Monday, March 02, 2009

Big 12 2008 SDPI

One of my favorite set of posts from the past two offseasons has been the SDPI recap/early preview. Don't know what SDPI is? It stands for Standard Deviation Power Index and is a tool Eddie Epstein used in his book Dominance to rate pro football's best teams. The basic idea is to look at how far above or below average (by standard deviations) a specific team is relative to their conference brethren. Since each team plays the same number of conference games, it can give us a good idea about who the best team was within the conference. However, it cannot tell us which conference is better. But the purpose of these posts is not to determine which conference is superior, but rather project ahead which teams in a conference will be contenders or also rans in the upcoming season. In the first post on SDPI two years ago, I calculated SDPI based on points scored and allowed within conference play. Last season I used points scored and allowed as well as yards gained and allowed. This season, I'm sticking with yards only. The yardage version of SDPI has a better correlation with future performance than points, and including both last season made the post seem (at least to me) quite muddled. Of course, this is by no means, the end all be all rating system, but it can give us an idea of which teams will improve and decline in 2009. We've already looked at the ACC, Big East, and Big 10 so now we'll take a look at the Big 12.

If you want the meat of the article, skip this next paragraph as it just gives an example of how SDPI is calculated. The mean yardage for and against for all Big 12 teams in conference play (championship game excluded) was 3450.47 yards. The standard deviation for yards gained was 604.40. The standard deviation for yards allowed was 448.37. Kansas gained 3283 yards in conference play and allowed 3577 Their offensive SDPI was -0.28 = ([3283-3450.47]/604.40). Their defensive SDPI was -0.28 = ([3450.47-3577]/448.37). Their total SDPI was -0.56 which ranked 7th in the conference.

To refresh your memory, here are the 2008 Big 12 Standings.

Now here are the 2008 SDPI Standings sorted by total SDPI, with conference rank in offense, defense, and total SDPI in parentheses.

I'm a man who gives credit where credit is due, and it looks like the BCS got it right. Despite their head-to-head win over Oklahoma, in the other seven Big 12 conference games, the Sooners were simply more dominant than the Longhorns. Statistically, they were the best team and I have no problem with them earning the Big 12 Championship bid, and later the BCS National Championship Game bid. The Sooners offense was historically prodigious in 2008. In conference play, the second best offense (Texas Tech) was was nearly three quarters of a standard deviation worse, and the third best offense (Nebraska) was more than an entire standard deviation below the Sooners. Vehement SEC partisans might point out that Florida shut-down the mighty Oklahoma attack, but while the Sooners did gain a season-low 363 yards against the Gators, if not for two red-zone gaffes, they would have been in a great position to win the national title. Those same SEC partisans might also point out that Oklahoma put up those yardage and point totals against a set of Big 12 defenses that were not exactly rugged. It also bears mentioning that the other eleven teams in the conference also got to play those defenses. The Sooners separated themselves, at least offensively, from their conference brethren like few teams before them. Not surprisngly, as three teams in the division finished 7-1 in league play, the South was once again superior to the North as the top three and four of the top six teams resided in that division. What may come as a shock to some folks is that Nebraska rates out as the top team in the North. After being bludgeoned by Missouri on their homefield, on national television, by 35 points, the Huskers would lose only twice more the rest of the season--by six at Texas Tech and in embarrassing fashion at Oklahoma. Meanwhile Missouri would prove their pretender status, at least on the national scene, by dropping a home game to Oklahoma State, being blown out at Texas, and losing at a neutral site to arch-rival Kansas. Missouri certainly owned the Huskers in their head-to-head meeting, but over the rest of the Big 12 season, Nebraska was the better team.

Best Offense: Oklahoma 1.86
The most remarkable fact about the Okahoma offense is not how many yards and points they put up, but how many negative plays they avoided. Sam Bradford threw four interceptions and was sacked six times in eight Big 12 games. If we include the Big 12 Championship Game against Missouri, those numbers do not change. Two of those interceptions and three of those sacks came in the loss to Texas. I hope Mr. Bradford took his offensive linemen out for some quality steaks after he won the Heisman.

Worst Offense: Colorado -1.64
Think Dan Hawkins regrets leaving the cozy confines of Boise, Idaho?

Best Defense: Texas 1.30
Maybe this Will Muschamp character knows a thing or two about defense. Not surprisingly, the only game where the Longhorns really struggled defensively was the upset loss in Lubbock where the Red Raiders gained 579 yards.

Worst Defense: Kansas State -1.82
In other years, a performance like this could be enough to get the greatest coach in program history to make a comeback...

Hardest Schedule (based on cumulative SDPI of opponents): Baylor 5.29
The Bears played in the more arduous South division, meaning they had to play Oklahoma, Texas Tech, Texas, and Oklahoma State. In two of their three games against the North, they drew the best teams from that division--Nebraska and Missouri. All in all, they faced the top six teams in the conference.

Easiest Schedule (based on cumulative SDPI of opponents): Iowa State -4.58
If I wanted to fan the flames of the Gene Chizik hiring fiasco, I might remind you that the team with the easiest conference schedule went winless in Big 12 play. Obviously, they played in the weaker division, and in their games against the South, they managed to avoid the triumverate of Oklahoma, Texas Tech, and Texas.

Entire Schedule Strength (hardest to easiest)
Baylor 5.29
Kansas 3.60
Kansas State 1.34
Oklahoma State 0.86
Texas 0.27
Texas A&M 0.17
Texas Tech -0.25
Oklahoma -0.46
Nebraska -1.72
Colorado -1.91
Missouri -2.77
Iowa State -4.58

Looking ahead to next season, the prohibitive favorite should be...

North: Nebraska
Projecting Nebraska as the favorite to take the North in 2009 is based on their good standing in 2008 (statistically the best team in the division with the third ranked offense and second ranked defense) and because frankly, its hard to come up with a convincing argument for any of the other five teams. Statistically, Kansas State was the league's worst team last season, and while they have a good shot at improving this season (more on that later), a division title may be the coaching job of Bill Snyder's career. Iowa State may finally have a coach who wants to to be their, but they were winless in the Big 12 last season despite owning the league's easiest schedule. Colorado has yet to find an offense in three seasons under Dan Hawkins, and actually regressed in Year Three after making significant progress in Year Two. Missouri, the two-time defending division champ, loses one of the best quarterbacks they've ever had (Chase Daniel), their number one playmaker (Jeremy Maclin), one of the nation's best tight ends (Chase Coffman), an over-looked senior receiver (Tommy Saunders) and two starting offensive linemen. On the other side of the ball, the Tigers will be without the services of alliterative defensive end Stryker Sulak who led the team with 10.5 sacks last season. Along with Sulak, six of the top eight tacklers from last season's mediocre defense are gone. Fortunately for the Tigers, game-changing linebacker Sean Weatherspoon will be around for his senior year. It looks like a rebuilding year in Columbia, and while rebuilding may mean the Texas Bowl instead of 4-8, it likely precludes a division title. That leaves Kansas as the likely only legitimate contender the Huskers will have to deal with. Kansas returns a great deal of talent on the offensive side of the ball with quarterback Todd Reesing, running back Jake Sharp, and receiver Kerry Meier (the best hands in college football) all back for their senior seasons. The Jayhawks also have a pair of talented junior receivers in Dezmon Briscoe (led the team in catches, yards, and touchdowns last season) and Jonathan Wilson. However, despite the glut of returning skill position players, Kansas may have some issues along the offensive line as three or four starters (depending on your definition) depart. The Jayhawks will also be without linebacker James Holt who led the team in tackles and sacks last season, as well as five other defensive starters. Oh yeah, and there's also the matter of the schedule. While the Jayhawks do host Nebraska, they have only three conference home games, and one is against Oklahoma. Still repaying the football gods for their tranquil 2007 conference schedule, the Jayhawks must travel to Colorado (one win at Boulder since 1995), Kansas State (one win in Manhattan since 1989), Texas Tech, and Texas. Ouch. I don't doubt that the Jayhawks can win some of those games, but they can't win enough to take the division. That leaves one team. Nebraska. Though it was overshadowed, the job Bo Pelini did in resurrecting the Nebraska defense in his first year on the job was amazing. In 2007, the Huskers finished dead last in the Big 12 in yards allowed per game in conference play, permitting 512 yards per game. In 2008, the Huskers finished second in yards allowed per game, permitting only 365 per contest. The defense also improved as the season wore on as the table below illustrates. Those first six games include clashes with powerful offenses like Missouri and Texas Tech, but also games with San Jose State (113th in total offense), New Mexico State (62nd in total offense), and Virginia Tech (103rd in total offense). In their last seven games, the Huskers tussled with Oklahoma and faced only two teams ranked lower than 44th in total offense (Colorado at 95th and Clemson at 87th). The primary difference between the first and second half of the season was the pass rush. In their first six games, the Huskers managed only 11 sacks. In their last seven, they totalled 24, including 19 in the last four games. The bulk of the players who accumulated those sacks will be back in 2009, including defensive tackle Ndamukong Suh, who led the team with 7.5 sacks. Another positive indicator going forward is that Nebraska played very good defense despite almost never forcing turnovers. The Huskers forced only nine turnovers in eight conference games, tied with Iowa State for the fewest forced turnovers in Big 12 play. The good news for the Huskers is that turnovers, while the result of skill and practice, are also highly random. A team that forces a minuscule number of turnovers one season will tend to force more the next season. Of course, the Huskers do have a few areas they will need to address to actually win the North in 2009. Foremost among them is quarterback. Joe Ganz tossed 25 touchdowns last season and posted the 14th best passer efficiency rating in the nation. If the Huskers can just get competent quarterback play, they should take the North crown.

South: Oklahoma or Texas
After my 'War and Peace' breakdown of the North, here's an abriged version on the South. Oklahoma or Texas have won the South every season since 1999. Texas Tech had their chance last season, and almost broke through. Unfortunately, the two biggest stars of that team, quarterback Graham Harrell and receiver Michael Crabtree, are gone. I'm sure Mike Leach will find an air apparrent under center and lead the Red Raiders to eight or nine regular season victories, but Oklahoma or Texas will represent the South in the Big 12 Championship Game.

The team(s) you should be buying are...

Iowa State and Kansas State
New head coach Paul Rhodes is not exactly stepping into an ideal situation, but the good news is, it only takes one win to improve upon last season's dismal campaign. Though they were winless last season (in conference play), the Cyclones actually rated out better than both Texas A&M and Kansas State (although those two teams did beat them on the field). The Cyclones were far from a 'good' or even 'mediocre' team (eighth in offense and tenth in defense in the Big 12), but they were certainly better than most winless squads. They were undone by a poor record in close games. Three of their eight conference games were decided by a single score (Kansas, Colorado, and Kansas State). The Cyclones lost each of those games by a combined 14 points. A few lucky bounces here or there and Iowa State would have been another middling 2-6 team, rather than one that bore the stigma of a winless conference season. Along with the likelihood of a little better luck in 2009, the Cyclones also bring back their starting quarterback (Austen Arnaud), running back (Alexander Robinson), eight of their top nine receivers, and seven starters on defense. Once again the schedule is very inviting as the Cyclones avoid Oklahoma, Texas Tech, and Texas from the Big 12 South. They also host Baylor and Colorado, as well as getting Kansas State at a neutral site in Kansas City. They won't be anywhere close to the best team in the division, but with North Dakota State, Kent State, and Army on the non-conference slate (as well as the annual showdown with rival Iowa), a bowl bid is at least within the realm of possibility. Elsewhere, in Manhattan, Kansas, the architect of the greatest turnaround in college football history returns to try to right the program once again. Let's face it, Ron Prince was a terrible hire. When he was the offensive coordinator at Virginia, did anyone ever stop and say to themselves, 'Wow. We have to get this man to coach our team.' I doubt it. Especially after this game. Much like two decades ago, Snyder has a long way to go to get this team back on track, but luckily for him the schedule may give him the chance to at least get the Wildcats back to the postseason. To be fair, the road trips are nightmarish (Texas Tech, Oklahoma, and Nebraska in conference play, along with a non-conference trip to UCLA), but the home conference schedule includes dates with Texas A&M (another statistically awful team last season), Colorado, and Kansas (Jayhawks win in 2007 was their first in Manhattan since 1989). Besides the schedule, the other good news for Kansas State is that they were partucularly young on defense last season. The top six tacklers from last season are back, and it would be hard to envision them being any worse. Three Big 12 teams gashed the Wildcats for over 600 yards last season and three others topped the 500-yard mark. Defenses that bad that are not coached by complete morons tend to get better (see Nebraska from last season as a luminating example). On offense, Snyder must find a quarterback to replace Josh Freeman, who departed for the NFL. However, even with Freeman, the Wildcats were only ninth in the conference in total offense. And against Big 12 defenses, well he wasn't that good.And a number of prognosticators have this guy going in the first round? Am I missing something here? He played below average ball in a conference not known for its defense. OK. Back to the matter at hand. Whoever (Iowa State or Kansas State) wins the showdown in Kansas City on October 3rd, will have a great shot at getting to a bowl game and wiping the taste of an awful season from their palates.

The team(s) you should be selling are...

Missouri and Texas Tech
I've already examined the myriad of personnel losses Missouri is facing heading into 2009, but I can't emphasize how much Missouri will miss those skill position players. A bowl bid is a distinct possibility, but definitely not assured. Texas Tech will also be breaking in new skill position talent, and it should also be noted that the Red Raiders were very fortunate in one-score games last season. Their victories over Nebraska, Texas, and Baylor were by a combined 19 points and were all decided late. If the Red Raiders were a little less fortunate, that 7-1 breakthrough is just another ho-hum 5-3. The home schedule outside of Oklahoma (and who knows, the Red Raiders have beaten the Sooners the past two times in Lubbock) is very winnable, but if Texas Tech breaks even on the road in conference play (Texas, Nebraska, Oklahoma State, and Baylor), 2009 will have to be considered a success.

The team(s) you should be holding are...

Believe me, no one is rooting for the Bears more so then yours truly. However, Baylor's first bowl bid since 1994 will probably have to wait until at least 2010. For starters, despite showing substantial improvement under Art Briles, the Bears were still only the fifth best team in their division. Quarterback Robert Griffin certainly gives the casual fan a reason to watch Baylor. I know as a Wake fan, I am certainly glad Kirby Freeman started the opener in Waco. Still, even with Griffin, the Bears finished eleventh in the Big 12 in offense, beating out only Colorado. The bad news for Baylor is that they only managed two conference wins despite an outstanding turnover margin of +11 in Big 12 play (second to Oklahoma at +14). The primary reason for that great turnover margin was the fact that Baylor had only five turnovers in their eight conference games. Part of that was the efficient play of Griffin, who threw only three interceptions in conference play, and part of it was the fact that Baylor only lost six of their twenty fumbles on the season. Griffin will probably throw more interceptions in 2009 (see Colt McCoy in 2007) and the Bears will certainly fail to recover 70% of their own fumbles. More turnovers are bad for any team, but especially for teams operating at a distinct talent disadvantage like Baylor. Another factor, and this will be the case for the foreseeable future, is the schedule. The Bears four conference home games are against Oklahoma State, Nebraska, Texas, and Texas Tech, all of which are likely bowl teams. The road schedule is not much easier with trips to Oklahoma, Iowa State, Missouri, and Texas A&M on the horizon. The Bears did beat Iowa State and Texas A&M in Waco last season, but the Bears have just two road wins since joining the Big 12 in 1996! The non-conference schedule will give the Bears a chance at a fast start (Wake Forest, Connecticut, Northwestern State, and Kent State), but barring multiple league upsets, Baylor will be home for the holidays yet again.


Anonymous said...


I don't get how this works. I see that SDPI is a good indicator of strengths and weaknesses LAST year.

But you say "This season, I'm sticking with yards only. The yardage version of SDPI has a better correlation with future performance than points,"

How is that so? In your predictions analysis for 2009, you don't mention yardage SDPI at all. When you mention teams that we should be buying and selling, you discuss mostly roster changes or schedule strength. I don't see yardage SDPI being applied. If the 2008 yardage SDPI is a good predictor of future strength, then it's telling us that the Big 12 team wins and losses will be the same as last year, right?

Again, I see how this gives insight for last year, but I don't see yardage SDPI, or any other SDPI for that matter, predicts future performance.

matt said...

I beg to differ. I mention where a team finished on either offense or defense numerous times (this is a reference to the table near he beginning of the post that liasts their SDPI finish).

There's more to analyzing a team than their overall record. Yardage is a good indicator of how a team performed on a down-to-down basis.

To summaraize my 4 biggest points from this post were as follows:

1. Nebraska was the best team in the North last season (despite not winning the division) and should be favored to win in 2009.

2. Iowa State, despite a winless season, was far from the worst team in the division. When they challenge for a bowl game in 2009, don't be shocked.

3. Kansas State, despite an 'NFL' quarterback, was below average on offense. His departure will not hurt the team as much as expected.

4. Baylor, despite the play of Robert Griffin, was significantly below average on offense. Their high turnover margin is not likely to be repeated and they will not improve in Art Briles' second season.

Plus not using schedule strength and roster attrition/addition would be just plain stupid. SDPI is a better predictor of future performance than won/loss record, but of course it is far from perfect. That is why we watch sports: the outcome is undecided.