Advanced Stats Season Preview, Part 2: Fixing What’s Broke

Welcome back to the 2020 Advanced Stats Season Preview. Check out Part 1 here, and be sure to consult the glossary if you see a stat you’re not familiar with.

This article will dive deeper in the overall trends we’ve seen since Bobby Petrino’s first season in 2008, and then we’ll start to dig into the run game.

Trends to reverse

I have more data than I know what to do with, so I decided to graph every single play from scrimmage (FBS games only) for the Hogs since 2008, color-coded by coach:

Note: Hover your mouse over a point to see the play-by-play. This probably won’t work if you’re on mobile.

The main thing that stands out to me is that Bobby Petrino’s firing didn’t actually destroy Arkansas’ offense. Petrino led the Arkansas offense on a meteoric rise across 2010 and the beginning of 2011 before showing signs of a plateau. But after him, both John L. Smith and Bret Bielema left the offense better than they found it. The offense “peaked” during the 2017 New Mexico State game in Bielema’s final season.

Under Morris, the Hogs lost 160 EVA, wiping out ten years of offensive progress in just two seasons. That’s incredible. Petrino’s offense sank as low as minus-82 during his first season, but improved over the rest of 2008 and all of 2009, getting back to zero against Troy in November 2009. Over the next 20 games, from the 2009 Troy game to the 2011 Troy game, the offense gained 180 EVA. It took almost seven seasons to max out at +345, but Sam Pittman will inherit a team just 145 points better than the one Bobby Petrino found in 2008, having lost 200 points in fewer than three seasons.

Now let’s check out the defense:

It’s weird that the firing of Bobby Petrino seemed to impact the defense more than the offense, but it’s clear that the defense declined starting in 2012. Bielema’s defense got off a disastrous start in 2013 and early 2014, but the weird 2014 season with Robb Smith as DC and Trey Flowers, Martrell Spaight, and Darius Philon wreaking havoc caused a brief resurgence that didn’t end until early 2015.

Last year continues to look ridiculous on this graph even if you saw it all in person. Yikes.

Rakeem Boyd moves up the charts

One key part of the 2020 season will be running back Rakeem Boyd. He’s totaled over 1,700 yards in two seasons and has a chance to leave Fayetteville as one of the most prolific running backs in recent Hog history, which is saying something.

Take a quick guess, of all Hog running backs since 2008, who was the best?

Let’s find out:

Alex Collins leads all backs in carries since 2008, but the advanced stats don’t like him. He was less efficient than Jonathan Williams during the 2013 and 2014 seasons, and only had a strong year according to the EV+ system in 2015.

So who was the best? Knile Davis leads all backs in total EVA, with 21.6 points attributed to his skill. He was only moderately explosive and only decent at avoiding tackles for loss, but the real key to his success was when he carried the ball: Arkansas’ strong passing game in 2010 allowed Davis to carry on a lot of second-and-shorts, allowing him to rack up EVA with ease.

But on a per-play basis? I bet you didn’t guess David Williams, the South Carolina grad transfer who had a solid season in 2017. Amid all the quarterback drama and Bielema getting fired, Williams had a nice year, posting a 50% success rate, 2.9 line yards per rush (average rushing yards gained up to 5), and +0.12 EVA per rush, all easily the highest since 2008. He wasn’t explosive, but he never went backwards. He was also very good in short yardage and on third down, a rarity for Arkansas backs during this time period.

Note that Rakeem Boyd is middle of the pack in overall EVA, but he’s the leader in bonus yards per opportunity (average yards gained after 6 on runs of 6+ yards), making him easily the most explosive Hog back since 2008. He could stand to generate more big runs (just 32.6% opportunity rate, although opportunity rate is largely a function of offensive line skill), but when he gets a chance, he can collect yards in a hurry.

Speaking of offensive line, here’s how it’s looked in run blocking since 2008:

Remember that these are marginal (opponent-adjusted) numbers, so 0% is 6-6 level. Negative numbers mean that stat was below bowl-level, while positive numbers are winning record level. Pittman’s lines (2013-2015) where very good at avoiding stuffs (runs for a loss) and gaining line yards, but mediocre at generating big runs. Much of this may be a function of Bielema’s offensive scheme. You can see how big the dropoff was in 2016 when Pittman left. Kurt Anderson did make some progress with the line in his second season in 2017, but he was already struggling in recruiting and other things had broken down in the Bielema era.

A good target for 2020 is to get that stuff rate back from -7% to 0%, then try and get the line yards and opportunity rate numbers into the positives. We already know what Boyd can do with opportunities (10.0 bonus yards per opportunity), so the line needs to give him chances. If they can turn 2019’s -7%, 0%, 0% into something like 0%, 2%, 2%, then that will officially be an SEC-level line. For the first time in a while.

Up next

Our two-part lookback at a decade-plus of Razorback football is now complete. Part 3 will start to tackle the 2020 squad stat-by-stat as we prepare for the start of the Sam Pittman era.