bulldogs averaged 5.5 points per red zone opportunity, which is far outside the normal range. The best teams in the nation are usually around 5.2 points per RZ and the national average is about 4.8 points per RZ opportunity. Louisiana Tech projects to be at 4.9 points per RZ based on their overall offensive stats and the difference works out to 2.4 points per game of red zone variance. The Bulldogs also had 4 defensive touchdowns, which is more than average also, so their 37.5 points per game was randomly high, which is another reason we have some line value on the under. Louisiana Tech may not have been affected by the academic suspensions on their defense but the offense really struggled against Marshall (just 268 yards at 4.1 yards per play) without two suspended starting offensive linemen, Tre Carter and Mitchell Bell. Carter and Bell are two of the most experienced starters on the line (combined for 41 career starts) and Bell was named first team All-CUSA. An offensive line that had given up just 5 sacks on 282 pass plays (1.8%) in the final 8 regular season games allowed 2 sacks on just 22 pass plays (9.1%) in their CUSA Championship game against Marshall. The Herd also had 7 quarterback hurries in that game and Sokol was a horrible 7 for 20 passing for just 72 yards (59 yards with sacks included) while under constant duress. That’s 9 sacks and hurries in just 22 pass plays (41%) compared to 43 sacks and hurries in 431 pass plays (10%) in the other 12 games with Carter and Bell. The patchwork offensive line should be better than they were against Marshall with a few weeks to prepare for this game but I think it’s reasonable to assume that the offensive line will not be as good without their 1st team All-Conference tackle and their best guard.
The Illini don’t look too good defensively, as they allowed 33.9 points per game and 6.1 yards per play to a schedule of opponents that would combine to average 30.5 points and 5.8 yppl against an average defensive team. Illinois was actually pretty solid defensive early in the season and they were decent late in the season but they gave up an average of 8.4 yards per play in week 6 and 7 before playing better down the stretch. At 0.3 yppl worse than average the Illini defense matches up pretty evenly with a Louisiana Tech offense that is just average from a yards per play perspective this season and is likely to be a bit worse than average without their two best offensive linemen. The math model projects 379 yards at 5.7 yards per play for the Bulldogs, which is just barely better than the national average for yards per play and less than the national average of 396 total yards per game. It’s unlikely that Louisiana Tech will top 30 points even if they continue to average 5.5 yards per red zone possession.
Overall the math favors Louisiana Tech by 5 points, which is where this line opened, but the Illini apply to a very good 73-15-1 ATS bowl game situation that is more significant than a 50-23-2 ATS statistical match up indicator that applies to Louisiana Tech. I’ll lean with Illinois at +5 or more
The projected statistics in this game project just 49 total points but there should be a few more points than that given that Louisiana Tech is still likely to have a better red zone scoring efficiency than my model would project – although not as high of an average as they’ve had this season. The combination of 5 fewer projected plays than the season numbers would project and Louisiana Tech’s positive variance in points scored has supplied us with enough line value on the under to make a play. I should make this Under a 2-Star Best Bet but I’ve had bad luck with totals this season so I’ll play it conservatively and just make this a 1-Star play. I’ll go UNDER 59 points or higher in a 1-Star Best Bet (2-Stars at 61 or higher) and I’d consider the Under a Strong Opinion down to 58 points (and a lean under at less than that number).
**Virginia Tech (+3) 28 Cincinnati 24
Sat Dec-27-2014 at 10:00 AM Pacific Rotation: 227 Over/Under 51.0
Virginia Tech is no longer a national power but the Hokies still play great defense and taking the defensively superior underdog in bowl games is a good recipe for success. That’s especially the case if the good defensive team is otherwise mediocre, as is certainly the case with Virginia Tech. Bowl underdogs that have allowed less than 26.0 points per game are 40-21 ATS against a favorite that has allowed more than 26.0 points per game. You don’t get as much value with those teams if they have a really good record (just 5-6 with teams that lost 2 or fewer regular season games) and teams that are better than average offensively tend to get some respect from the odds makers and the public too. What you really want is a defensively superior underdog that is worse than average offensively, as that 40-21 ATS record is 16-4 ATS if the dog has lost 3 or more regular season games and averages 26 points or less per game. That’s Virginia Tech and Central Michigan this season and my math model likes both of those teams.
Virginia Tech’s shining moment this season was their 35-21 win at Ohio State in which they held the explosive Buckeyes’ attack to just 342 yards at 5.0 yards per play. I’ve heard plenty of so called experts dismiss that win with the explanation that Ohio State quarterback J.T. Barrett hadn’t yet gotten comfortable in the offense, which is why he had a bad game against the Hokies. I don’t see that being the case at all given that Barrett averaged 13.6 yards per pass play the week before playing Virginia Tech and he averaged 10.1 yppp the week after playing the Hokies - and then averaged 8.4 yppp or more in the 3 games after that. Virginia Tech’s good defensive effort against Ohio State was about Virginia Tech not about Ohio State’s quarterback not being in good form. The Hokies’ defense can make a lot of quarterbacks look like they’re in bad form, as that unit allowed just 4.8 yards per pass play for the season while facing a lot of good quarterbacks (the quarterbacks they faced would average 6.6 yppp against an average defensive team). The Hokies will face another good quarterback in this game, as Cincinnati Gunner Kiel averaged 7.9 yppp this season while facing teams that would combine to allow 6.4 yppp to an average quarterback. Kiel had mixed results against better than average defensive teams, as he had a great game against Ohio State but played poorly against Miami (until late in the game when the Canes had a big lead) and against Temple while having a decent game against Memphis. Overall, Kiel’s compensated numbers against good defenses were about the same as his overall rating and my math model projects 5.8 yards per pass play for Kiel in this game. That projection could even be a bit high given that the Hokies allowed 5.8 yppp or more just twice all season (to ECU and the option attack of Geogia Tech). Cincinnati should have decent success running the ball (the math projects 5.4 yards per rushing play for the Bearcats), as Cincy’s rushing attack went from bad to better than average once freshman Mike Boone started getting the bulk of the carries in week 9. Virginia Tech’s run defense was a bit worse than average for the season (5.4 yprp allowed to teams that would average 5.3 yprp) but that’s only because they couldn’t stop either Miami or Boston College when leading tackler LB Chase Williams was injured. The Hokies were 0.5 yprp worse than average in the 4 games that Williams missed but he returned in the finale against Virginia and registered 12 tackles in helping limit the Cavaliers to just 76 yards at 2.5 yprp. Virginia Tech is 0.3 yprp better than average with Williams but Cincinnati has a bit of an advantage running against that defense with Boone as the featured back. Overall, Cincinnati is projected to gain 373 yards at 5.6 yppl in this game.
Virginia Tech’s problem in recent years has been a stagnant offense and this year’s edition managed just 5.0 yards per play despite facing teams that would combine to allow 5.7 yppl to an average attack. That attack is a bit better with RB Marshawn Williams out for the season. Williams was horrible, averaging just 3.8 ypr, and J.C. Coleman has done a good job since taking over the position, as he averaged 104 yards at 5.6 ypr over the final 3 games. The rushing attack should work well against a sub-par Cincy defense that allowed 6.0 yards per rushing play this season (to teams that would average 5.2 yprp against an average defense). That number was skewed by the 363 yards at 11.7 yprp that the Bearcats allowed to Miami but the math model projects a solid 5.2 yprp for the Hokies in this game even after adjusting for that outlier. Cincy is also worse than average defending the pass (6.4 yppp allowed to quarterbacks that would combine to average 5.8 yppp) but Virginia Tech quarterback Michael Brewer is 1.1 yppp worse than average and is expected to average just 5.6 yppp in this game. Overall the Hokies are projected to gain 423 yards at 5.4 yppl against Cincinnati’s leaky defense that allowed 6.2 yppl this season (to teams that would average 5.6 yppl against an average stop unit).
While Cincinnati is expected to average slightly more yards per play (5.6 yppl to 5.4 yppl) the Hokies should run more plays, as their defense doesn’t allow many long drives. In fact, Virginia Tech’s defense allowed just 4.6 plays per drive, which is the lowest in the nation and the reason the Hokies were +8.3 in play differential. The extra plays that Virginia Tech is expected to have more than makes up for the difference in projected yards per play and my math model favors Virginia Tech by 3 points in this game (with a total of 53 ½ points). In addition to the line and the 16-4 ATS defensive dog angle the Bearcats apply to a negative 5-32 ATS bowl situation that is based on their 7 game win streak. I’ll take Virginia Tech in a 2-Star Best Bet at +3 points or more and I have no opinion on the total.
Strong Opinion – Duke (+7 ½) 29 Arizona State 31
Sat Dec-27-2014 at 11:00 AM Pacific Rotation: 229 Over/Under 65.5
I can’t imagine Arizona State being too excited about a minor bowl game when they were at one point ranked 7th in the college football playoff rankings before losing to Oregon State in week 12 and then blew their chance for the Pac-12 Championship game and a major bowl game by losing to rival Arizona in their final regular season game. I actually have some evidence that Arizona State may not be 100% invested in preparing for this game, as team’s from the power 5 conferences (plus Notre Dame and including the old Big 8) are just 9-34-3 ATS as favorites of more than 7 points in pre-New Year’s bowl games after losing their final regular season game (or conference championship game). These are generally teams that had higher goals than to be playing a much weaker team in a bowl such as the Sun Bowl. Arizona State was in the exact same situation last year and lost 23-37 as a 14 ½ point favorite against Texas Tech in the Holiday Bowl. Duke also benefited from this situation last year when they nearly upset Texas A&M as a double-digit dog before losing 48-52 to Johnny Manziel and the Aggies on New Year’s Eve.
There is also some line value in favor of Duke in this game, as the Sun Devils aren’t as good with Taylor Kelly at quarterback. Kelly had some good numbers early in the season against 3 horrible defensive teams and the Sun Devils’ attack got much better when Kelly injured his foot and Mike Bercovici took over the starting role for 3 games. Bercovici was great in those 3 starts, averaging 7.7 yards per pass play against 3 good defensive teams (UCLA, USC and Stanford) that would combine to allow just 5.0 yppp to an average quarterback. Overall Bercovici was 1.3 yppp better than average on his 194 pass plays this season, averaging 6.6 yppp (excluding the 46 yard Hail Mary pass against USC) against teams that would allow just 5.3 yppp to an average quarterback. Kelly, meanwhile, has averaged just 6.7 yppp while facing teams that would allow 6.3 yppp to an average quarterback and he was 0.4 yppp worse than average in 6 games against Pac 12 competition after returning from his injury. The rushing attack also struggled in those games, as teams realized that Kelly wasn’t much of a threat through the air and could focus more on defending the run. Arizona’s offense was 0.2 yards per play worse than average over those last 6 games with Kelly back as the starting quarterback (5.3 yppl against teams that would allow 5.5 yppl) and I rate the Sun Devils’ attack at just 0.1 yppl better than average using Kelly’s numbers for the entire season. Duke’s defense is 0.2 yppl worse than average (5.4 yppl allowed to teams that would average 5.2 yppl against an average team). Arizona State doesn’t have much of an edge with their offense and the Sun Devils are projected to gain 462 yards at 5.9 yards per play in this game.
Arizona State isn’t good defensively either, as the Sun Devils allowed 5.8 yppl this season to teams that would combine to average 5.8 yppl against an average team. Duke is a sub-par offensive team this season (5.5 yppl against teams that would allow 5.8 yppl) but their disadvantage is not that significant and the Blue Devils are projected to gain 396 yards at 5.3 yppl. Overall, Arizona State has a 0.6 yppl advantage but the Blue Devils have excellent special teams and turnovers are projected pretty evenly. The math model favors Arizona State by just 3 ½ points (and 60 total points) and the Sun Devils have been overrated ever since Kelly returned to the lineup. Kelly has started 9 games this season and the 4 games in which the Sun Devils covered with Kelly at quarterback were all games in which they were +2 or more in turnover margin. In other words, Arizona has needed some good fortune to cover the spread with Kelly at quarterback and they’ll probably need some turnover luck to win by more than 7 points in this game too. I’ll consider Duke a Strong Opinion at +7 points or more and I’ll consider the Under a Strong Opinion at 65 points or higher (a lean under at less than 65), as the combination of ASU’s offense being worse with Kelly at quarterback and the 11 non-offensive touchdowns in Arizona State games (which is very high) has the total higher than it should be.
***MIAMI-FLORIDA (-1 ½) 38 South Carolina 27
Sat Dec-27-2014 at 12:30 PM Pacific Rotation: 231 Over/Under 61.0
As most of you know I’ve been anti-South Carolina since before the season started (I had 2-Stars on Under 9.5 wins) and I started playing on Miami in week 7 and went 3-1 on my Best Bets on the Hurricanes, with the only loss being a bit unlucky (a 4 point loss as a 3 point dog to Florida State). What prompted those mid-season Best Bets (easy wins against Cincy, Virginia Tech and North Carolina prior to the tough loss to FSU) was extremely negative variance in 3rd down efficiency that kept the Hurricanes from looking as good as they actually were those first 6 games. Through week 6 Miami was averaging 6.6 yards per play and allowing just 4.9 yppl against a good schedule of FBS opponents (Louisville, Arkansas State, Nebraska, Duke, and Georgia Tech) that would combine to outgain an average FBS team by 0.5 yppl. However, Miami won only 2 of those 5 games against FBS opponents while being outscored by an average margin of 8.1 points, which made the Hurricanes appear to be a pretty mediocre team. I knew that was far from the truth, as they had actually been 11.6 points better than an average team from the line of scrimmage. The discrepancy between how good Miami actually was and how they appeared to be based on scoring margin was partially due to a -3 in turnovers but was mostly due to extremely negative 3rd down variance. Miami, despite being a very good offensive team overall, had converted on just 15 of 63 (24%) 3rd downs through the first 6 weeks of the season, which would be low even for the worst offensive teams in the nation, and was extremely low for an offense that was as good as Miami’s offense. That negative variance in 3rd down conversions through 6 games supplied us with the line value to tab Miami for Best Bets in their next 4 games. The Hurricanes beat Cincinnati, Virginia Tech, and North Carolina by an average of 24 points before blowing their 23-7 lead against Florida State by playing the second half not to lose rather than being aggressive and confident like they were in the early stages of that game. As you know, Miami lost a heartbreaker 26-30 despite outgaining the Seminoles by 73 yards and that defeat crushed the Hurricanes’ growing spirit. I passed on making Miami a Best Bet the next week against Virginia, as it was a huge letdown coming off the disappointing loss to the Seminoles, and Canes lost that game to the Cavs by 17 points and then lost by 12 points to Pitt in their finale. Perhaps Miami lost some of their enthusiasm after the deflating loss to FSU but those two seemingly bad losses weren’t actually that bad given that Miami outgained Virginia and Pitt by an average of 380 yard at 6.0 yppl to 354 yards at 5.8 yppl. What those losses did do was provide us with the line value on Miami for this game that was starting to fade after their mid-season surge.
Miami may appear to be an inconsistent team but the Hurricanes’ line of scrimmage rating, which is a rating based on a combination of total yards and yards per play for and against and adjusted for site and quality of opponent, was very consistent. Miami’s worst line of scrimmage (LOS) rating this season was +7.6 points in their opener against Louisville and their average LOS rating is +14.7 points while their variance in LOS ratings is among the lowest in the nation. Where Miami is inconsistent is turning their yardage advantage into a points advantage. Early in the season the problem was due to 3rd down conversions (24% through 6 games) but the Hurricanes have converted on 45% of their 3rd downs in their last 6 games, which is more in line with what would be expected from an offense as good as their offense is. The problem in their final two games was negative red zone variance, as Miami averaged just 2.0 points per red zone opportunity and allowed 5.8 points per red zone opportunity in their losses to Virginia and Pitt, which explains how the Canes could lose by an average of 14.5 points in those two games despite outplaying those teams from the line of scrimmage. Miami’s red zone efficiency prior to their last 2 games was 4.7 points per RZ on offense and 4.8 points per RZ allowed on defense, so the extreme red zone variance was just a two week thing that made Miami once again look much worse than they actually are (the last two games put Miami’s season red zone scoring to just 4.4 points per RZ on offense and 5.0 points per RZ on defense). I see no reason why Miami would have trouble on 3rd downs or with red zone offense or defense in this game given how good they are overall from the line of scrimmage.
How good is Miami from the line of scrimmage? Miami has averaged 436 yards and 6.9 yards per play on offense while facing a schedule of FBS opponents that would combine to allow 5.6 yppl to an average team. Miami’s defense has yielded just 358 yards at 5.2 yppl to teams that would combine to average 6.1 yppl against an average FBS defense. That’s an overall compensated yards per play differential of +2.2 yppl, which ranks 9th in the nation. South Carolina, meanwhile, has a compensated yards per play differential of just +0.4 yppl as the offensively strong Gamecocks (6.1 yppl against teams that would allow 5.1 yppl to an average attack) have struggled this season due to a horrible defense that is 0.6 yppl worse than average (6.5 yppl allowed to teams that would combine to average 5.9 yppl against an average team). South Carolina has run more offensive plays than their opponents because they tend to give up big plays (and thus fewer long drives) and overall the Gamecocks’ line of scrimmage rating is +6.8 points. South Carolina is expected to run 9 more plays than Miami in this game, as the Hurricanes should hit on a few big plays against a defense that tends to give them up, but overall the Hurricanes have been 8 points better from the line of scrimmage, 0.6 points better in special teams and 0.9 points better in projected turnovers. Add it all up and Miami would be favored to win this game by 9 ½ points with neutral variance. Of course, there is a reasonable chance that Miami’s negative differential in 3rd down conversions and red zone scoring averages is not due to variance. However, that chance is factored into my model’s projection of Miami having a 57.5% chance of covering based solely on the math model. That percentage is based on the historical performance of my model and the math model plays (games with a significant difference from the line) have been particularly good in bowl games when choosing a team that is favored by 4 or less or getting points (those are 35-18 ATS since 2004, the first year of my current math model). In addition to the math, the Hurricanes also apply to a 73-29-3 ATS bowl situation that plays on teams to rebound from an upset loss in their regular season finale and teams in that situation that have lost 3 or more games in a row are 9-1 ATS with the only spread loss by just ½ a point. In general, teams that lost their final 3 regular season games are 60% ATS in bowl games, including 21-8 ATS when not favored by 7 points or more and facing a team that has 3 or more losses on the season. Miami has had time to refocus themselves after their disappointing end to the season and the Hurricanes are a much better team than South Carolina, who they are likely to beat even if the Gamecocks also have a renewed enthusiasm for this game. There is another situation that plays on good defensive teams in bowls when facing a bad defensive team and that angle is 64-32-1 ATS, which also applies to Miami. I’ll take Miami in a 3-Star Best Bet at -2 ½ points or less, for 2-Stars at -3 and for 1-Star up to -4 points.
**Boise State (+4) 35 Arizona 29
Wed Dec-31-2014 at 01:00 PM Pacific Rotation: 251 Over/Under 69.5
Arizona is playing this game in their home state but the Wildcats don’t have any other advantages and I’m not sure the proximity to home will be that much of an advantage given how Boise State’s fans tend to travel well for bowl games. The Broncos and their fans are excited about being the non-power conference team to get invited to a major bowl and I expect Boise to win this game straight up against an overrated Arizona team that got stomped by Oregon in the Pac 12 Championship game. Arizona also beat Oregon in Eugene but that impressive win was one of just 2 really impressive games that the Wildcats played this season (the other being a 42-10 win at Utah). The rest of the year was pretty mediocre and the Cats’ 10-3 record is due in large part to their good fortune in close games. The Wildcats are 6-1 in games decided by 7 points or less and they are more like an 8-5 or 7-6 team than a 10-3 team. That mediocrity shows in their stats, as Arizona only outgained their opponents by 7 total yards per game and 5.8 yards per play to 5.7 yppl. Arizona’s schedule was only 4.9 points tougher than an average FBS team and the Wildcats are only 5.4 points better than an average team from the line of scrimmage and only 7 points better than average overall (that includes special teams and projected turnovers). That rating is a few points lower than their rating based purely on points because their point margin was influenced by some turnover luck (+3 in defensive touchdowns off turnovers) and a Hail Mary pass to beat Cal (which I don’t include in my stats because I consider Hail Mary passes random).
Boise State is better than Arizona on both sides of the ball, especially offensively. The Broncos averaged 501 total yards and 6.7 yards per play this season while facing a schedule of teams that would combine to allow 5.8 yppl to an average attack. The Broncos are well balanced with Jay Ajayi (1689 yards and 25 rushing touchdowns) leading a ground attack that averages 237 yards and 5.8 yards per rushing play (against teams that would allow 5.1 yprp) while quarterback Grant Hedrick takes care of the aerial attack with 71% completions and 7.9 yards per pass play (against teams that would allow 6.7 yppp to an average quarterback). The passing game is even better now that WR Dallas Burroughs no longer gets snaps and sophomore Thomas Sperbeck is a featured receiver. Burroughs averaged just 4.3 yards on the 28 passes directed at him in the first half of the season and a mid-season injury to senior WR Matt Miller was a blessing in disguise, as it put Sperbeck in the lineup. Sperbeck leads the Broncos in receiving yards despite not playing the first 4 games of the season and his 12.0 yards per pass thrown to him also leads the team. Boise’s pass attack has been 0.2 yppp better since week 5 when Sperbeck starting playing and Boise State’s offense currently rates at 1.0 yppl better than average, which gives the Broncos a big advantage over an Arizona defense that is only 0.2 yppl better than average (5.7 yppl allowed to teams that would combine to average 5.9 yppl against an average team). My math model projects 513 yards at 6.4 yppl for Boise State in this game. Boise State has averaged 39.8 points per game on an average of 501 yards and they’re projected to score 35 points in this game.
Arizona’s offense tallies a lot of yards because they run a lot of plays but the Wildcats are just 0.1 yppl better than average, averaging a modest 5.8 yards per play while facing teams that would allow 5.7 yppl to an average team. I decided not to include the last two games since an injured foot hindered the mobility of quarterback Anu Solomon, which made him more prone to getting sacked (8 sacks the last two games) and kept him from running effectively. Solomon should be pretty close to 100% for this game but the Wildcats’ are still just 0.2 yppl better than average offensively if I exclude those final two games. Boise State’s defense was just 0.1 yppl better than average overall (5.4 yppl allowed to teams that would combine to average 5.5 yppl) but that average was skewed by the 627 yards at 9.7 yppl in their game against the New Mexico triple-option. My model adjusts for outliers and Boise’s defense rates at 0.3 yppl better than average after that adjustment. Boise also struggled defending the run against the Air Force option and while the Broncos were just barely better than average defending the run overall they were very good against the run when not facing a team that runs the triple-option. In those 11 games the Broncos allowed just 4.2 yards per rushing play to teams that would average 5.1 yprp against an average team. Boise allowed less than 4.0 yprp in 7 of 11 games against non-option teams so their run defense is actually very good despite their overall mediocre numbers that were skewed by their two games against option offenses. That math projects 414 yards at 5.5 yppl for Arizona in this game, which equates to about 29 points after factoring in the Wildcats’ 2.3 points advantage in special teams (projected turnovers are even).
Boise State has a projected advantage of 99 total yards at 0.9 yards per play and the Broncos appear to be the better team. I didn’t give any points to Arizona for playing in their home state since there isn’t any compelling evidence that playing in your home state (and not your actual home stadium) is an advantage. I’ll take Boise State in a 2-Star Best Bet at +3 points or more and for 1-Star as a dog of less than 3 points. The math also projects just 63 ½ total points and I’ll consider the Under a Strong Opinion at 69 points or higher.