Statistically Speaking - Cal at Utah

Cal has created havoc, but can they sustain it?

The Bears offense is a well-known juggernaut. They showcase the best quarterback in the country, a stable of big, talented wide receivers, and running backs who handle their assignments well in an offense that doesn’t ask them to carry the ball a whole lot. Utah’s defense against this offense will be a fascinating battle to watch, but it’s only half of the story. Cal has buttered its bread with an improved defense that excels at creating havoc plays; the Bears lead the nation in take-aways with 18, including 10 interceptions, good for 2nd nationally. They’ve also been successful at getting to the quarterback, racking up 3.6 sacks per game, 7th most in the country.

What’s particularly interesting about these excellent havoc numbers is that they don’t come along with indicators that the Bears are a very sound defense. They are giving up 23.4 points per game, and rank 83rd nationally in yards per pass attempt. The defense is also giving up a ton of long plays; they have surrendered 20 pass plays of 20 yards or more, 116th in the nation. Advanced metrics agree; Cal is ranked 93rd or worse in rushing defense, passing defense, success rate allowed, and explosiveness defense.

For as massive an impact as interceptions and fumbles have on the course and final score of a football game, they are poorly understood. There’s little evidence that particular players, coaches, or schemes are more successful than others at generating turnovers; there’s a strong correlation between passes broken up and INTs, indicating (duh) that players who are better at coverage are also better at getting picks, but that’s about it. Sacks are a different animal; we know how they work, why they happen, and can predict roughly which players, coaches, and schemes are better at generating them.

I broke down a number of team defensive metrics for 2014 and correlated them with interceptions, sacks, and passes defensed (interceptions plus PBUs). Very few team metrics looked to be related to havoc plays. The strongest was points per game; teams with many havoc plays tend to do very well at limiting the points opponents score, even though they don’t necessarily correlate with reducing explosiveness or success rate.

These tables give the correlations that were over .25*; I also checked for relationships with explosive pass plays, total explosive plays, yards per carry, red zone efficiency, and red zone touchdown efficiency, finding no significant relationship.

The upshot of this analysis is about what you would expect. Teams which create picks, PBUs, and sacks prevent explosive run plays, restrain yards per attempt, and limit points scored. They also tend to be better at controlling third downs, particularly teams which generate a lot of sacks.

These numbers are correlations, so there’s no way to know if one variable is causing the other or if something else I’m not measuring is causing both. However, it does indicate that what we’re seeing with the Cal defense is unusual. They haven’t been great at controlling explosive run plays, they’ve been terrible at preventing third down conversions, and they’ve been challenged through the air to the tune of 7.3 yards per attempt.

One way or another, the Bears appear due for a correction. Either they are going to stop getting so many picks and sacks, or their sound defense is going to start showing in their other stats as teams continue to struggle to move the ball without allowing negative havoc plays. Travis Wilson and his offensive line provide a challenge unlike those this Cal defense has faced before.

On average, Cal’s defense has faced opponents which have given up 13 sacks, which would tie for 102nd nationally. Utah’s offensive line has given up only once sack, third in the country. They also haven’t seen a quarterback like Travis Wilson before; while he has one interception this season, it was on a tipped hail mary pass as time expired in the first half of the Michigan game. Wilson has been making excellent decisions and accurate throws.

The weakness of Cal’s opening schedule has me leaning towards the sack numbers in particular being a mirage. I project Cal’s havoc numbers to drop across the rest of the season as they face better quarterbacks and offensive lines, and some of the protection that’s been given to their defense’s per-play and points numbers to go away. That decline starts on Saturday, and it may be a rude awakening for Cal fans.

Advanced Metrics Overview

The Utes had been muddling along in advanced metrics, up until pounding Oregon while Michigan continued its complete domination of every opponent to face them since (they’ve given up 14 points since the Ute game. Not average. Total.) The Utes’ offense is starting to look like it might be really, really good, and the Utes defense continue to impress as well, holding very good teams to very low scores. The Utes are in the middle of a massive upward adjustment from their preseason projections, and the only question is how far they’ll go.

The ascension of the Utah offense stands out; Michigan’s defensive excellence and Utah’s impressive performance against Oregon have created a team with an offense-first advanced profile. Cal’s defense is not favored by advanced metrics, and these numbers suggest they’ll struggle to stop the Utes. They are better than Oregon, though, so don’t expect 62. There aren’t a lot of advanced special teams metrics available yet, but the Utes have the edge there too, ranking 34th nationally compared to Cal’s 51st in ESPN’s efficiency metrics. If that seems low, given the Utes’ explosive return game, remember that Andy Phillip’s two missed field goals in the Michigan game are still sharply dragging down the kick efficiency numbers.

Vegas and the polls

The Utes were 6 point favorites to begin with, and most of the money has gone their way, driving the line up to 7.5 points. Massey’s aggregate ratings have been more accurate this year than advanced metrics, and have the Utes ranked 4th. Cal’s aggregate rating is dragged down by its soft opening schedule; the Bears are ranked 28th.

Stat-Head Pick

I’ve never been happier to be wrong about the Utes as I was watching them beat the tar out of Oregon. The offensive explosiveness caught me by surprise. In light of that game and the incredible performance of the Michigan defense since the Utes offense put up 17 on them with two missed field goals, and given the difference in performance when Wilson was lost to injury, I think it’s fair to say that, so long as Wilson remains healthy, the Utes are truly elite in 2015. Cal is a good team with a very good offense, but they’ve benefited from a doughy schedule and a bit of havoc play luck. They aren’t elite this year, and the Utes are about to prove it.

The Bears will struggle to stop the Utes offense, who will take advantage of their weakness with explosive plays through the air and a steady diet of Devontae Booker between the tackles. Meanwhile, the Utes defense, sound at every position, should contain a Bears offense that isn’t blowing doors off quite the way they were expected to. The more I look at this game, the bigger the scoring gap I expect to see.

Utah 37, Cal 24.

*A correlation coefficient ranges on a scale from -1 to 1, with 0 indicating no relationship between the variables and -1 or 1 indicating a completely dependent relationship. A negative number means that when one variable goes up, the other goes down (like miles run per day and BMI) whereas a positive number means that the numbers rise and fall together (like doughnuts eaten per day and BMI).