Lindy's Five Essential Websites (Non-Major Media) for 2013
[+] Team Summaries

Friday, October 26, 2012

Week 9: An EPA Preview

Now that I'm able to more quickly crank out the EPA (Expected Points Added) rankings for QBs and RBs, I thought it'd be fun to use them in a way to preview five of the weekend's biggest or most interesting games.  So, here's my EPA-based preview of each:

Texas Tech at Kansas St (3:30 ET, Fox)

Quarterbacks: Seth Doege (93.89 EPA; .308 EP per play) vs. Collin Klein (90.03; .349 EPpp)
Running backs: Eric Stephens (-1.20 EPA; -0.015 EPpp) & Kenny Williams (-1.51 EPA; -0.016 EPpp) vs John Hubert (6.74 EPA; 0.054 EPpp)
If you want to see the two most productive quarterbacks in the country so far this year, watch this game.  Doege is #1 in our EPA rankings while Klein leads in the NEPA rankings, which slightly differs from EPA in that it accounts for future drives.  The running game, while taking a backseat to the Heisman contenders under center, shows an advantage for K-State.

Now, this is all information you can glean from the rankings I linked to.  However, I can also shed a little light on the two defenses, and hopefully come up with an idea of what to expect on Saturday.  I should note that I only have defensive statistics against FBS competition.

Texas Tech has allowed -32.30 EPA and -0.112 EPApp this year to quarterbacks.  They've already seen one Heisman contender, Geno Smith, and rendered him not only mortal but downright bad (-17.03).  It's not only Smith they've beaten up on.  Texas St's Shaun Rutherford (-8.2) and Iowa St's Steele Jantz (-16.55) also had their worst production of the season against Tommy Tuberville's Red Raider defense.  The high against TT this year?  Landry Jones' decent but still meager 6.13.  It could be a lean day for Collin Klein.

Primary running backs have managed 4.21 EPA (.035 EPApp) against Tech, which is slightly better, but could be explained by the fact Tech is usually playing with a lead, so they're a little more willing to surrender runs.  Only two backs have even garnered 20 touches - Oklahoma's Damien Williams (1.87 EPA) and West Virginia's Andrew Buie (-5.76).   Tech's defense appears to be for real.

The Wildcats have also been tough on opposing QBs this year, limiting them to -25.07 EPA and -0.102 EPApp this year, comparable numbers to Tech.  K-State hasn't given any QB their worst game of the year, but they're close to it for nearly the entire list of opponents.  Also, their defensive stats might be buoyed by having been allowed to face Kansas and Dayne Crist.  Tech hasn't had that luxury.  Against common opponents, Geno Smith was around 7 points better (less terrible would be more accurate) against Kansas St, while Landry Jones was 3 points better against Tech.

RBs have had a rough go of it against Kansas St, at -7.15 EPA (-0.048 EPApp) this year, and similar to Tech, only two opposing backs have managed 20 touches in a game this year - North Texas' Brandin Byrd (-0.14) and Kansas' James Sims (-1.54).

Like Scott, I give the edge to the Wildcats, for a couple of reasons.  First, the running game.  Tech by no means is a turnstile against the run, but running backs have had positive success this year.  Running backs have not had similar success against Kansas St.  Fitting into my reasoning, John Hubert has been more successful than Tech's tandem this season as well.  The second is that Doege has been rendered ineffective at times this year, against Iowa St and especially against Oklahoma.  Klein hasn't (which is why Klein is rightfully getting more Heisman attention right now).  It should still be worth watching, I think.  Tech's defense should be good enough to keep the game close.

Florida vs. Georgia (3:30, CBS)

Quarterbacks: Jeff Driskel (12.22 EPA; .062 EPApp) vs. Aaron Murray (49.62 EPA; .209 EPApp)
Running Backs: Mike Gillislee (-18.22 EPA; -.121 EPApp) & Matt Jones (-0.69 EPA; -.031 EPApp) vs. Todd Gurley (18.49 EPA; .161 EPApp) & Keith Marshall (10.12 EPA; .082 EPApp)

If given the above numbers without names and told that one team is a top 2 team while the other is #11, having skated by with some underwhelming performances and one blowout loss, and asked to pick which is which, you might say Georgia is the top 2 team.  But you'd be wrong.  Based on offensive EPA, Georgia seems like the cinch pick to win.  Let's look at the D's.

Florida has held opposing QBs to a whopping -76.45 EPA & -.252 EPApp.  They've allowed -23.64 EPA & -.137 EPApp rushing, so that's not any easier.  In fact, it'd be easier to make a list of individual players with above-average performances against Florida this year:  Texas A&M's Trey Williams had 1.93 EPA on 9 plays, and LSU's Spencer Ware had 0.09 on 10.  Two running backs, both barely above average on limited carries.  Now we're seeing why Florida is undefeated despite a very underwhelming offense.

UGA has allowed -16.24 EPA & -.057 EPApp passing and -16.28 EPA & -.082 EPApp rushing, so they're hardly a pushover, despite what some UGA fans would tell you after slow defensive starts against South Carolina and Kentucky.  The South Carolina concerns were legitimate - Connor Shaw gashed UGA's D for over 15 EPA - Shaw remains the only player to have managed 10+ EPA in a single game against either defense on the field Saturday.

Florida should have a clear advantage thanks to their monster defense, but two factors could keep this closer than Georgia fans, despondent since the trip to Columbia, might think:
1. "Gurshall" - UF's rush defense has been stellar, but have they really been tested?  A&M's duo of Michael and Williams have been somewhat effective, but only somewhat.  The same thing could be said of Tennessee's Raijon Neal and USC's Lattimore/Miles combo.  LSU's backs might be the best the Gators have seen, and their combined EPA for the year don't even match Keith Marshall - and he's Georgia's second best RB.
2. The UF offense - Despite as amazing a defense as college football has seen this year, take a look at the scores after 3 quarters for the following Florida games:

Florida 17, Bowling Green 14
Texas A&M 17, Florida 13
Florida 27, Tennessee 20
Florida 7, LSU 6
Florida 18, Vanderbilt 8

The Gators have typically pulled away in the 4th as better conditioning or more desperate opposing offense tends to work in their favor, but for a team currently slotted as a national title contender, they lack a certain dominance, and I think the offense is to thank for that.  Even one of their blowouts, the South Carolina game, is largely due to defense, with the Gators scoring 21 points in the first half on just 29 yards of total offense.  The offense has never looked like that of a top 5 team.  Gillislee has plenty of rushing yards, yes, but they're not particularly valuable yards aside from the effect they might have on the clock.  If UF plays inefficiently, as they're somewhat wont to do, Georgia's defense is good enough to take advantage.

Now, will these two factors even matter (or rather, factor)? I don't know.  Just because UF hasn't faced running backs of Gurley & Marshall's caliber doesn't mean they won't punish them the way they've punished the others that have tried them.  If UF's offense is inefficient, it doesn't mean it'll hold back the defense - it hasn't yet.  I think Florida wins, but nothing in Jacksonville would surprise me.

Ohio St at Penn St (5:30, ESPN)

Quarterbacks: Braxton Miller (46.38 EPA; .141 EPApp) vs. Matt McGloin (43.24 EPA; .146 EPApp)
Running Backs: Carlos Hyde (21.89 EPA; .190 EPApp) & Jordan Hall (0.12 EPA; .002 EPApp) vs. Bill Belton (5.30 EPA; .133 EPApp) & Zach Zwinak (1.42 EPA; .020 EPApp)

The only Big Ten game of the year that will have absolutely no effect on the Big Ten race might very well feature the two best teams in the conference.  One, Ohio St, is not exactly a surprise.  The other, Penn St, is a huge surprise, as most had left the Nittany Lions' chances for dead after their roster was picked apart by vulturing opponents.

The Buckeye defense has held passers to -30.37 EPA & -.078 EPApp this year while holding runners to -4.20 EPA & -.020 EPApp.  While that pass defense has remained pretty solid all year long, the run defense total is misleading.  Over the first 5 weeks of the season, the D allowed -22.37 EPA to opposing running backs.  Over the last 3 games that number has bloated to +18.17.  That's not good, especially when you consider the last 3 opponents included Indiana and Purdue (I understand that Akeem Hunt has been very good for Purdue this year, but Ohio St still should have had enough of a talent advantage to overcome his abilities).  So, that will be a concern for Saturday, as Belton is coming off a good game and Zwinak had been good prior to the last game.

Penn State's D has held opposing QBs to -28.18 EPA & -.086 EPApp this year; they've held rushers to -8.59 EPA & -.054 EPApp.  The run D has been pretty consistent.  Carlos Hyde will probably be the best RB they've faced yet, but Venric Mark is next on the EPA list and he managed just +2.59 EPA against the Lions, so it's unlikely that Hyde runs wild.  My guess is he will be above average.

I think this one comes down to Braxton Miller.  A good QB (Tyler Tettleton) has gashed Penn St already this season, and Miller is capable of doing it again.  Both QBs represent the best the other has faced, but I get the feeling we know what we'll get from McGloin - I think he'll be within 3-6 points of average, either negative or positive.  Miller is the wild card, especially coming off a game in which he left due to injury (officially ruled as whiplash).  I find this one particularly tough to call, and a quick perusal of Scott's projections tells me his statistical analysis agrees.

Notre Dame at Oklahoma (8:00, ABC)

Quarterbacks: Everett Golson (-7.46 EPA; -.041 EPApp) vs. Landry Jones (26.24 EPA; .117 EPApp)
Running Backs: Theo Riddick (-6.13 EPA; -.048 EPApp) & Cierre Wood (1.53 EPA; .031 EPApp) vs. Damien Williams (19.38 EPA/ .206 EPApp) & Dominique Whaley (1.01 EPA; .021 EPApp)

There's no doubt here that you want to watch this one for when Oklahoma sends their offense out against Notre Dame's defense, but before we get to that, let's look at the less intriguing matchup, ND's offense against the Sooner D.

Oklahoma's D has allowed -39.62 EPA & -.197 EPApp this year, and they've really only struggled with backups, suggesting they're quite good in preparation and film study.  Seriously, only 3 QBs have posted positive EPAs against Oklahoma this year.  One is the Heisman frontrunner, and the other 2 were backups, Case McCoy and Michael Brewer, who threw a combined 14 passes (but did so for an impressive 11.86 EPA).  There's no reason to expect Everett Golson to have anything except a very difficult time against this Sooner defense.  OU has allowed -13.69 EPA & -.096 EPA rushing this year, but those who can afford to try running typically can succeed.  Three running backs have gotten at least 20 touches against the Sooners this year and all 3 yielded positive EPAs - Nathan Jeffery of UTEP (.69), John Hubert of KSU (2.05), and James Sims of Kansas (.86).  We're not talking about All-American performances here, but all slightly above-average.  If Notre Dame's defense can keep the game close and have reason to run, they could have some success on the ground.

Now, about that.  Notre Dame's defense has allowed -77.84 EPA & -.262 EPApp to passers this year and -29.94 EPA & -.184 EPApp to runners.  The best total team passing against Notre Dame was probably Stanford's, and they were still 7 points worse than average.  Akeem Shavers of Purdue had some moderate success against ND in week 2, but it was merely moderate.  This defense looks impenetrable.  That doesn't bode well for Oklahoma, as Landry Jones hasn't looked great when confronted with good defense this year, and Damien Williams has looked worse.  I'm using the K-State game as my point of reference, and the Jones/Williams duo combined for just 0.19 EPA in that game.  And K-State's defense doesn't looke nearly as good as Notre "D", as they shall henceforth be referred.  If Notre Dame does give up points in bunches on Saturday, it could tell us a lot about the teams they've faced prior to this.  That being said, I'm not expecting a high scoring game, but other than that, I have no idea what will happen.  Looking forward to watching this one, as we'll finally get to really see if Manti Te'o deserves his Heisman hype.

Mississippi State at Alabama (8:30, ESPN)

Quarterbacks: Tyler Russell (43.41 EPA; .190 EPApp) vs. AJ McCarron (38.15 EPA; .207 EPApp)
Running backs: LaDarius Perkins (14.01 EPA; .097 EPApp) vs. Eddie Lacy (12.65 EPA; .115 EPApp) & TJ Yeldon (23.95 EPA; .263 EPApp)

Vegas isn't expecting this one to be close, but they are two undefeated teams meeting in late October, so the matchup deserves attention for that reason alone.  Also, my other choice was Nebraska/Michigan, and I wanted this to be about teams doing as well or better than expected, not worse.

MSU's defense has held opposing passers to -34.66 EPA & -.144 EPApp while holding runners to -14.12 EPA & -.100 EPApp.  The problem is that it's tough to know what to think of that - they haven't exactly played the cream of the FBS crop: Auburn, Kentucky, Tennessee, Middle Tennessee, Troy, and South Alabama.  To be fair, Alabama has actually fared similarly against Sun Belt competition this year, so perhaps we should at least consider MSU a viable contender, at least on defense.  McCarron has rarely been brilliant this year (last week being the exception) and while it seems a tall order to shut down Lacy & Yeldon, you could legitimately expect to keep them from beating you.

I must publicly admit my giddiness over what I'm about to do, and that's calculate Bama's defensive EPA allowed.  Having seen Notre Dame and Florida check in the -70's against passers, reputation and expectation might lead me to think Bama's would be in the -80's or even -90's.  They don't quite live up to their super-human reputation, though they're right there with the other 2 elite defenses covered in this preview.  Alabama has allowed -76.74 EPA & -.346 EPApp to passers and -31.01 EPA & -.214 EPApp rushing.  It's going to be a long afternoon for the MSU offense.  Denard Robinson and Kawaun Jakes, both ahead of Russell in EPA, had their worst games of the year when they faced this defense.  However, they were fairly close to average, so Russell might have a chance.  I think MSU's chances do depend on Russell, though.  They need a low scoring game and the game of the year from their QB.

They're unlikely to get it, but still, two undefeated SEC teams meeting this close to Halloween always gets my attention.  It should get yours as well.

No comments:

Post a Comment