clock menu more-arrow no yes

Filed under:

Arkansas vs. Mississippi State Advanced Stats Preview: Let's Try This Again

New, 4 comments

I'm running out of clever things to say here about playing top 10 teams.

Wesley Hitt

The Hogs are back in conference play and just in time to set an NCAA record for most consecutive top 10 conference opponents played. Now it's Mississippi State, who is actually ranked #1 in the nation on account of the fact that they beat (badly) two teams that Arkansas lost to.

The chatter on the message boards is that Mississippi State isn't your typical #1 team. Let's see if the numbers back that claim up.

Can the Hogs run the ball?

ARK off MSU def
S&P+ 16th 9th
Rushing S&P+ 13th 5th
Passing S&P+ 41st 15th
Pass Downs S&P+ 19th 7th
Drive Efficiency 12th 3rd
FEI 47th 12th

Ouch. Mississippi State's defense, a relative weakness, actually, has advantages everywhere, albeit small ones. MSU has only played one good rushing team (Auburn) but these numbers are opponent-adjusted, so folks aren't hitting their averages - or even coming close - against the Bulldog defensive front.

I've added a new stat ranking - Fremeau Efficiency Index (FEI) - at the bottom. FEI is a drive-based efficiency stat similar to S&P+. When combined with S&P+, it forms a stat called F/+, which is the overall qualitative ranking of a team used by stat guru Bill Connelly. Connelly uses team's F/+ ranking to make weekly picks in a computer. Some have asked me why I only post S&P+ (essentially half of the F/+ formula) and not FEI or F/+ itself. The short answer is that I do not believe FEI is a very good stat, and it therefore causes F/+ to be inaccurate. South Carolina (4-4) has the 3rd-best offense in college football, per FEI. Georgia Tech is #1.

For further proof, check out Connelly's F/+ picks. He has two formulas for picking games, "old" and "new." The "new" is below .500 against the spread this season, and the "old" is barely above .500 for the year. That's not particularly perceptive. Here's how his numbers have picked the Hogs this season:

Game "Old" Pick "New" Pick Line Bet Actual Correct?
at Auburn AUB by 20.6 AUB by 17.7 AUB -19 Split AUB by 24 Split
at Texas Tech TTU by 3.9 TTU by 4.1 TTU -3.5 Texas Tech ARK by 21 No
Northern Illinois ARK by 6.7 ARK by 8.3 ARK -14 NIU ARK by 38 No
vs. Texas A&M TAM by 13.9 TAM by 17 TAM -10 Texas A&M TAM by 7 No
Alabama ALA by 14.2 ALA by 12.8 ALA -9.5 Alabama ALA by 1 No
Georgia UGA by 18.1 UGA by 11.6 UGA -3.5 Georgia UGA by 13 Yes
UAB ARK by 11.5 ARK by 21.5 ARK -25 UAB ARK by 28 No
Record: 1-5-1

The "old" picks, the ones that are above .500, have picked against Arkansas in every game, going 2-5. The "new" picks, which are below .500, are 1-6. I didn't count his Nicholls State pick because it's FBS vs. FCS, but his projected score was 39-21, and Arkansas' chance of winning was only 86 percent, which is shockingly low for an SEC school against an FCS foe.

So if I'm going to post F/+ and draw significant analysis from it, I might as well ask a six-year-old, who could probably at least get close to .500 ATS. There's something about Arkansas that causes these F/+ picks to undervalue the Hogs pretty much every week. I'm not against the concept of stat-based predictions, but the current formulation is not worth using.

Can the Hogs stop anything Mississippi State does?

ARK def MSU off
S&P+ 24th 2nd
Rushing S&P+ 9th 7th
Passing S&P+ 28th 4th
Pass Downs S&P+ 20th 6th
Drive Efficiency 6th 5th
FEI 43rd 23rd

These figures don't point to a massacre, although again, Mississippi State has slight advantages pretty much everywhere. The Bulldogs have the dangerous combo of a quarterback that can throw (Dak Prescott) and a good running back (Josh Robinson), meaning that Arkansas is going to have to stop (or at least slow down) something without focusing on it.

As crazy as it sounds, Arkansas' best bet is probably to play nickel defense and have the linebackers - Brooks Ellis, whose return should be huge, and Martrell Spaight - focus on stopping the run, and force Prescott to win the Heisman through the air. The risk is, of course, that Prescott does just that, bombing for 400 yards and leading the Bulldogs to a fairly easy win. Arkansas' defense does have an ace up its sleeve, though.

Here's how opponents' #1 receiver have fared against the Hogs' coverage this season:

Team Name Receptions/Game Yards/Game Receptions vs ARK Yards vs ARK Yards Margin
Auburn Sammie Coates* 1 13
Texas Tech Bradley Marquez 5.8 72.7 4 34 -53.23%
NIU Da'Ron Brown 4.6 82.7 5 29 -64.93%
Texas A&M Malcome Kennedy 4.6 47.7 3 44 -7.76%
Alabama Amari Cooper 8.9 141.5 2 22 -84.45%
Georgia Malcolm Mitchell* 1 12
Average: -62.57%

I left out UAB because they had a backup quarterback playing, and I left an asterisk next to Sammie Coates and Malcolm Mitchell. Entering the season, Coates was expected to be Auburn's top receiver. He had just one catch for 13 yards, but that game became a breakout party for D'haquille Williams, who had 154 yards. Mitchell, on the other hand, has been hurt much of 2014 but was back for the Hogs.

Overall, you can see a trend. Robb Smith knows that his defense probably has less talent (and certainly less depth) than pretty much every offense the Hogs will face in SEC play. So rather than try to cover everybody evenly, which would lead to players getting burned all over, Smith's D shades coverage to a quarterback's favorite target and forces him to find someone else. That's why a quarterback like Blake Sims, who has a ton of natural talent, had a rough game against the Hogs: he couldn't come off of Amari Cooper, who was smothered by the Razorback D.

So if Arkansas with a healthy Ellis can slow down the run and take away Prescott's favorite target, how does Mississippi State's offense look?

Name Receptions Yards Touchdowns
De'Runnya Wilson 20 349 6
Jameon Lewis 15 223 1
Fred Brown 12 210 1
Robert Johnson 8 163 1

It's pretty clear that De'Runnya Wilson, who is 6-foot-5, will be the guy that Henre' Toliver shades in coverage with safety help to his side. However, Misssissippi State does spread the ball around pretty well. Robinson, the running back, is fifth in yards, and State has three other guys with 100 to 150 receiving yards this year.

Mississippi State's receivers aren't overly talented - they certainly aren't the caliber of Alabama's or Auburn's, and probably not even Texas A&M's - but the fact that they spread the ball around is either is either really good or really bad news for Arkansas. The good news is that if the Hog coverage can take Wilson out of the picture, it's possible that no one else will step up, since most of these guys are not used to being a primary target. Arkansas' second-best pass defender is Tevin Mitchel at nickelback, which puts him lined up with Jameon Lewis most often. Arkansas' primary defensive weaknesses in the passing game appear to be the field corner - usually Jared Collins - and the safeties down the seams, specifically Rohan Gaines. After lining up against guys like Williams from Auburn, De'Andrew White, Ricky Seals-Jones, and Chris Conley, Collins will probably be a little bit relieved to square off with Brown and Johnson - decent players, certainly, but not playmakers.

Of course, the bad news is that if Arkansas does in fact take Wilson away and slow down Lewis, Prescott may not be flustered at all, and some of these other guys may step up against coverage slanted away from them. That's what Georgia was able to do with Chris Conley. Regardless, Arkansas' going to make Prescott work.

Stat of the game: Line yards per carry

Line yards are the number of rushing yards a team has if all runs are capped at six yards. So a 14-yard gain is counted as 6. The reasoning is that the offensive line is really only responsible for the first six yards or so of a run. So if a team has runs of 0, 0, and 80, that's only 2.0 line yards per carry, since the back breaking off an 80-yard run isn't all due to the line.

Arkansas averaged 4.0 against UAB, which is really good, but UAB's run defense was really bad. Mississippi State's defensive strength is its front seven, but given that Arkansas has already faced Alabama and Georgia, this isn't anything new. Alabama's front is probably a little better than State's overall. Georgia's may actually be better as well. Nevertheless, finding room to run is going to be difficult. A good goal is at least 3.5 (Arkansas got 3.2 against Alabama, and it wasn't quite enough). Jonathan Williams gets more line yards per carry than does Alex Collins, especially in SEC, so Williams getting 25 touches may be necessary. If Collins isn't playing as well as Williams, then AC3 vulturing carries is not a good thing.

Keys to the game

1. Generate at least 3.5 line yards per carry. That's a fairly tall order against this front, but Arkansas did get 3.2 against Alabama, the most allowed by the Tide this season. The Hogs will also need to break off a couple of long runs, which is the main thing that DIDN'T happen against Alabama. But consistently moving the football keeps the offense out of obvious passing situations and keeps Prescott and Robinson on the sideline.

2. Post a success rate of 45 percent when passing from under center. I spent all of last week, between the Georgia review and the UAB preview, harping on Brandon Allen's passing from shotgun (really good) and passing from under center (not very good). I wrote that if Arkansas was going to change anything about the offense to put Allen in shotgun throwing on early downs, the UAB game would be the time to make the change. Well, it didn't happen. Outside of a well-executed two-minute drill late in the first half, Allen was only in shotgun on obvious passing downs, although Arkansas did run the ball a few times from shotgun, which is encouraging. The biggest problem has been pass protection, as Allen was been heavily pressured when trying to throw from under center in all four SEC games. If Arkansas isn't going to show more shotgun, then the Hogs are going to have to execute better when throwing from under center, and that starts with protection. Thankfully, Mississippi State's secondary is a relative weakness, so if Allen does have time, someone will get open.

3. Go +1 or better in the turnover margin. After turnovers didn't really hurt the Hogs in the first two SEC games (one each), Arkansas has gone -5 over its last two conference games. Mississippi State is +3 on the year, but even in conference play. Winning the turnover margin could be significant.

4. Make all field goals and extra points. The kicking game cost Arkansas two games (Texas A&M, Alabama) so if it's a close game, the Hogs can't waste a good effort only to miss a kick. Adam McFain's 49-yard field goal against UAB (and his perfect 8-of-8 on PATs this year) suggests this problem could be solved, but we'll see.

5. Hold Josh Robinson under 100 rushing yards. As improved as Arkansas' secondary is, it's not good enough to completely contain the Dak Attack. It could be good enough to slant coverage against De'Runnya Wilson and force Prescott to find other targets, where the Bulldogs may be plagued by the occasional drop or sack. Arkansas is 10th nationally in pass break-ups, so I'll take the Hogs' chances with slowing down drives where Prescott is having to throw. If all Prescott needs to do is hand the ball off to Josh Robinson, this game will be over in a hurry. Arkansas absolutely has to take away what comes easy to the Bulldogs.

Overall, this one's a long shot to win, especially on the road. It is a trap game for a Bulldogs team not used to having a target on its back, so there's that. But the reality is that Arkansas' defense probably isn't quite at the level necessary to stop a Bulldog team with lots of weapons. On the other side of the ball, Arkansas can and will move the football, but isn't built to attack State's weaknesses (deep passing), while State's impressive depth up front - a product of 16 of their 28 signees in 2010 now playing as redshirt seniors, compared to 4 of 25 at Arkansas (thanks Bobby and John L.) - won't be easy to wear down as the game goes on, meaning that Arkansas is going to have to throw way more than it wants to. So this isn't really a good matchup on either side of the ball.

The good news is that each of the last three opponents - LSU, Ole Miss, and Missouri - are much better matchups for the Hogs. This is certainly a winnable game, but not likely. That said, the breakthrough is coming, and Lord have mercy on the team Arkansas breaks through against.