This season, Free Analysis sides were 287-273-8 overall and 99-81-2 on games with a difference or 4+ points from the spread. Totals are just 284-277-1 but they’re a profitable 37-29 on differences of 6 points or more from the line.


The record on all Free Analysis sides is now 1626-1431-62 ATS since 2013, which is very good picking nearly every game over 5 years. The Free Analysis sides in which my predicted margin was 4 points or more away from the line are now 569-473-22. Totals on the Free pages are now 1347-1294-16 in the 4-plus seasons I’ve been tracking them.


The spreads posted are the consensus lines at the time I posted each analysis.

Bowl Best Bets were 8-3 and I am now 52-35-2 on my Best Bets (75-49-3 on a Star Basis) and 22-18 on my Strong Opinions this season.

College Bowl Games

Michigan St. vs

Mon, Dec 31 12:00 PM

Odds: Oregon -2, Total: 48

Game Analysis view matchup stats

Create an account to get analysis and updates sent to your inbox.

Strong Opinion – Under (48) – Oregon (-2)  21   Michigan State  19

Michigan State games averaged a total of just 37.8 points this season and the Spartans have gone under the total in each of their last 7 games with an average of just 28.6 total points. I see that under trend continuing even with Michigan State’s star cornerback Justin Layne skipping this game. The offense went from decent to worse for Michigan State when they started losing key receivers. Cody White missed four games starting in week 6 and star receiver Felton Davis III was injured early in week 8 after averaging 79 yards per game and 7.9 yards per target in the team’s first 6 games. The passing numbers really dropped off when White got hurt and I thought they’d improve when he came back in week 10, but White got a lot more attention from opposing defenses with Davis out of the lineup and the pass attack continued to struggle – regardless of whether it was starter Brian Lewerke or Rocky Lombardi throwing the passes (Lewerke’s compensated numbers since week 6 are the same as Lombardi’s compensated numbers). White averaged 11.5 yards per target when Davis was getting the double-teams those first 4 games but he averaged only 4.9 yards per target in the 4 games since his return to the lineup now that he’s the marked man. Michigan State’s top pass-catcher became #3 receiver Darrell Stewart, who averages just 4.8 YPT for the season. Michigan State’s receivers, without Davis, are simply terrible and Spartans have averaged just 4.0 yards per pass play (against teams that would allow 5.7 yppp to an average team) since Davis was injured, including just 3.3 yppp on Lewerke’s 77 pass plays – so don’t expect his return to improve the pass attack. I actually feel there was some negative variance involved in those horrible passing numbers (and some bad weather) and I have the Spartans’ pass attack rated at 1.2 yards per pass play worse than average rather than the -1.7 yppp they’ve actually been without Davis. Michigan State’s rushing attack is 0.6 yards per rushing play worse than average and Oregon defends the run well (0.5 yprp better than average) so the Spartans can’t rely on their rushing game to make up for their anemic aerial attack. Overall Oregon is 0.3 yards per play better than average defensively and I project just 4.3 yprp, 4.9 yppp and only 334 yards at 4.6 yards per play for the Spartans in this game.

Thankfully for Michigan State, their defense is good enough to keep them in this game. The Spartans allowed just 18 points per game and only 4.8 yards per play to teams that would combine to average 6.1 yppl against an average defensive unit. The defense got even better over the final 4 games after last year’s top cornerback Josiah Scott became healthy enough to play after missing the first 8 games. Scott had two interceptions and 5 total passes defended in just 4 games and he teamed with 2nd team All-Big 10 CB Justin Layne to completely shut down opposing wide receivers. Losing Layne to selfish behavior is unfortunate but Scott is just as good and he only played in 4 of Michigan State’s games, so the overall adjustment to the season stats for losing Layne isn’t that drastic (just 0.24 yppp higher based on my algorithm). Michigan State was 1.2 yppp better than average during the regular season and they’ll still be very good with Scott neutralizing Oregon’s star receiver Dillon Mitchell (1114 yards and 9.4 yards per target). Oregon’s other 3 wide receivers combined for just 776 yards and 6.4 yards per target so you only need one elite corner to defend the Ducks’ receivers. Michigan State also has the #2 rated run defense in the nation (behind Clemson), allowing just 3.7 yards per rushing play to teams that would combine for 5.5 yprp against an average team). Oregon’s offense isn’t nearly as good as their reputation, as Ducks are just 0.4 yards per play better than average (6.2 yppl against teams that would allow 5.8 yppl to an average team) and I project only 335 yards at 4.8 yppl for Oregon in this game even with the Ducks Justin Herbert projected to throw the ball more often than normal.

I expect both teams to struggle to move the ball in this game and playing on natural grass with some wind expected (13 mph with gusts into the mid-20s) should make it even tougher to consistently move the chains. Both teams played most of their games on artificial surfaces so playing on grass, with some wind, should lead to lower scoring than the math model would project without a surface/weather adjustment. There isn’t quite enough value to play the under as a Best Bet but I’ll consider the Under a Strong Opinion at 47 points or higher and I have no opinion on the side (the math leans with Oregon but Michigan State applies to a 46-11-1 ATS bowl situation that plays on elite defensive teams as underdogs).

Missouri @
Oklahoma St.

Mon, Dec 31 12:45 PM

Game Analysis view matchup stats

Create an account to get analysis and updates sent to your inbox.
Detailed analysis of every Bowl game is only available to subscribers to the College Bowl Package (details and updated pricing on the College Football page).

Northwestern @

Mon, Dec 31 4:00 PM

Game Analysis view matchup stats

Create an account to get analysis and updates sent to your inbox.
Detailed analysis of every Bowl game is only available to subscribers to the College Bowl Package (details and updated pricing on the College Football page).

No Carolina St. @
Texas A&M

Mon, Dec 31 4:30 PM

Game Analysis view matchup stats

Create an account to get analysis and updates sent to your inbox.
Detailed analysis of every Bowl game is only available to subscribers to the College Bowl Package (details and updated pricing on the College Football page).

Mississippi St. vs

Tue, Jan 1 9:00 AM

Odds: Iowa +7, Total: 40.5

Game Analysis view matchup stats

Create an account to get analysis and updates sent to your inbox.

Best Bet – *Iowa (+7 at -115)  17   Mississippi State  16

Best Bet Team Total – *Under (24) Mississippi State

Lean – Under (40.5)

The math model projects 35.7 total points scored but I expect this game to be lower scoring than that given how much worse, relatively, Mississippi State’s offense was against good defensive teams. The one-dimensional Bulldogs struggled against good defensive teams while racking up huge numbers against weaker defenses. In fact, Miss State averaged 7.6 yards per play and 41.2 points in 6 games against average or worse FBS defenses that would allow 6.1 yppl to an average team. Quarterback Nick Fitzgerald is a great runner but he’s an extremely inaccurate passer and he really struggled against better than average defenses, who could afford to single cover his receivers while putting a spy on him to limit his running. Mississippi State averaged only 4.2 yards per play and 7.8 points against 5 better than average defensive teams (Kentucky, Florida, Auburn, LSU, and Alabama) while scoring 7 points or fewer in 4 of those 5 games. Those teams would combine to allow 4.7 yppl to an average offense, so the Bulldogs were 0.5 yppl worse than average against better than average defensive teams while being 1.5 yppl better than average against mediocre or bad defensive teams. Overall, Miss State was 0.6 yppl better than average offensively but that rating was skewed by those huge offensive numbers against bad defensive teams and they’re not going to be 0.6 yppl better than average against Iowa’s strong defense. The Hawkeyes are 1.1 yppl better than average defensively (4.7 yppl allowed to teams that would average 5.8 yppl against an average defense) and their defensive rating is slightly better than the average rating of the 5 good defensive teams that Miss State averaged only 7.8 points against. The math model projects only 5.1 yards per play and 18.6 points for Mississippi State even if they really are 0.6 yppl better than average and a regression equation predicting the Bulldogs’ yards per play as a function of the opposing defensive rating (adjusted for site) would predict just 4.4 yppl for them in this game, which is think is very reasonable considering that Fitzgerald averaged just 3.4 yards per pass play against the 6 teams he faced that have a better than average pass defense. Those teams would allow 5.0 yppp to an average team, which is pretty close to the 5.2 yppp that Iowa’s defense would allow to an average quarterback on a neutral field. The regression equation for Fitzgerald’s passing predicts just 4.0 yppp against a team with Iowa’s pass defense rating. Iowa’s run defense is also elite, as the Hawkeyes rank 5th in the nation in compensated run defense (4.4 yprp allowed to teams that would average 5.7 yprp against an average defense). The Hawkeyes two most recent games were actually against teams with good running quarterbacks (Illinois and Nebraska) and they held those two elite running teams to just 4.8 yprp, which is 2.1 yprp lower than an average defense would hold those teams to. I project just 271 yards at 4.6 yppl for Mississippi State in this game.

Mississippi State has a dominating defense that allowed just 12.5 points per game and even held Alabama to 4.8 yards per play. Iowa’s offense rates as just average (5.5 yards per play against teams that would allow 5.5 yppl) and the Hawkeyes will be without one of their two star tight ends (Fant is sitting out). Fant’s absence actually isn’t that detrimental, as Fant averaged 8.1 yards per target while the rest of the tight ends and receivers combined to average 7.9 YPT, including 11.7 yards per target from the other two tight ends (758 yards on 65 targets). So, the adjustment for being without Fant is very minimal. Iowa’s not going to score much either, as they’re projected at just 263 yards at 4.06 yppl based on the straight math and I get 4.04 yppl using a regression equation for Iowa’s offense as a function of opposing defense. Iowa only faced one really good defense in Penn State and the Hawkeyes were held to just 4.1 yppl in that game but quarterback Nate Stanley injured his thumb on a helmet early in the game and could grip the ball for most of the game, which showed in his 18 for 49 passing.

The straight math only favored Mississippi State by 1.8 points with a total of 35.4 points, which assumed that the Bulldogs’ offense was relatively as good against good defenses as they were overall. There is plenty of evidence that is not the case given the high statistical significant of the regression equations and I now get a predicted score of 17-16 in favor of Iowa. I actually think there is a decent chance that Iowa scores fewer than 17 points given that Mississippi State allowed just 12.5 points per game but there is certainly value on Iowa plus the points and even more value on the Miss State Team Total under 24 points, which is a ridiculous number given the 7.8 points per game they scored against good defensive teams and Iowa’s 17.4 points per game allowed average. Mississippi State’s games against those 5 good defensive teams obviously all went well under the total, averaging just 26.4 TOTAL points in those games against Kentucky, Florida, Auburn, LSU, and Alabama with the average Over/Under line in those games being 51.4 points.

I’ll take Iowa in a 1-Star Best Bet at +6 points or more. I’ll also play 1-Star on Mississippi State’s Team Total Under 21.5 points or higher, And, I’ll lean with the Under in the game.

LSU vs

Tue, Jan 1 10:00 AM

Odds: UCF +7.5, Total: 58

Game Analysis view matchup stats

Create an account to get analysis and updates sent to your inbox.

Lean – Central Florida (+7.5)  27   Louisiana State  30

This game means a lot more to UCF than it does to LSU, as the Knights continue to have a chip on their shoulder and want to further validate their 25 game win streak with another upset of an SEC team after beating a highly ranked Auburn team 34-27 as a 10 point dog (they also beat Baylor by 10 points as a 16.5 point dog in their only other New Year’s day bowl game back in 2014). I would have picked UCF to win straight up if McKenzie Milton were healthy, but the math still leans with the Knights to keep this close with Darriel Mack Jr. at quarterback. The Knights become very run heavy with Mack behind center and they have a lot of speed that could give LSU some problems. Leading rusher Greg McCrae average 9.0 yards per run and Mack is a gifted runner as well (372 yards on 52 runs). Mack’s passing was bad in his first start back in week 8 against East Carolina (just 56 yards on 23 pass plays) but he averaged 12.3 yards per pass play with 70% completions in the AAC Championship game against Memphis as the Knights racked up 707 total yards at 8.0 yards per play. Obviously, Mack’s passing will fall somewhere in the middle and overall he averaged 7.2 yards per pass play on 66 pass plays (against teams that would allow 6.8 yppp to an average QB) in the 3 games where he was running the offense for more than just garbage time snaps. That’s a significant drop from Milton, but LSU will be playing without both starting cornerbacks, as Kristian Fulton is injured and All-American Greedy Williams is protection himself for the NFL draft. UCF has some very fast receivers running against those backup corners so it’s possible that Mack could put up big numbers if he can be accurate on down the field throws – although I project him at 5.1 yards per pass play. Overall, I have UCF gaining 405 yards at 5.3 yppl with Mack at quarterback, which should be enough to keep them in this game.

LSU’s offense is just mediocre (5.5 yppl against teams that would allow 5.4 yppl to an average team) and UCF is 0.2 yppl better than average defensively (5.5 yppl allowed to teams that would average 5.7 yppl against an average defense). LSU is projected to gain 412 yards at 5.6 yppl. The yardage is pretty close but LSU does have superior special teams and overall the math favors The Tigers by 4 points with a total of 56.5 points. However, I was conservative on my adjustment for both of LSU’s starting cornerbacks being out (just a 0.3 yppp adjustment) and I think UCF cares more about this game than LSU does. I’ll lean with the Golden Knights plus the points.

Kentucky vs
Penn St.

Tue, Jan 1 10:00 AM

Odds: Penn St. -6, Total: 47

Game Analysis view matchup stats

Create an account to get analysis and updates sent to your inbox.

Lean – Kentucky (+6)  22   Penn State  26

Kentucky was a surprising 9-3 in the regular season and the Wildcats played relatively better against better teams, going 4-2 straight up and 4-2 ATS as an underdog (with one of those spread losses being a 6 point overtime loss as a 5.5 point dog) while recording a 1-5 ATS mark when favored. Even without assuming that Kentucky will play relatively better as a dog my math has the Wildcats keeping this game close.

Kentucky has a slightly better than average offense (5.7 yppl against teams that would allow 5.5 yppl to an average team) and the Wildcats’ run-oriented attack has a bit of a match-up advantage against a Penn State defense that is much better defending the pass (1.7 yppp better than average) than they are defending the run (0.5 yprp better than average). The math model projects 229 rushing yards at 5.4 yprp and just 124 rushing yards at 4.4 yppp for the Wildcats in this game.

Penn State’s offense also depends on the run, as the Nittany Lions averaged 227 rushing yards at 6.1 yprp and just 192 pass yards at 6.2 yppp. Kentucky is also better against the pass (1.2 yppp better than average) than they are against the run (0.6 yprp better than average) and Penn State should also be able to run the ball (202 yards at 5.6 yprp projected) while average a modest 5.4 yppp for 153 passing yards.

Overall the total yards are pretty even, as Kentucky is projected to run more plays and the math favors Penn State by just 2.2 points with a total of 47 points after adding a bit for the good weather in Orlando. However, Penn State was very good at turning yards into points, as they averaged 5.8 points per redzone opportunity, which ranks among the best in the nation. There is some variance in that number but I’ll make a slight adjustment and call for the Lions to win by 4 points.

Washington vs
Ohio St.

Tue, Jan 1 2:00 PM

Odds: Ohio St. -6.5, Total: 57

Game Analysis view matchup stats

Create an account to get analysis and updates sent to your inbox.

Ohio State (-6.5)  32   Washington  25

Ohio State has looked very good in their last two games, beating Michigan 62-39 and beating Northwestern 45-24 in the Big 10 Championship game. However, people seem to be forgetting about the 1-6 ATS run in the 7 games prior to those two good performances. My math model does not. Actually, the math picks this game right at the number, favoring Ohio State by 6.3 points. I do expect a good effort from the Buckeyes in Urban Meyer’s final game as head coach and Meyer’s history when he’s had more than a week to prepare over his career (36-13-1 ATS) and his record against non-conference opponents (52-19-2 ATS) would have me leaning their way in this game if not for a 48-18 ATS bowl angle that plays on underdogs of 4 points or more that have an advantage in defensive scoring average of 5.5 points or more, as Washington does.

Washington is a balanced team that is good on offense (0.8 yards per play better than average) and even better on the defensive side of the ball (1.3 yppl better than average). Ohio State, meanwhile, has an elite offense that averages 6.9 yppl and rates at 1.6 yppl better than average with Dwayne Haskins at quarterback but is 0.1 yppl worse than average defensively (6.0 yppl allowed to teams that would average 5.8 yppl against an average defense). The biggest problem for Ohio State’s defense is big plays and they could be pretty good if they can fix that issue.

The math projects Ohio State with 450 yards at 6.0 yppl and Washington with 423 yards at 6.7 yppl. Ohio State is projected to have 11 more offensive plays, which is the case because their defense does have a lot of short possession, either from 3-and-outs or from giving up a big play and over the course of the season the Buckeyes have run 174 more plays than their opponents. The Buckeyes also have elite special teams units that add some points to their side of the ledger and overall I get Ohio State by 6.3 points and 61.1 total points. I’m going to shave a few points off of the total because this game applies to a 33-8 bowl game Under situation that plays against teams off 3 straight overs and a 33-11-1 Under angle that plays under in bowl games involving a team that averages over 40 points per game and is facing a good defensive team. I’ll pass on both the side and the total.

Texas vs

Tue, Jan 1 5:45 PM

Odds: Georgia -11.5, Total: 58.5

Game Analysis view matchup stats

Create an account to get analysis and updates sent to your inbox.

Strong Opinion – Georgia (-11.5)  37   Texas  18

This game comes down to motivation. Some are speculating that Georgia is disappointed to be playing in this game after feeling that they were one of the nation’s 4 best teams and should have been in the playoffs (both true) after barely losing to #1 Alabama in the SEC Championship game. Others feel that Georgia wants to send a message with a dominating performance to prove that they should have been a playoff team. It’s hard to know how Georgia’s players are feeling about this game or how focused they were in preparation. I do know that Georgia will destroy Texas if they prepared well for this game.

I looked up how teams that finished 5th and 6th in the final playoff ratings have done in bowl games and such teams favored by more than 3 points in their bowl game are 4-0 ATS and last season I had a winning Best Bet on Ohio State over USC after the Buckeyes finished 5th and felt they should have been a playoff team. In 2014 TCU went from 3rd in the penultimate rankings, won their final game, and ended up 6th in the final ratings. I remember the same speculation about TCU not being psyched for their bowl game because of the letdown of not making the playoffs. Instead, TCU won their bowl game 42-3 as a 3 ½ point favorite over a good Ole’ Miss team to prove that the committee was wrong not to include them. I’ll just assume Georgia plays at their normal level, which should be more than enough to win this game by two touchdowns or more.

Texas is not really that impressive from the line of scrimmage, as the Longhorns averaged just 5.7 yards per play and allowed 5.8 yppl this season while rating at 0.2 yppl better than average offensively with Sam Ehlinger at quarterback (he missed parts of two games) and 0.6 yppl better than average defensively (that 5.8 yppl allowed was against a schedule of teams that would average 6.4 yppl against an average defensive team). Georgia, meanwhile, outgained their opponents 7.3 yppl to 4.9 yppl and their average opponent has been 0.2 yppl better than Texas from the line of scrimmage. The Bulldogs are one of the most efficient offensive teams in the nation, rating at 2.2 yards per play better than average when quarterback Jake Fromm is in the game (I doubt that backup Justin Fields will play given that he’s rumored to be transferring), and the math model projects 460 yards at 7.25 yards per play for the Bulldogs in this game.

Georgia’s defense, rating at 1.4 yards per play better than average (4.9 yppl allowed to teams that would average 6.3 yppl against an average team) should have no trouble limiting the Longhorns’ mediocre offense. Texas is a worse than average running team (0.4 yprp worse than average) and Ehlinger is 0.2 yppp worse than the average quarterback that Georgia faced this season. The Bulldogs allowed just 4.9 yards per pass play to that good group of opposing passers and the only team that was able to average more than 5.9 yards per pass play against them was Alabama, who managed just 6.25 yppp despite having a historically good quarterback. The math model projects just 294 yards at 4.5 yards per play for Texas.

I’m well aware that Texas head coach Tom Herman has an excellent 12-2 ATS mark as an underdog but the Longhorns haven’t faced a team that is great on both sides of the ball. Texas only faced 4 better than average defensive teams this season (Maryland, USC, TCU, and Iowa State) and they averaged 5.3 yards per play against those teams, which is only 0.4 yppl better than the 4.9 yppl an average offensive team would have tallied against those teams and only 0.2 yppl better than their overall offensive rating. Texas had 3 games against elite offensive teams, facing West Virginia and Oklahoma twice, and the Longhorns allowed an average of 42 points and 8.0 yards per play in those 3 games, which is 0.5 yppl better than what an average defensive would have allowed and 0.1 yppl worse than their overall defensive rating. So, while Texas has been good as an underdog they really aren’t any better from a yardage perspective against better defenses and elite offensive teams.

In addition to their significant advantage from the line of scrimmage Georgia also has better special teams units and is less likely to turn the ball over. That math model predicts Georgia to win by 20 points, which isn’t that unrealistic considering that the Bulldogs’ average scoring margin is +18.6 points and Texas is 0.2 yppl worse than the average team that they faced. This game qualifies as a Best Bet based solely on the math but there is a chance that the Bulldogs aren’t as excited to be back in the Sugar Bowl as I hope they are so I’ll consider Georgia a Strong Opinion at -14 points or less.