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

Monday, September 30, 2013

Week 5's best performances

EP3 is the effective points per possession, or how effective an offense is with each possession. I adjust the raw EP3s to account for the strength of the competition.

Illinois comes out on top after scoring a 2.88 EP3. Overall in 2013, the Illini are 13th in EP3. Arizona State got Lane Kiffin fired, so you can guess that was a pretty good performance. Oregon had a very odd game. They got a big lead early on short fields and then cruised. The result is an unimpressive EP3, but still a fairly good Perf-O. The model got a little confused with Oklahoma State's -0.89. It decided that OSU might have been a little off their game but West Virginia was also playing world beater defense.

Florida, Washington, Virginia Tech and Alabama were also credited with playing good defense. Air Force was swiss cheese. The model gave little credit to Nevada and New Mexico for big offensive performances, instead placing the blame on the defense.

Friday, September 27, 2013

Week 5 Saturday Preview: The 3:30-6:30 games

An EPA-centric preview of every game this weekend.  To find out more about Expected Points Added, click here.

#6 LSU (4-0) at #9 Georgia (2-1), 3:30, CBS

LSU rush O (.17) vs. UGA rush D (-.006)
LSU pass O (.59) vs. UGA pass D (.185)
UGA rush O (.08) vs. LSU rush D (-.011)
UGA pass O (.61) vs. LSU pass D (-.081)
Zach Mettenberger indeed has been much better this season, and it seems that's the national story of the week.  As good as he's been, Aaron Murray has been better, with slightly better per-play stats against considerably better competition.  Murray always seems to thrive out of the spotlight, and despite his elite level of play so far, that's exactly where he finds himself in the days leading up to this game.  LSU's defense has hardly proven to be elite, registering as slightly above average against underwhelming competition so far.  While EPA suggests LSU will have a clear mismatch when throwing the ball, adjusting for opponent suggests it might not be as much of an advantage as it seems.  Georgia has already faced two teams that could very well be better than LSU this year.  LSU hasn't faced anything comparable to Georgia, and they're on the road.  Am I predicting a Georgia win?  I'm not predicting anything either way.  I've simply noticed a disproportionate number of analysts taking the Tigers in what should be a close, exciting game.   
Player to watch: Georgia QB Aaron Murray (53.5 EPA) ranks 2nd in the country in opponent-adjusted EPA.

UTEP (1-2) at Colorado State (1-3) - I like the Rams in this one, mostly due to UTEP being bad at defense this year.

#14 Oklahoma (3-0) at #22 Notre Dame (3-1), 3:30, NBC

OU rush O (.17) vs. ND rush D (-.029)
OU pass O (.07) vs. ND pass D (-.010)
ND rush O (-.10) vs. OU rush D (-.136)
ND pass O (.14) vs. OU pass D (-.211)
The Oklahoma defense has been very effective this year, and coupled with a somewhat balanced offense, the Sooners should have the advantage in this game.  It's a premier matchup between two storied programs.  This is what all non-conference games should be.
Player to watch: Oklahoma QB Blake Bell (31.3 EPA) ranks 9th nationally in opponent-adjusted EPA per pass.

#8 Florida State (3-0) at Boston College (2-1), 3:30, ABC/ESPN2

FSU rush O (.39) vs. BC rush D (.033)
FSU pass O (.59) vs. BC pass D (-.006)
BC rush O (-.08) vs. FSU rush D (-.041)
BC pass O (.01) vs. FSU pass D (-.354)
I'm surprised this game gets a national billing, as FSU seems poised for a blowout victory.  There's not a facet of the game where I expect the Eagles to get the upper hand.
Player to watch: Florida State QB Jameis Winston (51.5 EPA) ranks 7th in opponent-adjusted passing EPA.

Iowa (3-1) at Minnesota (4-0), 3:30, ABC/ESPN2

Iowa rush O (.05) vs. Minn rush D (-.173)
Iowa pass O (.06) vs. Minn pass D (.125)
Minn rush O (.28) vs. Iowa rush D (-.217)
Minn pass O (.00) vs. Iowa pass D (-.172)
Last week, I promised that I'd be a Gopher believer if they handled San Jose State.  They did, and now I am.  Unfortunately, I'm going to pick Iowa to win.  The Hawkeyes are more battle tested and their defense is pretty superior to Minnesota's.  However, the Gophers are improved yet again under Jerry Kill, and if they took home Floyd of Rosedale for the second straight year, it would be only a mild surprise.
Player to watch: Minnesota RB David Cobb (14.9 EPA) is 23rd in rushing EPA.

Central Michigan (1-3) at NC State (2-1) - NC State will win, but they haven't been quite right on offense since Brandon Mitchell's injury.

Week 5 Saturday Preview: The Early Games (12-3)

An EPA-centric preview of every game this weekend.  To find out more about Expected Points Added, click here.

#15 Miami (3-0) at South Florida (0-3), 12:00, ESPNU

Miami rush O (.25 EPA per play) vs. USF rush D (-.026)
Miami pass O (.27) vs. USF pass D (.045)
USF rush O (-.24) vs. Miami rush D (-.008)
USF pass O (-.49) vs. Miami pass D (-.473)
There's a good argument that South Florida is the nation's most disappointing team so far.  Willie Taggart seemed like a perfect fit.  He had legitimate success at Western Kentucky, and he hails from the greater Tampa area.  The thing is, he might be the perfect fit, but if the perfect fit inherits no talent in the passing or rushing attacks, the perfect fit won't look like the perfect fit.  USF ranks 123rd in EPA per pass, costing the team a point every 2 times they throw the ball.  The defense wasn't great in the 53-21 loss to McNeese State, but even that final score doesn't tell the full story.  9 points were scored when the D was off the field, one TD came on a drive that only required 7 yards, one FG was allowed on a 3-and-out that started in Bulls territory, and another TD capped a 16 yard drive.  That's 26 points either directly from or helped greatly by the problems of the offense or special teams.  Since that game, the defense has managed to keep things manageable, which is a very tall order with complete offensive ineptness, which is what the Bulls have seen this year.  Saturday at noon, you can catch that ineptness on display.  It's unlikely to change against a decent Miami defense.
Player to watch: Miami RB Duke Johnson (8.6 EPA) is 19th in opponent-adjusted rushing EPA.

#11 Oklahoma State (3-0) at West Virginia (2-2), 12:00, ESPN

OSU rush O (.16) vs. WVU rush D (-.095)
OSU pass O (.39) vs. WVU pass D (-.103)
WVU rush O (.06) vs. OSU rush D (-.101)
WVU pass O (-.13) vs. OSU pass D (-.067)
The Cowboys have been somewhat quiet for a top 15 team.  They slowly squeezed the life out of Mississippi State on opening weekend but were overshadowed by bigger matchups.  This game will also likely be overshadowed, but that has more to do with their opponent - West Virginia has been not at all as expected in pretty much all facets.  They've been better than average on defense.  They've been bad passing the ball.  It's a weird time to observe Mountaineer football.  If last week's game against Maryland is any indication, it's also going to be a long year for fans of Mountaineer football. 
Player to watch: Oklahoma State QB JW Walsh (42.6 EPA) is 18th in opponent-adjusted passing EPA.

Miami (OH) (0-3) vs. Illinois (2-1) - The Illini should eclipse last year's win total with this game.  After this, the schedule starts to toughen up: Nebraska, Wisconsin, Michigan State, and Penn State are the next 4 opponents.  After this win, for Illinois to go bowling, it's going to take wins at Indiana and at Purdue, plus an upset of a considerably better team along the way.  Still, 3-5 wins is an improvement.  Miami continues to regress in Don Treadwell's 3rd and likely final year.

Thursday, September 26, 2013

How good is Baylor?

In three games, Baylor has scored 69, 70 and 70 points for the highest points per game in the country. On the other end, they have allowed only 23 points and the second fewest points per game. Both of these numbers are a little biased by tempo: Baylor plays fast, so they have more opportunities to score in a game than most. Baylor is only 4th in points score per possession (4.11), but no one has allowed fewer points per possession than the Bears (.49).

But to say Baylor is 4th in points per possession is again misleading. Against Wofford, Baylor was up 28-0 at the end of the first quarter. Against Buffalo they had scored 56 points at the half. And in their last game against ULM Baylor was up 42-0 two minutes into the second quarter. It's hard to say a team is calling off the dogs when they are scoring 69.7 points per game, but Baylor is averaging more than 47 points in the first half.

The chart below shows points per possession, but breaks it down by drive number - points per possession in the first drive of each game, through the first two drives, etc - for the country's top scoring offenses. Essentially, early in games, through the first 10 possessions, Baylor is scoring more often than any team in the country. Combine that with the .49 points per possession and -1.63 effective points per possession allowed, and Baylor has been, hands down, the most dominant team in the country. Meanwhile, Texas allowed 550 rushing yards to a team that has lost their two other games, West Virginia lost 37-0 to Maryland and Oklahoma beat WVU 16-7. This could be Baylor's year.

Wednesday, September 25, 2013

Quick observation about Oregon (the state)

Oregon leads the nation in yards per rush. That's not uber surprising. Oregon State, on the other hand, is averaging 2.57 yards per carry. Only FIU is worse, and just barely.

Oregon State's Sean Mannion is leading all passers (and all players) with 80.6 expected points added. Teammate Brandin Cooks is leading all receivers with 54.5 expected points added.

Tuesday, September 24, 2013

Week 5 picks

We're getting closer to reliable picks from the model, but not quite there yet. 

The model likes Oklahoma by 8 over Notre Dame, Ohio State by less than 3 and Georgia by 5.

Monday, September 23, 2013

Week 4's top performances

I'm going to wait another week before posting opponent-adjusted EP3s. 

Bowling Green had the most dominant offensive performance of the week. They scored touchdowns on 7 of 10 possessions, including drives of 99 and 97 yards. Savannah State is maintaining a streak of horrific performances against FBS schools. Mississippi State's offensive explosion against Troy was unsuspected (by me, at least).

Miami scored a -2.33 against Cincinnati. That is almost impossible. The typical possession is worth about 2.1 points, so in most cases a team can do better than -2.33 by just punting the ball away on first down. But Miami gained positive yards on only 8 of 14 possessions, so they were almost better punting on first down. Seven yards on 29 rushes is amazing. West Virginia's -2.28 is little better, but they got there mostly through turnovers.

Sunday, September 22, 2013

Week 4 updates

What is NOT updated:
Picks - I will have unofficial picks on Tuesday
Conference stats - Unless someone convinces me otherwise, this will be relegated to the dustbin of history
Team Summaries - Technical issue. Hope to have some this week to fix it
Teams: Team, Team: Summary - see Team Summaries
Teams: Special Teams - I don't actually have anything to update yet

Also, keep in mind that any opponent-adjusted state ("+") is suspect at this point. Not until week 4 or 5 do I put any confidence in these numbers.

And Georgia State is excluded for now. It's not because I don't love them, but I don't love them enough to make the necessary adjustments just yet for a team that is completely irrelevantly.

Saturday, September 21, 2013

Week 4 Saturday Preview: The Night Games (7:00-)

#23 Arizona State (2-0) at #5 Stanford (2-0), 7:00, FOX

ASU rush O (.04) vs. Stanford rush D (-.091)
ASU pass O (.28) vs. Stanford pass D (.030)
Stanford rush O (.16) vs. ASU rush D (-.178)
Stanford pass O (.28) vs. ASU pass D (-.128)
ASU might have a better shot than people really think.  Taylor Kelly is a good QB, and the Sun Devil D has been pretty solid so far, and it's not as if they've only played vastly inferior opponents - Wisconsin isn't bad at all.  I think the Stanford offense is good enough to match the Sun Devils, but if this winds up an upset, it won't be a stunning one.  EPA star to watch: Arizona State QB Taylor Kelly (27.1).

Morgan State (0-3) at Western Kentucky (1-2) - Hilltoppers will win big.

SMU (1-1) at #10 Texas A&M (2-1), 7:00, ESPNU

SMU rush O (.21) vs. aTm rush D (.186)
SMU pass O (-.01) vs. aTm pass D (.273)
aTm rush O (.29) vs. SMU rush D (.016)
aTm pass O (.47) vs. SMU pass d (.259)
Texas A&M's defense has been pretty bad at football so far this year, but it won't come back to bit them in this one.  Johnny Manziel won't have any problems with the Mustang defense.  EPA star to watch: Texas A&M QB Johnny Manziel (53.1).

Colorado State (1-2) at #1 Alabama (2-0) - This one is on ESPN2, but it's not because anyone expects it to be close.  Do you really need a preview?

Week 4 Saturday Preview: The Mid-Afternoon Games (3:30-6:30)

Utah State (2-1) at USC (2-1), 3:30, ABC/ESPN2

USU rush O (.07) vs. USC rush D (-.230)
USU pass O (.58) vs. USC pass D (-.349)
USC rush O (.09) vs. USU rush D (-.171)
USC pass O (-.18) vs. USU pass D (-.002)
In backyard football growing up, we used to occasionally have 'full time QB', one kid who played QB for both teams.  This game would be much better if Chuckie Keeton were full time QB.  Unfortunately, USC has to provide their own QB.  Fortunately for us, we get to see Keeton go up against a stifling Trojan defense.  EPA star to watch: Utah State QB Chuckie Keeton (71.2, #1 in the country) is the nation's most productive player so far, and with the right performance Saturday, he could move into the Heisman discussion.

Tennessee (2-1) at #19 Florida (1-1), 3:30, CBS

Tenn rush O (.16) vs. Florida rush D (-.342)
Tenn pass O (.14) vs. Florida pass D (-.202)
Florida rush O (.01) vs. Tenn rush D (.055)
Florida pass O (-.06) vs. Tenn pass D (.025)
It's hard to know what to make of Tennessee, as they really haven't played any normal competition yet.  This is yet another tough matchup, but not in the same way as Oregon.  Florida is not likely to score over 60 points.  Florida's tendency to be bumbling on offense could keep this game too close for comfort for the Gators, but I think they'll still win.  The Florida defense is still one to behold, and while they may only have one great unit, that's one more than Tennessee.  EPA star to watch: Tennessee QB Justin Worley (10.7).

Maine (3-0) at #18 Northwestern (3-0) - Maine is no pushover, but Northwestern is well-coached and not the kind of team to be unprepared for cupcake matchups.

Week 4 - Saturday Preview: The Early Games (12-3)

Apologies if you were looking forward to a weekday games preview.  I'm sure I would have predicted Clemson to struggle and Fresno to edge Boise by a point.  Take my word for it.  This week's preview is again divided into three posts: one detailing the games starting 12-3, one for games 3:30-6:30, and one for games 7:00 and later.  It worked well last week, and it will probably be the norm.  Let's delve right in, with EPA details for the nationally televised matchups.

North Carolina (1-1) at Georgia Tech (2-0), 12:00, ESPN

UNC rush O (.02) vs. Tech rush D (-.129)
UNC pass O (.11) vs. Tech pass D (-.437)
Tech rush O (.33) vs. UNC rush D (.020)
Tech pass O (.97) vs. UNC pass D (.037)
Georgia Tech ranks 1st in EPA per pass, for two big reasons - they only throw 19.1% of the time (122th in the country), and they've played bad teams so far.  UNC's defensive numbers don't look great, but they played one really good team (South Carolina).  The feeling I get here is that Tech isn't nearly as good as the EPA stats suggest, and North Carolina probably isn't quite as bad.  Evening the teams out means this should still be a good game, a potentially key one in the ACC Coastal race.  The two teams combined for 118 points in regulation last year.  EPA star to watch:  Georgia Tech QB Vad "The Impaler" Lee (27.9).

FIU (0-3) at #7 Louisville (3-0) - Every FIU preview for this entire season shall mention Mario Cristobal.  So, Mario Cristobal.

Vanderbilt (1-2) at Massachusetts (0-3) - Is ESPNews a national broadcast?  Technically, yes, but let's pretend it's not just so we don't have to analyze this.  All we should think about going into this one is that no matter how much James Franklin talks up Vanderbilt and how much their fans think they're moving into the upper echelon of the SEC, they're still taking road trips to Massachusetts.

San Jose State (1-1) at Minnesota (3-0), 12:00, ESPN2

SJSU rush O (.08) vs. Minn. rush D (-.141)
SJSU pass O (-.02) vs. Minn pass D (.051)
Minn rush O (.26) vs. SJSU rush D (-.030)
Minn pass O (.03) vs. SJSU pass D (.073)

Friday, September 20, 2013

Johnny Manziel is also really good at football

Two days ago I did some analysis on Mike Evans' performance against Alabama. It was good (Evans, not the analysis . . . though the analysis was pretty good, too). I promised a follow up on Manziel, and I deliver:

Manziel threw for 464 yards on 39 attempts and rushed for another 98 on 14 carries (a sack cost him 5 yards, so Manziel exceeded 100 "rushing" yards). Given the opposition, this marks the best offensive performance in the history of college football . . . if it weren't for two interceptions.

Manziel totaled 562 yards on Saturday; that's the 19th most yards by a player in a game since 2005. In those 8 1/4 seasons, the 550 mark has been eclipsed 25 times. Manziel has done it three times, Colt Brennan did it twice, and 20 others did it once.

But we should also account for the quality of the opposition. Geno Smith, for example, totaled 687 yards against Baylor, but Baylor allowed over 500 yards per game in 2012. To account for this, next to the yards I've listed the average yards per game allowed by each opponent. In Manziel's case, I used Alabama's 2012 average yards per game, but I substituted A&M's 628 yard game Saturday for their 418 yard game against Alabama last year; Alabama's average yards allowed increased from 250 to 265 as a result. In the final column, I divided the two: Yards/Average yards allowed.

When we do this, Manziel's game against Alabama comes out on top. After his 2.12, Colt Brennan is a distant second with 1.75 followed by Cody Hodges in 2005 with a 1.68. In fact, Manziel's 2.12 is the biggest number not just on this list but on any list. When we look at all players in all games, the second highest yards/average yards allowed ratio is 1.82 from Chase Holbrook against Boise State in 2005 followed by Geno Smith's 1.79 against LSU in 2011. Manziel more than doubled the average yards allowed per game of his opponent; no one else has come close . . . at least recently . In short, Manziel's 562 yards was the most impressive yardage total of the last 9 years, and not by a little.

The best stat for measuring an individual players' production (which, I admit, is different than performance, but we'll get back to that) is the EPA, or expected points added. Essentially, it looks at where a team was before a play, in terms of score and field position, and after a play and assigns a point value to the play. That point value is based on how the score and field position changed. Field position, down and distance are converted to an expected point value based on what teams do on average over the course of the possession from that situation. 

If that doesn't make any sense, remember this: the higher the EPA the more a player did to help his team score more points and make it harder for the other team to score points.

Against Alabama, Manziel scored a 22.6. That's good. It is the 15th highest single-game EPA by a player this season (AJ McCarron scored a 23 in that same game). But it could have been much better. Sunseri grabbed the tipped pass and took it back for a touchdown. On that one play Manziel scored a -7.2. Remove that play and Manziel scores a 29.8 and the 4th highest single-game EPA of the season (2 points behind Mariota vs. Tennessee). If Manziel had managed to bring down Sunseri after the interception he finishes with a 26.4 (8th best).

We can't just remove bad plays and highlight good plays, but we also shouldn't compare apples and oranges. The three highest EPA performances this season came from Mariota (vs. Tennessee), Blake Bell (vs. Tulsa) and Keeton (vs. Air Force). Each performance is impressive in its own right, but racking up stats against Tennessee, Tulsa and Air Force, or even McCarron's 23 against A&M, is not the same as doing it against Alabama.

Fortunately, we can adjust a player's performance for the strength of the competition. Basically, we look at what other players have done against that defense and against all other defenses, throw it in a fancy algorithm, and come up with a value that reflects how hard it is to run, throw and catch against each team.

When we adjust for the competition, Manziel's 22.6 becomes a 50.5, the highest-single game EPA+ ("+"=adjusted) of the season (Mariota's 40.8 against Tennessee is tied for second). But we shouldn't put too much stock in these results. Manziel has only played one half against another FBS opponent and Alabama had only played Virginia Tech. The 50.5 is the computer's way of saying, "Holy smokes, this Manziel guy is a lot better than Logan Thomas." That's not news.

So instead, I threw this game in the 2012 season - what if 2012 Johnny Manziel had done this against 2012 Alabama? How would the computer interpret that? When we do that, and adjust for competition, Manziel against Alabama comes in 22nd since 2005. Great, but not historical. Sam Bradford had three better games, RGIII, Colt Brennan, and Vince Young had two. Even Manziel himself scored higher against Missouri last season. Geno Smith against Baylor, the leader of the pack, was 5.7 points better (Smith's raw EPA from that game was a 51.3, almost 30 points higher than Manziel's). This is not an insignificant gap, but if that tipped ball had somehow fallen through Sunseri's hands, Manziel would have had the single-most productive game for an individual player in the last nine years . . . and probably of all time. Throw out the second interception and this performance would be a full touchdown better than any we've ever seen before. Unfortunately, football doesn't work like that.

Subjectively, given the moment, I've said that I think this was the most impressive offensive performance in the last decade and maybe of all time (though I'm admittedly too young to speak of such things). The accumulated stats with the (insufficient) 4th quarter comeback are awe worthy. If the A&M defense had come up with a couple more stops, the rest of the college football world might see it the same way.

Thursday, September 19, 2013

A few quick thoughts on 550

You might have heard that Texas gave up 550 rushing yards to BYU. While I'm almost two weeks late, I wanted to take a little time and put that game in historical perspective.

First, 550 rushing yards in a game is a lot. It's the most so far this season and the 8th most since 2005. But BYU needed significantly more carries than the other teams on the list. On the other hand, this list puts Texas in the company of Kansas (and not 2008 Kansas), New Mexico State, North Texas and Eastern Michigan.

But perhaps more important than the total number of yards is the total number of carries. If BYU had averaged 7.6 yards over 30 carries Manny Diaz would still be employed . . . even though his defense was no more successful at doing its job. In fact, BYU's 72 rush attempts against Texas is 20th most by a team in a game since 2005. Texas also allowed more yards per carry in a game in which the opponent attempted more than 65 carries in that period, so that's not good. The 72 attempts were the result of 1) BYU averaging 7.6 yards per carry, but also 2) BYU's passing game was in the bottom quintile in terms of efficiency this season.

And that brings us to efficiency. BYU's EPA (expected points added) on running plays was 25.1, the sixth most this season. In other words, while the Texas defense was bad against the run, 5 teams have been worst. More accurately, 4 teams because Nicholls State made the list twice, with Colgate, Elon and Indiana rounding things out. And from whom did we get these dynamic rushing performances? Exactly who you'd expect: Oregon, Georgia Tech, Air Force, Navy and Louisiana-Lafayette. And now BYU.

Taysom Hill's performance, though, was particularly impressive. He contributed 259 rushing yards and 15.2 per attempt. His rushing EPA of 23.3 for the game is the most this season, the only to top 20 and one of only two to top 15 points. It is more than all but four other players have totaled for the entire season. In other words it is, hands down, the best rushing performance of 2013.

Deconstructing the myth of Jadeveon Clowney

After The Hit, the Jadeveon Clowney hype went out of control, creating unrealistic expectations. As you would learn in any intro sociology or psychology class, unmet expectations lead to frustration and conflict. Clowney is frustrated with the coaches, the coaches are frustrated with each other, the fans are frustrated with the universe, and ESPN has a new explanation every week for why their low-watt star has been, to this point, all smoke.

The newest explanation is my favorite - bone spurs. Apparently, a condition that he has dealt with since high school and didn't merit attention during the offseason is an explanation for his struggles. I played football and baseball with bone spurs in my foot. I developed a stress fracture as a result. It's not ideal, but it has nothing to do with Clowney's lack of production.

The real explanation for Clowney's lack of production is that there is no lack of production. We confused myth with reality (and by "we" I mean you, not me; I've seen through this from the beginning). The fact of the matter is that Clowney 2013 looks very much like Clowney 2012.

In the chart below I've laid out Clowney's statistical production through 2 games, 3 games and for the entire season, and then below that I calculated the production per game. Like 2013, he got off to a slow start in 2012; in fact, his 2013 numbers are better. Only in his third game of 2012 did Clowney make his presence felt (statistically).

Using his production per game so far, in the second to last column I project his production through 13 games. He is on pace to significantly lower totals in 2013, but the gap is not as large as we might believe. In the final column I calculate what Clowney would need to do in game 4 to put him back on his 2012 pace. Seven tackles, 4.5 TFLs and 2 sacks - less than what he did against Clemson last year - and Clowney would be ahead of schedule.

The bigger issue is that the USC defense has struggled, and this is ultimately where we have been misled. On offense, the quarterback is the offense. Other players make contributions, but a great quarterback can make any offense good, and a bad quarterback can make any offense dysfunctional. There is no equivalent on defense. At the end of the day, Clowney is one player, one very good player, on a defense with serious question marks. He alone can't overcome that.

Wednesday, September 18, 2013

Mike Evans is really good at football

I start with the obligatory disclaimer: Alabama won the football game. They have the best team in the country and will probably win another national championship. AJ McCarron did exactly what he was supposed to in this game. The offensive line either made the necessary adjustments after Virginia Tech or the A&M front seven is an embarrassment to the SEC (probably a bit of both). T.J. Yeldon is a man, except for his annual tradition of fumbling against A&M at the worst possible moment. Nick Saban is a defensive genius; he needs only 10 months of preparation to take the nation's most talented defense and hold a 6'1" - 200 lb quarterback to under 100 yards rushing and 500 yards passing. Bow before the rolling Tide.

But as a football fan, that's not interesting. What is interesting is that Mike Evans had 279 receiving yards on 9 targets and Manziel had the second most yards by an SEC player in the history of the conference, and they did  it against the two-time defending best defense in the country. I hope to break down Manziel's performance in a part two, but here I focus on Mike Evans:

Since 2005 (where my play-by-play data begins), players have gained 200 yards receiving 170 times and and 250 yards 24 times. Evans' 279 yards ranks 12th over that period. Looking over the list of the top 20 games by receiving yards, a few things stand out. First, most of these were against bad defenses. That's not to say the accomplishment is unimpressive, but we should keep the strength of the competition in mind. Some receivers amassed their numbers against solid defenses: Marqise Lee and Austin Hill both topped 250 in the same game; Cobi Hamilton gained over 300 yards against a good Rutgers defense; Dezmon Briscoe's 269 against Oklahoma is notable. But Mike Evans had 279 yards receiving against the best defense in college football.

Second, Mike Evans is the only player on this list in single digits in targets. His 31 yards per target is 7.5 more than any other on the list. Manziel could have targetted Evans more often, but Kevin Sumlin also noted that much of what they were able to do on the ground was based on Alabama's approach to containing Evans defensively. Unfortunately, this is the kind of contribution I have no way of measuring (for now).

While this post is about Evans, Jheranie Boyd against LSU deserves a little attention
No player has gained more receiving yards on 9 or fewer targets in a single game (see above), but there is a better way of handling this. I calculated yardage totals for any 9 consecutive targets to a player in the same game. For example, if a player was targeted 12 times in a game, I would look at targets 1-9, targets 2-10, 3-11, and 4-12 independently. Still, Mike Evans' 279 yards in the most since 2005. In fact, the gap between Evans and the second best (27 yards) is as big as the gap between the second best and the 20th best series of 9 consecutive targets. And, as I've said before in this post, Mike Evans did it against Alabama. Quod erat demonstratum, Mike Evans is good at football.

Tuesday, September 17, 2013

Week 4 unofficial picks

I won't start posting the official picks for at least another week - you need at least four weeks of data for the model to really work - but I promised week 4 picks so here they are. 

Louisiana Tech. Last year they took Texas A&M to the wire. This year, Vegas likes Kansas by 9 and the model likes Kansas by more than a dozen. The model likes Stanford by less than a field goal; that is to say, the model likes Stanford by less than Joel Stave deciding to lay the ball on the ground so it could get dogpiled and the ref failing to get the ball set in time so the clock expires. It gives Georgia Tech a massive 18 point advantage against UNC, and the model likes UCLA by 50, which is its way of saying UCLA can pick the score. 

Monday, September 16, 2013

Best and Worst Performances, Week 3

The EP3 (effective points per possession) measures a team's offensive performance by considering both the points scored per possession and the impact of each possession on field position. For example, Cincinnati scored a 3.69 against Northwestern State. This means that Cincinnati was 3.69 points per possession more effective on offense than the average team, both in terms of scoring points based on their starting field position and leaving the opponent with relatively bad field position. In later weeks I will adjust the weekly EP3s for the strength of the competition, but for now I am just reporting the raw values.

Cincinnati against Northwestern State was the second most effective offensive performance of the season. Brendon Kay averaged 18.3 yards and 1.56 expected points added per pass attempt, the highest figure of the season. Zach Mettenberger wasn't much worse as he averaged 14.7 yards and 1.17 points added per pass. 

Against Wagner, Syracuse allowed 0 points, 87 yards and forced 3 turnovers. Wagner scored a -1.9 against the Orange; the average possession is with just over 2 points, so Wagner would have been little worse off if they had just started every possession with a punt. And never would have guessed that both TCU and Texas Tech would make the best defenses list when they were playing each other.

Sunday, September 15, 2013

Week 3 updates

What is NOT updated:
Picks - coming soon
Conference stats - Unless someone convinces me otherwise, this will be relegated to the dustbin of history
Team Summaries - Technical issue. Hope to have some this week to fix it
Teams: Team, Team: Summary - see Team Summaries
Teams: Special Teams - I don't actually have anything to update yet

Also, keep in mind that any opponent-adjusted state ("+") is suspect at this point. Not until week 4 or 5 do I put any confidence in these numbers.

And Georgia State is excluded for now. It's not because I don't love them, but I don't love them enough to make the necessary adjustments just yet for a team that is completely irrelevantly.

What we learned, week 3

1) Alabama has a better football team than Texas A&M. Either the 'Bama o-line got a lot better in two weeks or the Aggie front 7 seven is atrocious. My guess is something between those two.

2) But Manziel and Evans were hands the best players on the field and are the best players in the country at their respective positions (yeah, Marqise Lee, I'm talking to you).

3) Blake Bell can run an offense against Tulsa and, more surprising, we're supposed to be impressed by that. For those that didn't get the memo, Tulsa's defense is not good. Consider this, Bowling Green scored 34 on Tulsa and 10 on Indiana. A little transitive property, and OUs 51-20 win becomes a 27-20 win if Tulsa borrowed Indiana's D. I'm not impressed.

4) Eastern Washington might have been a fluke. After an amazing offensive performance in week 1 against Oregon St, they put up 21 on Toledo. Elsewhere, Oregon St allowed 12 yards against Utah in the 1st quarter and 528 in the game.

5) Nebraska. What I didn't learn is why we don't spend more time ragging on the Huskers and instead focus our attention on . . .

6) Texas. We didn't learn anything here. They lost to a superior opponent. It happens.

7) Devin Gardner really let me down. I set him up as the anti-Braxton Miller, but yesterday he laid an egg. Speaking of . . .

8) Miller, Guiton steps in and puts up 368 yards and no INTs. He didn't set the world on fire (you should probably take anything anyone does against Cal and divide it by 2), but it's no worse than Miller would have done.

9) FSUs Winston is good, but Nevada is even better at being bad. Winston is particularly good at completing passes, something Teddy Bridgewater used to do better than the rest.

10) I'd be getting nervous if I were a Stanford fan. They knew exactly what Army was going to do, Stanford countered with what is supposed to be an elite run defense, and Army still rolled up 284 yards on 61 tries. That's not terrible, it's not 550, but it won't beat Oregon.

Saturday, September 14, 2013

Week 3 Saturday Preview - The Night Games

A preview of games starting 7:00 or later.

Mississippi State (1-1) at Auburn (2-0), 7:00, ESPN2

MSU rush O (.02) vs. Auburn rush D (-.022)
MSU pass O (.09) vs. Auburn pass D (-.074)
Auburn rush O (.13) vs. MSU rush D (-.084)
Auburn pass O (.15) vs. MSU pass D (-.212)
So far this year Auburn has the better offense, and MSU has the better defense.  Obviously that's the matchup to watch.  EPA star to watch: Auburn RB Corey Grant (15.0), who ranks 14th nationally in rushing EPA.

#4 Ohio State (2-0) at California (1-1), 7:00, FOX

OSU rush O (.30) vs. Cal rush D (.131)
OSU pass O (.11) vs. Cal pass D (.171)
Cal rush O (-.12) vs. OSU rush D (-.192)
Cal pass O (.14) vs. OSU pass D (-.123)
The Bears don't have the defense to really contain the Buckeyes, so I'm not expecting an upset.  However, as the Cal players get more experience in Sonny Dykes' Bear Raid (which is a name I unabashedly love), they're only going to improve.  That alone should make this game worth watching.  EPA star to watch: Cal WR duo Chris Harper (19.6) and Bryce Treggs (19.5), who rank 12th and 13th respectively in receiving EPA.

Florida Atlantic (0-2) at South Florida (0-2) - If we had some kind of pillowfight of the week designation, this would probably win.  The highlights for these two teams this year include losing to McNeese State by 30+ and spiking the ball on 4th down.

Kent State (1-1) at #8 LSU (2-0), 7:00, ESPNU

KSU rush O (-.10) vs. LSU rush D (.018)
KSU pass O (.25) vs. LSU pass D (-.095)
LSU rush O (.09) vs. KSU rush D (.010)
LSU pass O (.54) vs. KSU pass D (.353)
The LSU from the opening weeks isn't the LSU we're used to, playing sub-par rush defense and doing most of their offensive damage through the air.  The Kent State we've seen has also been unexpected, mostly because Dri Archer, the country's most productive RB in 2012, has played only one series.  Archer is expected back in this game.  He's hoping that LSU's trademark rush D isn't.  EPA star to watch:  LSU QB Zach Mettenberger (32.2).

Nicholls State (1-1) at UL-Lafayette (0-2) - The Ragin' Cajuns should get win number one tonight.
Northwestern State (2-0) at Cincinnati (1-1) - With Munchie Legaux out for the year, the Bearcats get an FCS opponent to break in the new starter.
Massachusetts (0-2) at Kansas State (1-1) - Bill Snyder either beats opposing QBs or writes lovely, complimentary notes to them.  And Bill Snyder's all out of stationary.
Memphis (0-1) at Middle Tennessee State (1-1) - That moment you realize Memphis is in the AAC, an automatic BCS qualifying conference - that's a weird moment.

Vanderbilt (1-1) at #13 South Carolina (1-1), 7:00, ESPN

Vandy rush O (.20) vs. SC rush D (.002)
Vandy pass O (.24) vs. SC pass D (.346)
SC rush O (.13) vs. Vandy rush D (.169)
SC pass O (.30) vs. Vandy pass D (-.152)
Both offenses have looked very good so far.  Vandy's pass defense has been impressive.  South Carolina's pass defense looks awful, but they've faced UNC (who they did pretty well against) and UGA (who they didn't), two much better passing teams than Vandy has faced.  Likewise, SC's rush defense numbers look pretty good considering they've faced Todd Gurley.  The numbers here like Vanderbilt in the upset, but I think this is why we have to beware the numbers to an extent this early in the year.  This is potentially SC's easiest game of the year so far, and the Gamecocks should win.  EPA star to watch: Vandy QB Austyn Carta-Samuels (26.1).

Eastern Washington (2-0) at Toledo (0-2) - EWU upset Oregon State, and Toledo takes a break from its grueling SEC East schedule to face them.  The game should be fairly high scoring.
Western Kentucky (1-1) at South Alabama (1-1) - Western Kentucky leads Toledo by one game in the SEC East standings.
Maryland (2-0) at Connecticut (0-1) - Randy Edsall returns to Connecticut, and he'll hear boos.  Whether they're for him or Paul Pasqualoni, we can't be sure.  Note: We can be sure they're for both.
Lamar (1-1) at #12 Oklahoma State (2-0) - A lot of attention has been given to the Oklahoma State scandal, but little seems to be focused on OSU playing against a man named Lamar in week 3.  That hardly seems fair.
Kansas (1-0) at Rice (0-1) - Technically this is nationally telecast on CBS Sports Network, but I'm not sure they want that advertised.  Also, because I may not get the chance to do so much longer, allow me to type the words "undefeated, Charlie Weis-coached".  Thanks.
UTEP (0-1) at New Mexico State (0-2) - Someone has to win, and that someone will probably be UTEP.
Marshall (2-0) at Ohio (1-1) - Rakeem Cato has just 75 plays in 2 games.  Is he injured?
#25 Ole Miss (2-0) at Texas (1-1) - Manny Diaz got scapegoated last week, but its doubtful much will change.  That being said, last week was just one week, and Texas is still probably a pretty decent team, so a win today would not be out of the question or even particularly surprising.  This one is on the Longhorn Network if you get it.
Weber State (1-1) at Utah State (1-1) - Chuckie Keeton could put up some monster stats against this FCS team.

#21 Notre Dame (1-1) at Purdue (1-1), 8:00, ABC

ND rush O (.05) vs. Purdue rush D (.092)
ND pass O (.14) vs. Pudue pass D (.017)
Purdue rush O (-.10) vs. ND rush D (.041)
Purdue pass O (-.30) vs. ND pass D (.012)
Notre Dame's strongest asset is their passing offense.  In fact, the Notre Dame offense is the only unit in this game that has graded as above average so far this year.  I feel weird.  EPA star to watch:  Notre Dame WR TJ Jones (18.6).

Western Michigan (0-2) at #17 Northwestern (2-0) - The phrase "young energetic coach" will get a lot of mileage tonight.

Oregon State (1-1) at Utah (2-0), 10:00, Fox Sports 1

OSU rush O (.03) vs. Utah rush D (-.070)
OSU pass O (.36) vs. Utah pass D (.003)
Utah rush O (.24) vs. OSU rush D (.053)
Utah pass O (.48) vs. OSU pass D (.302)
When the two largest numbers in either column are on the same line, that's a mismatch, and tonight Utah's passing game represents a mismatch for Oregon State's pass defense.  It's taken two years, but Kyle Whittingham's Utes are finally starting to look like a Pac-12 team rather than a MW team playing a Pac-12 schedule.  The Beavers' passing game should be able to keep them competitive in what should be an exciting late night game.  EPA star to watch: Both QBs, Oregon State's Sean Mannion (40.1, #4 nationally) and Utah's Travis Wilson (39.3, #6 nationally).

Central Michigan (1-1) at UNLV (0-2) - UNLV is one of the worst passing teams in the country and has a whopping 31% of runs go for negative yardage.  CMU isn't good, but against UNLV, I'm not sure you have to be.

#20 Wisconsin (2-0) at Arizona State (1-0), 10:30, ESPN

Wisconsin rush O (.36) vs. ASU rush D (-.565)
Wisconsin pass O (.49) vs. ASU pass D (-.484)
ASU rush O (.04) vs. Wisconsin rush D (-.432)
ASU pass O (.60) vs. Wisconsin pass D (-.234)
Fun fact: Neither team has allowed a point this season.  Why that's meaningless: the best opponent either has faced is Massachusetts.  So, basically, we get to see both teams for the first real time.  EPA star to watch: Arizona State QB Taylor Kelly (24.3).

UTSA (1-1) at Arizona (2-0) - Arizona RB KaDeem Carey has 13.2 EPA on just 16 plays.  

Brent Blackwell compiles the NEPA rankings for  Follow Brent on Twitter by mashing the pretty button below.

Week 3 Saturday Preview - 3:30-6:30 Games

I went back and forth between calling these games the "mid-afternoon games" and "early evening games", and settled on the exact parameters instead.  It's boring, but it lets you find a specific matchup.

Cal Poly (1-1) at Colorado State (0-2) - The Rams had bowl hopes for this year, and if they expect that to come to fruition, they need to get a win in this one.

Nevada (1-1) at #10 Florida State (1-0), 3:30, ESPN

Nevada rush O (.01) vs. FSU rush D (.073)
Nevada pass O (.27) vs. FSU pass D (-.203)
FSU rush O (.08) vs. Nevada rush D (.273)
FSU pass O (.81) vs. Nevada pass D (.198)
FSU matches up well with Nevada.  FSU could have some weakness against the run down the road, but Nevada hasn't indicated yet that they will be the team to take advantage of that.  FSU should neutralize the Pack's passing offense, which has been decent so far.  Little of that will matter, though, because FSU will probably score so many points that even some defensive slips will be irrelevant.  EPA star to watch:  FSU QB Jameis Winston (26.3).

Georgia Tech (1-0) at Duke (2-0), 3:30, ESPNU

GT rush O (.55) vs. Duke rush D (-.290)
GT pass O (1.29) vs. Duke pass D (-.309)
Duke rush O (.19) vs. GT rush D (-.333)
Duke pass O (.18) vs. GT pass D (-.544)
The two ACC teams we really know the least about so far.  67% of their games have come against FCS competition and the other 33% was Memphis.  These teams could be good, but there's no way to know now.  If anything, that might be the reason to watch this game.  By virtue of whipping Elon and then taking a week off, Tech has video game type rate stats.  Seriously, averaging 1.29 points every time they throw the ball?  I would expect Tech to win this game, as I think Duke's offense may be the weak link of the four, but considering this is the first real game for either team, it's hard to really make any kind of educated guess.  EPA star to watch:  Georgia Tech QB Vad Lee (16.7).

#1 Alabama (1-0) at #6 Texas A&M (2-0), 3:30, CBS

Bama rush O (-.15) vs. aTm rush D (.157)
Bama pass O (-.17) vs. aTm pass D (.004)
aTm rush O (.35) vs. Bama rush D (-.112)
aTm pass O (.47) vs. Bama pass D (-.722)
Bama looked bad on offense against what looks like a very good Hokie defense.  Fortunately for Bama, A&M does not have a very good Hokie defense.  Bama looked great on defense against Logan Thomas.  Unfortunately for Bama, A&M doesn't have Logan Thomas.  There's really not a lot we've learned about these teams since the preseason, unless you're one of those people who thought Johnny Manziel would be distracted by off the field issues, in which case you've learned that such assumptions are not particularly intelligent.  Whichever team you thought would win this game in July, there's no reason to change your thoughts now.  All I'm really sure of is that it should be a fun one to watch.  EPA star to watch: Johnny Manziel (30.5).

Stony Brook (1-0) at Buffalo (0-2) - The Bulls just spent two weeks facing Ohio State and Baylor.  By comparison, Stony Brook will be like a high school scrimmage.
Delaware (2-0) at Navy (1-0) - This one is on CBS Sports Network if you want.  I see no reason an unaffiliated fan would.

Tennessee (2-0) at #2 Oregon (2-0), 3:30, ABC

Tenn rush O (.22) vs. Oregon rush D (-.118)
Tenn pass O (.32) vs. Oregon pass D (-.345)
Oregon rush O (.62) vs. Tenn rush D (-.137)
Oregon pass O (.34) vs. Tenn pass D (-.375)
Tennessee has been quietly decent over the first two weeks.  Sure, a win over Austin Peay isn't one to really hang your hat on, but one over Western Kentucky win is, to an extent.  Consider that Tennessee blew out Bobby Petrino's Hilltoppers.  Then consider that those same Hilltoppers were one week removed from handily beating one of Tennessee's conference rivals, Kentucky.  It was a game that slipped through the cracks in terms of national attention, but I think an important one for Butch Jones' tenure in Knoxville.  This was precisely the kind of team the Derek Dooley Vols would struggle with, as a 55-48 affair with 5-7 Troy last season will attest to.  That has little to do with this game other than me wanting to get out a few nice words about Tennessee before I predict them to lose by 30+ points.  Because they're going to lose by 30+ points.  EPA star to watch: Oregon QB Marcus Mariota (36.4).

Wagner (1-1) at Syracuse (0-2) - Everyone knows Georgia had the toughest opening two games of the season, and I'd put Buffalo as a close second.  Third might be Syracuse, 0-2 after dealing with Penn State and Northwestern.  Syracuse should be 2-2 when they hit their bye week to prepare for Clemson.
Ball State (2-0) at North Texas (1-1) - Ball State is top 25 in points per possession, and should continue to roll offensively to their 3rd win.
North Colorado (1-1) at Wyoming (1-1) - Wyoming nearly beat Nebraska and whipped Idaho, so they should win this one pretty easily.
Northern Illinois (1-0) at Idaho (0-2) - NIU escaped Iowa with a win.  Now Jordan Lynch gets to return to video game stat land with a pretty bad Idaho defense.

Iowa (1-1) at Iowa State (0-1), 6:00, Fox Sports 1

Iowa rush O (.16) vs. ISU rush D (.153)
Iowa pass O (-.17) vs. ISU pass D (.290)
ISU rush O (.10) vs. Iowa rush D (-.171)
ISU pass O (-.02) vs. Iowa pass D (-.070
Early season rivalries are usually great because there's so much optimism still alive for the season.  Unfortunately, the first two weeks have already been pretty crushing to the entire state of Iowa.  The teams are collectively 1-1 against FCS competition and 0-1 against the MAC.  Yikes.  Iowa State's defense has been porous, so I think the Hawkeyes have the edge here, but as watchable as I find rivalry games, this one could potentially test my limits.  EPA star to watch:  Iowa WR Kevonte Martin-Manley (8.2).

UCF (2-0) at Penn State (2-0) - This game should be televised on something other than the Big Ten Network.  UCF has outscored their two (admittedly bad) opponents 76-7, so they could give the Nittany Lions a good game.
Howard (1-1) at Old Dominion (0-2) - Howard's defensive seven is really, really impressive, so ODU should be on upset alert!  *Note - I made all of that up, because I don't care about this game, and neither do you.
#19 Washington (1-0) at Illinois (2-0) - The Huskies whipped Boise State, and I suppose they'll get a lot of mileage from that win.  Illinois nearly collapsed against Southern Illinois, an outcome that thrilled the hell out of the Illini's football Twitter account.  Washington should win at this "neutral site" game.  The neutral site is Chicago, which isn't particularly neutral.
Bethune-Cookman (2-0) at FIU (0-2) - The ghost of Mario Cristobal has proven quite effective over the first two weeks, and if FIU drops this one, we'll know the Cristobal-Turner effect is a real one.
Southern Utah (2-0) at Washington State (1-1) - WSU is flying high after beating USC, and they should win again.

Brent Blackwell compiles the NEPA rankings for  Follow Brent on Twitter by mashing the pretty button below.

Week 3 - Saturday Preview: The Early Games

More way-too-soon EPA analysis, with EPA breakdowns for every nationally televised game on Saturday, and some thoughts on the un-televised as well.

This first part will deal with the early slate of games, starting 12:00-3:00.

Bowling Green (2-0) at Indiana (1-1), 12:00, ESPNU

BGSU rush O (.10) vs. IU rush D (.194)
BGSU pass O (.30) vs. IU pass D (-.151)
IU rush O (.16) vs. BGSU rush D (-.160)
IU pass O (.48) vs. BGSU pass D (-.035)
Indiana's passing offense has been impressive so far, and their pass defense is working well.  Unfortunately, the rush defense has been so bad it didn't really matter; last week they allowed 444 ground yards to Navy on 70 carries.  That's not catastrophic, but it's not good.  If you tune in to this game, you'll be tuning in to watch Indiana's offense take on Bowling Green's D.  Indiana has been really good at offense this year, thanks to some healthy and maturity from players who have been in Kevin Wilson's system a few years.  Bowling Green has been really good at defense for the past 15 or so games.  Should be fun to see who wins.  EPA star to watch: Indiana QB Nate Sudfield - Sudfield is currently 5th in the country in EPA with 39.8.

#7 Louisville (2-0) at Kentucky (1-1), 12:00, ESPN

Louisville rush O (-.02) vs. Kentucky rush D (-.012)
Louisville pass O (.84) vs. Kentucky pass D (-.099)
Kentucky rush O (.20) vs. Louisville rush D (-.266)
Kentucky pass O (.30) vs. Louisville pass D (-.167)
EPA suggests that Louisville could have some problems running the ball effectively against Kentucky, but it won't matter, because every other facet of Louisville has been phenomenal this season.  The defense has been outstanding, and Teddy Bridgewater is carrying the offense.  EPA star to watch: Louisville QB Teddy Bridgewater - #1 in the country in expected points added, Bridgewater is the Heisman winner for the first two weeks.

Akron (1-1) at #11 Michigan (2-0) - Michigan should dominate.
Western Illinois at Minnesota (2-0) - Minnesota is 2-0 despite not really doing anything particularly well over the first two weeks.  WIU may be FCS, but they're 2-0 as well and it seems to be a good year for an FCS uprising.

Virginia Tech (1-1) at East Carolina (2-0), 12:00, Fox Sports 1

VT rush O (.16) vs. ECU rush D (.085)
VT pass O (-.26) vs. ECU pass D (-.098)
ECU rush O (-.06) vs. VT rush D (-.247)
ECU pass O (.31) vs. VT pass D (-.409)
Virginia Tech should win this game by virtue of defense alone.  Their strength, pass defense, matches up well with ECU's strength.  We saw what VT did to AJ McCarron.  They're stout, to say the least.  However, worth watching will be how Scott Loeffler runs this offense.  The Hokies offense has run 128 plays this year.  62 have involved Logan Thomas doing something particularly ineffective.  ECU has shown in 2 games this year that they're kind of decent against passers, and they're not very good at stopping the run.  Virginia Tech has proven to be pretty awful when they drop back to throw, and pretty good when they run.  This gameplan should be really easy to put together, right?  You would think.  However, you'd also think that, after a 77 yard TD run against Alabama, they'd have steered the focus more toward RB Trey Edmunds, who got only 20 carries.  EPA star to watch: ECU QB Shane Carden.  Currently #11 nationally in EPA with 34.8, Carden won't face a tougher defense all year.  If he looks good in this game, it's time to start paying serious attention to the man.

#16 UCLA (1-0) at #23 Nebraska (2-0), 12:00, ABC

UCLA rush O (.45) vs. Nebraska rush D (.012)
UCLA pass O (.54) vs. Nebraska pass D (-.035)
Nebraska rush O (.11) vs. UCLA rush D (-.050)
Nebraska pass O (.56) vs. UCLA pass D (.036)
Exploding scoreboard!  The lowest scoring game either team has been involved in so far this year was Nebraska's 56-13 win over Southern Miss.  Tune in for points, and lots of them. EPA star to watch:  UCLA QB Brett Hundley, who is 25th in total EPA despite playing in only one game this season.

Tulsa (1-1) at #14 Oklahoma (2-0), 12:00, ESPN2

Tulsa rush O (-.14) vs. Oklahoma rush D (-.165)
Tulsa pass O (-.09) vs. Oklahoma pass D (-.293)
Oklahoma rush O (.20) vs. Tulsa rush D (.072)
Oklahoma pass O (-.37) vs. Tulsa pass D (-.264)
It's surprising how little punch the Sooners have had throwing the football this year.  In two games, the Sooners have 243 passing yards on a whopping 54 attempts.  54 attempts for 243 isn't that good if you're a running back.  If you're passing, it's dreadful.  Oklahoma is averaging 6 yards per play, and .20 expected points per play, rushing.  That's not good news for Tulsa, who excels against the pass but has struggled a bit against the run.  Tulsa was held to 7 points by a good Bowling Green defense, and then looked decent but not great against Colorado State.  The Sooner D has picked up the slack caused by the offense so far, so Tulsa could be in for a long day when they have the ball.  Expect a comfortable Sooner victory. EPA star to watch: Oklahoma RB Brennan Clay (7.0).

Georgia State (0-2) vs. West Virginia (1-1) - Mountaineers should win handily.

#5 Stanford (1-0) at Army (1-1), 12:00, CBS Sports Network

Stanford rush O (.15) vs. Army rush D (-.019)
Stanford pass O (.38) vs. Army pass D (.316)
Army rush O (.24) vs. Stanford rush D (-.079)
Army pass O (-.30) vs. Stanford pass D (-.025)

This one sets up to be an easy Stanford win, with pretty much every matchup going their way.  EPA star to watch: Stanford QB Kevin Hogan (12.0).

Southern Miss (0-2) at Arkansas (2-0) - USM's losing streak should reach 15 games.
ULM (1-1) at Wake Forest (1-1) - ULM offense vs. Wake D could be decent matchup.
New Mexico (1-1) at Pittsburgh (0-1) - New Mexico leads country so far in average time of possession per drive.  Pitt should still win.
Eastern Michigan (1-1) at Rutgers (1-1) - Knights will win.
Fordham (2-0) at Temple (0-2) - Fordham isn't a pushover, but Temple's hungry for a win and should get one.
Fresno State (2-0) at Colorado (2-0) - Mike MacIntyre's 2-0 start feels like a big win already for the Buffs.  This will probably be loss #1, but they seem to have gotten themselves a good coach.
Youngstown State (2-0) at Michigan State (2-0) - Michigan State gets yet another matchup they should win by a lot, but yet again they will win by 8-14 points, and the difference will most certainly be provided on defense.
Boston College (2-0) at USC (1-1) - BC has been better than advertised and USC has been worse.  Don't sleep on this one, and it's kind of a shame it's not televised.

Brent Blackwell compiles the NEPA rankings for  Follow Brent on Twitter by mashing the pretty button below.

Thursday, September 12, 2013

Week 3 Preview - The Weekday Slate

Week 2 was such an uninspiring slate of games, I couldn't muster a preview.  Week 3, on the other hand, looks like it could be a good one.  Let's delve right in.  And even better, we have Expected Points Added numbers to use!  I'm going to be breaking up the preview into sections, and we start with the pre-Saturday games.


  • Troy (2-0) @ Arkansas State (1-1), 7:30:
    Troy Rush O (.18) / ASU Rush D (-.050)
    Troy Pass O (.52) / ASU Pass D (-.109)
    ASU Rush O (.15) / Troy Rush D (-.240)
    ASU Pass O (.20) / Troy Pass D (-.181)

    In the preseason, I didn't think this would be a key SBC game, but the old Troy seems back.  This should be a good game.  ASU has considerably worse looking defensive stats, but they've played Auburn, while Troy's toughest opponent has been UAB.  It's still early to put too much stock into EPA (I'll feel better about it after 4 games or so), but it suggests the mismatch of this game will be Troy's Rush D overpowering the Red Wolves' running game.  One player to watch: Troy QB Corey Robinson, who is currently 14th in individual EPA with 33.4 points added over the first two weeks.
  • Tulane (1-1) @ Louisiana Tech (1-1), 7:30:
    Tulane Rush O (.05) vs. LT Rush D (.026)
    Tulane Pass O (.11) vs. LT Pass D (.052)
    LT Rush O (-.03) vs. Tulane Rush D (-.190)
    LT Pass O (-.02) vs. Tulane Pass D (.130)

    Tulane was outplayed by Jackson State on most plays but still won by 27 thanks to a whopping 6 turnovers going in their favor.  Last week they lost to South Alabama.  Tech had the toughest opponent so far for either, losing to NC State in the opening week.  While EPA suggests Tulane matches up well, I am skeptical.  Tech should be much better in this CUSA conference game.  Player to watch:  Tulane WR Ryan Grant (15.6 EPA).
  • #24 TCU (1-1) @ Texas Tech (2-0), 7:30:
    TCU rush O (.22) vs. TTU rush D (-.179)
    TCU pass O (.01) vs. TTU pass D (-.020)
    TTU rush O (.11) vs. TCU rush D (.056)
    TTU pass O (.33) vs. TCU pass D (.001)

    Both schools have wins over an FCS opponent.  Tech's other game was against SMU, while TCU opened the season with LSU.  Understandably, TCU's numbers might be slightly misleading.  This should be a good game, and a key early Big XII game.  The conference looks pretty wide-open.  With the Longhorns firing coordinators, the Sooners limping to 16 points against West Virginia (yes, THAT West Virginia), and the Cowboys under all kinds of fire off the field, almost every conference game could play a huge role.  This one's no different.  Tune in to see a good offensive mind, Kliff Kingsbury, match wits with two great defensive minds, Gary Patterson and Dick Bumpas.  Player to watch: Texas Tech QB Baker Mayfield.  Mayfield is currently 2nd in the nation in EPA with 48.3 points added over the first two games.  Oh, and he's a true freshman.  Oh, and he's a walk-on.  You read that right.


  • Air Force (1-1) @ Boise State (1-1), 8:00
    AF rush O (.15) vs. BSU rush D (.001)
    AF pass O (.00) vs. BSU pass D (.187)
    BSU rush O (-.02) vs. AF rush D (.035)
    BSU pass O (.16) vs. AF pass D (.423)

    BSU's early whipping at the hands of Washington has given other MWC teams some hope, and this game could be a key game in deciding the Mountain Division.  Startling is Air Force's .423 EPA per play allowed on passing plays, and they have Chuckie Keeton to thank for that.  Colgate's Gavin McCarney wasn't bad - 27 plays for 143 and 1 TD - but Keeton went ballistic on the Falcons D, racking up 360 and 5 TD on 40 plays.  McCarney is his conference's reigning player of the year, and we know how good Keeton can be, so Boise might have the worst opposing QB Air Force has faced so far.  That number can't really go anywhere but down.  Player to watch: Boise State QB Joe Southwick (15.7 EPA).

Brent Blackwell compiles the NEPA rankings for  Follow Brent on Twitter by mashing the pretty button below.

Tuesday, September 10, 2013

Week 2 updates

What is NOT updated:
Picks - I'll start posting picks next week
Conference stats - Unless someone convinces me otherwise, this will be relegated to the dustbin of history
Team Summaries - Technical issue. Hope to have some time next week to fix it.
Teams: Team, Team: Summary - see Team Summaries
Teams: Special Teams - I don't actually have anything to update yet

Also, keep in mind that any opponent-adjusted state ("+") is suspect at this point. Not until week 4 or 5 do I put any confidence in these numbers.

And Georgia State is excluded for now. It's not because I don't love them, but I don't love them enough to make the necessary adjustments just yet for a team that is completely irrelevantly.

Monday, September 9, 2013

The Preseason Bias

There is popular consensus that the preseason rankings are somewhat irrelevant.  Within a week or two, a number of the AP Preseason Top 25 will lose to unexpected teams and fall from the rankings.  This season alone we've already seen Florida upset by Miami, Boise State to Washington, Texas to Brigham Young, and Oregon State to FCS Eastern Washington.  After only two weeks of play, 7 of the preseason top 25, adding Georgia and Notre Dame to the list, have experienced an upset.

In theory, the preseason polls should be nothing more than a sideshow, as polling respondents will update their information as games are played throughout the season.  The logic goes, by the end of the season, the respondents are evaluating the total body of work for a team based upon careful observations of each team's behavior.

However, the preseason polls represent something very important: the polling respondents' expectations of how teams should perform.  In other words, respondent bias.  If Craig James ranks USC highly in the preseason, he believes that USC will be good.  Hypothetically, if USC and Baylor than provide equal evidence of their quality by the end of the year, but Craig James did not expect Baylor to be good and did not include them in his preseason top 25, he will be predisposed to rank USC higher for no other reason than his preseason faith in Lane Kiffin.

To illustrate the presence of this bias we take three pieces of evidence: 1) the preseason AP poll, 2) the final regular season AP poll, and 3) the final regular season Network Rankings which evaluate the total body of work based on wins and losses (more here, but any evidence based ranking should work).  I've collected the data and calculated the Network Rankings for the 2012, 2011, and 2004 seasons.  To demonstrate preseason bias, we expect to find that where there is dramatic divergence between the final regular season AP Poll and the Network Rankings, the preseason poll is influencing the outcome.

For example, in 2012, Boise State was ranked 19 in the final regular season AP poll and 40 in the Network Rankings.  Because it was included in the preseason top 25, polling respondents ranked them far more highly than they deserved.  Alternatively, we find that in the case of Texas A&M, unranked in the preseason, the Network Ranking is 8 and the final regular season AP poll 9.  A difference of only 1.

To demonstrate this effect statistically, I regress the final regular season AP poll on the preseason AP poll and the Network Rankings.  If the AP poll is simply capturing perceptions of a team's body of work, than the variation in final AP rank should be mostly accounted for by the Network Rankings.  However, more likely, the final AP poll is a combination of a polling respondents' perceptions of the actual evidence combined with the respondents' bias of how a team should perform represened by the preseason AP poll.

Indeed, in this (very) simple analysis, the preseason polls do appear to introduce bias.  As we would hope, the final regular season poll is contingent upon the evidence, and the Network Rankings are positively and significantly correlated with a team's final AP rank.  However, despite controlling for the Network Rankings, the preseason poll is also significant, increasing a team's end of season rank, on average, by 0.42.  For those interested in the statistical details, the r-squared is .49, both variables are significant at .01, and in coding the AP polls because they only rank the top 25, you can code the unranked teams as 26 (or code them as 0 and invert the ranking values), or 120, and you get about the same results.

Either the preseason poll represents some information we just can't pick up by evaluating the evidence that Craig James, in his sage wisdom, can, or the polling respondents are biased at the outset and maintain some portion of that bias throughout the season.  

So who benefits from this preseason bias?  Not the SEC, likely because the SEC usually provides actual evidence that meets even the most lofty expectations.  Here's a list of the top 10 offenders in the three seasons of data:

Team Year Network Preseason Final Rank
Boise State 2012 40 24 19
Florida State 2012 31 7 12
West Virginia 2011 35 24 23
Louisville 2012 33 25 21
Michigan State 2011 28 17 17
Oregon 2011 15 3 5
Michigan 2004 21 8 13
Wisconsin 2011 18 11 10
TCU 2011 26 14 18
Stanford 2012 12 21 6

Sunday, September 8, 2013

What we learned, week 2

1) The FCS picked up 4 more wins against FBS opponents and lost 2 others by a field goal or less. They also got another win against a ranked team.

What's that, Washington St. is FBS? Oh, then 3 wins against FBS opponents. And those 3 opponents were UMass, Georgia State and Western Michigan. On second thought, I'm not impressed. 

1a) Kiffin is a goner.

2) ESPN's low watt star, Jadeveon Clowney, backs up his talk on the field. Aaron Murray showed how scared he was when he threw 6 incompletions on 23 attempts and managed only 309 yards and 4 TDs, and Georgia was limited to a measly 41 points. How do you stop Clowney, who has 5 tackles and a sack through 2 games? You block him, apparently.

2a) I was tempted to insert a joke here about Clowney needing an imaginary girlfriend that he could sacrifice to resurrect his Heisman campaign, but I won't.

2b) This game reminded us of a basic principle that I've posted 15 times since the Clowney hype train got rolling: quarterbacks (like Aaron Murray) are many times more important to a team's success than any of their teammates.

3) We didn't learn if BYUs offense is good or bad, but we did learn that UT still hasn't learned to defend the run. Good news for the Horns is that they still play in the Big 12.

3a) USC and UT in the top 10, Mr. Steele? Good call.

3b) I would think Manny Diaz has to be Mack Brown's last sacrificial lamb.

3c) Baylor will win the Big 12, and they'll run (and pass) away with it

5) Between partying with Drake, signing some stuff, sleeping in at passing camps, and complaining about a parking ticket, Manziel didn't forget how to play football. In fact, he's really good at football. He threw for 400 yards, made a defender fall down without touching him and made another fall down by driving his shoulder threw the defender's chest. A win against Alabama and ESPN will be singing his praises from the mountain tops.

6) Bobby Petrino is still being punished for his crimes against Arkansas and the SEC. Six turnovers in the first half? You won't win a lot of games like that.

Sunday, September 1, 2013