Monday, February 22, 2010

Big East SDPI

Moving right along through major college football, we come to our second conference, the Big East. If you want to remember how clairvoyant I was last year at this time, here's the link to the Big East SDPI review from 2009. Before we jump into this edition of SDPI, let me go ahead and warn you to be sure you are sitting down. One of the results will certainly shock. I almost dropped my abacus, and certainly checked and rechecked my work when I saw it.

As usual, this first paragraph will explain how SDPI is calculated. So if you want the meat of this article skip on down. In the 2009 Big East regular season, conference play only, the average Big East team gained and allowed 2560 yards. The standard deviation for yards gained (offense) was 476.04 yards. The standard deviation for yards allowed (defense) was 335.60 yards. Connecticut gained 2957 yards and allowed 3208 yards. Their offensive SDPI was 0.83 = [(2957-2560)/476.04]. Their defensive SDPI was -1.93 = [(2560-3208)/335.60]. Their total SDPI was -1.10. This number ranked 6th in the Big East.

To refresh your memory, here are the 2009 Big East standings.Now here are the 2009 Big East SDPI standings. The standings are sorted by total SDPI with ranking for each category (out of 8 teams) in parentheses.Obviously, the huge shocker I was referring to earlier is the ranking of the Orange from Syracuse. Thanks to a superb defensive ranking, the Orange rated out as the 3rd best team in the league despite winning only a single conference games all season. How is that possible? Is the ranking system screwed up, or was Syracuse really better than you thought they were? I'm inclined to say that the Orange were a little better than you thought, especially on defense. On offense, the Orange were nothing special, ranking 6th in the league, ahead of only Louisville and Rutgers (!? more on the Knights later). In addition to this, the Orange offense under former Duke hoopster Greg Paulus repeatedly put the defense in harms way via the turnover. The Orange had the worst turnover margin in Big East play (see chart below) by a pretty healthy margin. Paulus threw 10 picks in 7 Big East games (versus just 6 touchdowns). Here are just a few examples of how the offense (and occasionally the special teams) hung the defense out to dry. In their Big East opener, the Orange held the South Florida Bulls to 333 yards of offense, but also committed 7 turnovers, including an interception that was run back for a touchdown in a 14 point loss. In mid-November against fellow cellar dweller Louisville, the Orange held the Cardinals to 151 yards of offense, but botched an extra point and gave up a huge 44-yard punt return late in the game that set up the winning score to allow the Cardinals to escape with a 10-9 win. In their season finale against Connecticut, the Orange outgained the Huskies by over 100 yards (489-377), but allowed Connecticut to return both a kickoff and a fumble for a touchdown in a 56-31 loss. Later on, we'll discuss how likely it is that Syracuse can maintain their solid defensive play in 2010. Elsewhere in the Big East, Cincinnati and Pitt were almost as close statistically as they were on the field in a game the Bearcats won thanks to a missed Pitt extra point. SDPI likes the Panthers just a hair more, but statistically speaking, its a wash. One other surprise in the SDPI rankings is the lowly position occupied by Rutgers. Only Louisville ranks lower in SDPI.

Conference Superlatives

Best Offense: Cincinnati 1.71
Brian Kelly's Bearcats moved the ball at will against everyone in the Big East. They opened the season with a 564-yard and 45-point performance at Rutgers and continued their domination no matter who played quarterback. The 2nd best offense, Connecticut, was less than half as prolific as the Bearcats.

Worst Offense: Rutgers -1.17
While they were not historically bad (the -1.17 is the highest lowest offensive rating for any team in 2009--hooray), the Knights did lay a few eggs in 2009. They netted 130 yards of offense in their blowout loss to Syracuse, and managed only 218 yards in the regular season finale against West Virginia.

Best Defense: Pitt 1.28
Not surprisingly, in their 2 losses, the Panthers had their worst defensive showing against Big East foes. They allowed 370 yards per game to West Virginia and Cincinnati, but only permitted 278 yards per game in their 5 Big East wins.

Worst Defense: Connecticut -1.93
Part of the culprit for this low ranking is the unholy 711 yards they allowed to Cincinnati. Still, even without that monstrosity weighing them down, the Huskies would have been the worst defense in the Big East (but it would have been a little closer).

What's in Store for Next Year?
So now I want to take a closer look at Syracuse. As previously mentioned, the Orange boasted a significant statistical jump on defense. How likely are they to maintain and perhaps build on that performance? I have calculated SDPI for the seasons from 2005-2009, so I may be able to answer that question. The jump Syracuse experienced on defense in SDPI from 2008 to 2009 was over 2 standard deviations (2.02). They went from 7th (2nd to last) in the conference to 2nd. How often has this occurred in the Big East in the recent past? Probably more than you would think. 3 teams have seen their defensive SDPI jump at least 1.5 standard deviations from one season to the next. In fact, it seems to happen at least once a year. The table below lists those jumps, along with conference ranks in defense for each season. As you can see, the leap made by Syracuse was hardly unprecedented. Cincinnati and Pitt both made larger single season leaps in 2006 and 2007 respectively. Now the big question is, what happened in the season after the large improvements? And the answer in the table below is not exactly encouraging for Syracuse fans. Cincinnati and Pitt experienced drops in performance of a little more than a single standard deviation, while Connecticut fell off the map. The Huskies boasted the league's number one defense in 2008 and the worst defense in 2009. One rudimentary way to guesstimate how Syracuse will do on defense in 2010 is to look at how many starters they will return and compare it to the number of starters these other defenses brought back. While looking at returning starters is inherently flawed, as defenses must deal with injuries and different personnel groupings over the course of a season, it can give us a reasonable place to start. Cincinnati returned 7 starters from their 2006 defense, Pitt returned only 5 from their 2007 defense, and Connecticut returned 6 from their 2008 defense. While we won't know until late in the summer how many starters Syracuse will return in 2010, only 2 of their top-12 tacklers were seniors. On the one hand, while this means the Orange will have a great deal of experience returning in 2010, looking at the other teams that improved significantly on defense in one season means the Orange were likely playing over their heads in 2009 and regression should be imminent. Still, if the offense can cut down on the turnovers, the Orange stand a very good chance of returning to postseason play for the first time since 2004. What about the other team in the New York/New Jersey area, the Scarlet Knights from Rutgers? While the Knights played in their 5th consecutive bowl (4th straight win), they were far from a dominant Big East outfit. In league play they were outscored by 2 points on the season and outgained by over 500 yards. The Knights were fantastic at winning the turnover battle, finishing +4 within the league and +20 overall. The Knights also scored 9 non-offensive touchdowns, which helped them to win games despite their offensive struggles. The Knights were breaking in a freshman quarterback in 2009, so the offense will likely be able to move out of the cellar, but on the whole the Knights enjoyed some good fortune that is unlikely to be repeated in 2010. With a non-conference slate that includes Norfolk State, Tulane, FIU, and Army, a 6th straight bowl bid is likely, but barring significant improvement, the Knights are not legitimate threats to win the conference. That designation likely belongs to the usual suspects--Cincinnati and Pitt. Butch Jones followed Brian Kelly at Central Michigan and won 2 MAC titles in 3 years there. He'll have an excellent quarterback in Zach Collaros (much like he had in Dan LeFevour at CMU) to keep the offense humming along. Pitt loses their starting quarterback, but they do return arguably the best running back in the league in super sophomore Dion Lewis. All the diminutive Lewis did as a freshman was rush for 1799 yards, closing the year with 8 straight 100-yard games. The biggest question mark in the Big East in 2010 for me will be Louisville. Steve Kragthorpe promptly took the Big East champion Cardinals and instead of wrapping the proverbial sports car around a tree at 100 mph, he simply neglected to change the oil and decided to fill the engine with 83 octane gas. The chart below shows Louisville's SDPI for the past 5 seasons. The first 2 were Bobby Petrino's doing, and the final 3 were the woeful Steve Kragthorpe era. Winning at Louisville is certainly not unprecedented, but Charlie Strong will have some house to clean in his first season.

Tuesday, February 16, 2010

ACC Recap: 2005-2009

Last season I took a look at how each of the 11 Division IA conferences had played out since 2005 (the last year of major realignment). We're gonna do the same thing this offseason with a few minor tweaks and also with a little commentary on the decade that was. So without further adieu, I present the our first installment: the ACC.

First off, here are combined ACC standings from 2005-2009.

No real surprise at the top. Virginia Tech has won 2 ACC titles in the 5 seasons since the league split into 2 divisions, and has made 3 ACC Championship Game appearances. The 2nd place team may surprise a few folks. Georgia Tech has not had a losing season since division play began in the ACC, and the Jackets own the other 2 Coastal Division titles not captured by the Hokies. Boston College may also be a surprise to some folks at number 3, but they are the only ACC team besides Virgina Tech to not have a league record worse than 5-3 since 2005. 3 teams have yet to post a winning record in league play since 2005, and they all reside in the state of North Carolina (and neither of them is Wake Forest). Duke had their best season since division play began in 2009 when they posted a 3-5 record. North Carolina and NC State have both maxed out at 4 conference wins.

Now here is the standard deviation of each team's conference record (in wins). Teams are ranked from the most inconsistent to the least.

It's a close race at the top with Miami, Virginia, and Maryland ranking as the most inconsistent teams. Each has competed (though never won) a division title--Miami was second in 2005, Maryland was second in 2006, and Virginia was second in 2007, and each has also finished at or near the cellar--Miami was ahead of only Duke in 2007, Virginia was last in 2008, and Maryland was last in 2009. The most consistent team has been Boston College. The Eagles have won 5 league games each season, except for 2007 when they won 6.

Now here is each team's point differential in conference play since 2005.

No surprise that the team with the best record (Virgina Tech) also has the best point differential by far. What is shocking is that Clemson is safely in 2nd place, with more than double the differential of 3rd place Georgia Tech. I think its safe to say Clemson has been a little unlucky in the close game department. Since 2005, the Tigers are only 6-12 in one-score games. If a few of those games had gone the other way, the Tigers could be looking at more than one Atlantic Division Championship.

With this being the end of the decade, here's the tally of conference and division titles.



And finally, what was the biggest takeway from the 00 decade in the ACC?
It's pretty simple, the fall of the mighty Florida State dynasty. In the 1990's, Florida State went 62-2 against the rest of the ACC. They had conference win streaks of 29 and 18 in that span. If we add in the first half of this decade, when the Seminoles still went a solid 34-6, Florida State was 96-8 in their first 13 seasons in the ACC. They had only lost a single time at home, and had never lost more than twice in a single season. They had won 11 of 13 conference titles. Not a bad run. In the last half of this decade, as we've already pointed out, the Seminoles went 21-19. They lost more than twice as many games in less than half the time. In fact, in two separate two seasons spans, they lost as many or more games as they did in their first 13 years in the league. From 2005-2006, they lost 8 games, and from 2006-2007, they lost 9. However, perhaps more amazing is the fact that Florida State has lost 9 conference home games since 2005, after losing just once at home from 1992-2004. It's highly doubtful that Florida State can ever resume the dominance they enjoyed in the 1990's and early 2000's, but a change was likely needed at the top. A team with the resources of Florida State should not struggle to win more than half of its home league games.

Tuesday, February 02, 2010

ACC SDPI

After a brief respite to recharge my batteries, your humble narrator has returned. This offseason will be a lot like the last few here at Statistically Speaking. Each conference will get the SDPI treatment (starting alphabetically with the BCS leagues), and we'll also do another multi-year review of each conference. So let's jump right in, shall we? We'll begin with a league very near and dear to my heart, the ACC. If you want a refresher on just what SDPI is trying to measure, here's the link to last year's article on the ACC.

As usual, this first paragraph will explain how SDPI is calculated. So if you want the meat of this article skip on down. In the 2009 ACC regular season, conference play only, championship games excluded, the average ACC team gained and allowed 2913.167 yards. The standard deviation for yards gained (offense) was 542.56 yards. The standard deviation for yards allowed (defense) was 436.64 yards. Maryland gained 2379 yards and allowed 3108 yards. Their offensive SDPI was -0.98 = [(2379-2913.167)/542.56]. Their defensive SDPI was -0.45 = [(2913.167-3108)/436.64]. Their total SDPI was -1.43. This number ranked 9th in the ACC.

To refresh your memory, here are the 2009 ACC standings.

Atlantic
Clemson 6-2
BC 5-3
Fla St 4-4
Wake 3-5
NC St 2-6
Maryland 1-7

Coastal
Ga Tech 7-1
Va Tech 6-2
Miami 5-3
UNC 4-4
Duke 3-5
Virginia 2-6

Now here are the 2009 ACC SDPI standings. The standings are sorted by division by total SDPI with ranking for each category (out of 12 teams) in parentheses.

Offense Defense Total
Atlantic
Clemson 0.17 (7) 1.15 (2) 1.33 (3)
Wake 0.86 (2) -0.75 (10) 0.11 (5)
Fla St 1.23 (1) -1.63 (12) -0.40 (8)
BC -1.04 (11) 0.15 (6) -0.89 (9)
NC St 0.35 (6) -1.41 (11) -1.06 (10)
Maryland -0.98 (9) -0.45 (9) -1.43 (11)

Coastal
Va Tech 0.61 (5) 1.56 (1) 2.17 (1)
Ga Tech 0.85 (4) 0.63 (4) 1.48 (2)
Miami 0.85 (3) 0.25 (5) 1.10 (4)
UNC -1.00 (10) 1.07 (3) 0.07 (6)
Duke 0.04 (8) -0.39 (8) -0.35 (7)
Virginia -1.93 (12) -0.19 (7) -2.12 (12)

As you can see, there's not a great deal of disconnect between the actual standings and the SDPI standings. According to SDPI, the 2nd and 3rd best teams in the conference faced off in the title game with the 2nd best team walking away with the Orange Bowl bid. SDPI sees Virginia Tech as the best team in the conference, but the Hokies had to play at Georgia Tech, where they lost a close game in what eventually decided the Coastal Division race. The Coastal Divsion was far stronger than the Atlantic with Virginia Tech, Georgia Tech, and Miami all rating out as one standard deviation above average, while Clemson was the lone team from the Atlantic to rate as one standard deviation better than average. The 2 teams the numbers and the actual standings disagree on are Boston College and Wake Forest. The Eagles won 4 of their 5 conference games by a combined 16 points, and each of their 3 league losses came by at least 18 points. Over their 8 league games, Boston College was outscored by 22 points, indicating their record was a little inflated. Meanwhile, Wake Forest boasted perhaps their best offense in school history behind senior quarterback Riley Skinner. However, the Deacons defense had equal trouble stopping opponents and Wake Forest lost 3 league games by a combined 7 points (including 2 in OT) resulting in their first losing season since 2005.

Conference Superlatives

Best Offense: Florida State 1.23
The Seminoles were the lone ACC team with an offense greater than one standard deviation above average. Florida State gained over 400 yards in 7 of their 8 conference games, missing that mark only once; against Clemson. As we'll see in a moment, that offense had to be good.

Worst Offense: Virginia -1.93
The Cavs offense was nearly one whole standard deviation worse than the 2nd worst offense (Boston College). The Cavs never gained more than 298 yards in ACC play, and during a 4-game stretch from mid-October to early-November, the Cavs averaged 186 yards per game. All the more bizarre, is that immediately before that span of suckitude, the Cavs hammered Indiana to the tune of 536 yards and 47 points in an easy 40-point win.

Best Defense: Virginia Tech 1.56
Ho-hum. Another season, another great defense for Virginia Tech. Kudos to Bud Foster.

Worst Defense: Florida State -1.63
The Seminoles gave back all the strides they made on offense by posting perhaps their worst defense in the Bobby Bowden era. Florida State gave up yards in bunches to anyone and everyone. Add to that the fact that the ACC was not an extremely offensive league, and you have the resume of perhaps the worst BCS-conference defense this side of Washington State.

What's in Store for Next Year?
Atlantic Division:
Florida State and Clemson should be your prohibitive favorites heading into the fall of 2010. The Seminoles bring back a ton of talent from the league's best offense with the lone major loss being leading receiver Rod Owens. While Owens did lead the team in catches and yards, he only hauled in 61 balls for 729 yards, indicating Christian Ponder did a great job of spreading the ball around. New head coach Jimbo Fisher proved he could bring the offense back around, now if he can just do something to being the defense back to 'average' Florida State could go back to being a force in the ACC. Fresh off their first Atlantic Division title, Clemson loses 3 of its biggest playmakers on offense (running back CJ Spiller, receiver Jacoby Ford, and tight end Michael Palmer), but does bring back an experienced quarterback and 4/5ths of the starting offensive line. Plus, for the last half decade or so, Clemson has made their mark on defense. From 2006-2009, Clemson finished 2nd, 1st, 2nd, and 2nd in the ACC in defensive SDPI. They're like a poor man's Virginia Tech on that side of the ball. Whatever kinks they may have to work out on offense, Clemson should still be in contention for another division title in 2010. Elsewhere, this may finally be the year Boston College drops from the top of the Atlantic standings. As mentioned earlier, despite their 5 league wins (won at least 5 league games in all 5 years in the ACC), the Eagles were a shell of their former selves. Take a look at the graph below that shows the yearly SDPI of Boston College in the ACC.As you can see, the defense, with the exception of 2008 has maintained a performance level either a little above or below average. With Matt Ryan at the helm from 2005-2007, the offense steadily improved, peaking with a phenomenal year in 2007. However, since Ryan's matriculation, the offense has fallen off not one, but two separate cliffs. The Eagles were not a good football team last year, and barring substantial improvement, should see their record fall more in line with their unimpressive SDPI numbers. The other 3 teams in the division are very much a mystery. As mentioned earlier, Wake Forest boasted perhaps their best offense ever, or at least in the 'modern era' of college football. Unfortunately, that offense loses its trigger man, quarterback Riley Skinner and 4 starters along the offensive line. Wake was a little better than their 3-5 conference record last season, but getting back to a bowl game is not a sure thing. NC State was a huge disappointment last season, and the majority of the blame can once again be laid at the feet of the defense. In Tom O'Brien's 3 seasons as head coach, the Wolfpack have ranked last, last, and 2nd to last in defensive SDPI. I guess an optimist would say 2nd to last equates to marginal improvement. I would agree, and then ask them which Bush campaign they worked on (00 or 04). NC State has a very talented quarterback in Russell Wilson, who has 2 years of eligibility remaining. They would be wise not to waste those years forcing Wilson to win shootout after shootout in the hopes of going 6-6. And finally we have Maryland. The Terps hit rock bottom in 2009, winning only one conference game for the first time since 1998. One of the biggest mysteries, at least to me, is how Ralph Friedgen was able to win 31 games (including an ACC title) in his first 3 years at a program with just one winning season in the previous decade, yet only win 35 games in the next 6 years. Maryland was very young on offense last season, so despite the loss of a mediocre senior quarterback in Chris Turner, the Terps should return to mediocrity, and in this day and age in college football, that means a potential bowl bid.

Coastal Division:
This is where the action should be in 2010 with 4 teams having legitimate shots at the division crown. North Carolina's defense was legit last year, finishing 3rd in the conference in SDPI behind Virginia Tech and Clemson. Almost everyone returns in 2010, so if the offense can improve upon its rather dismal showing, the Tar Heels could take their first division crown. Paul Johnson has proven his triple option attack can succeed in a BCS conference. Unfortunately, the defense let the Yellow Jackets down at times last season. Things won't get any easier in 2010 with the exodus of 4 players with eligibility remaining. Safety Morgan Burnett, defensive end Derrick Morgan, running back Jonathan Dwyer, and receiver Demaryius Thomas will all likely be on NFL rosters in the fall. Thankfully, the Jackets do not have to replace the most important member of their triple option offense, quarterback Josh Nesbitt, who will be in Atlanta for one more season. Since effectively razing their team in 2007 (see the graph below), the Hurricanes have been improving and could finally grab their first division title in 2010.The program was on a steady downward course at the end of Larry Coker's reign, and in Randy Shannon's first year (2007) hit rock bottom. The U has improved on both sides of the ball in both seasons since and with a talented young quarterback (Jacory Harris), the pieces are in place for a run at the division title in 2010. Even with those 3 teams in the mix, its hard to bet against the 'old money' program in the new ACC. Virginia Tech has won the Coastal Division 3 times in the 5 years the ACC has had the division format. In the 2 seasons they did not advance to the ACC Championship Game, the Hokies finished just one game back of the eventual division winner. 2009 marked the first time since I have been calculating SDPI that Duke did not finish dead last in the conference. That's a testament to head coach David Cutcliffe. The big problem for Cut and the Blue Devils in 2010 is that they must break in a new quarterback to replace the departed Thaddeus Lewis who threw for 10,065 yards and 67 touchdowns in his 4 seasons in Durham. The other problem for Duke is they play in the Coastal Division and must face the aforementioned Four Horsemen of the ACC each and every season. Once again, Duke will be undermanned, and well coached as they journey the ACC. The only other ACC team besides Florida State that will be starting over with a new coach in 2010 is Virginia, who plucked Mike London from nearby Richmond. London guided the Spiders to 2 playoff appearances in 2 seasons as coach, including a run to the National Championship in 2008. London was previously the defensive coordinator here at Virginia, holding that title in the 2006 and 2007 season. In those seasons, Virginia ranked 4th and 3rd respectively in the ACC in defensive SDPI. In the 2 seasons, since the Cavs have ranked 7th both times in that category. London clearly is a solid if not excellent defensive mind, but so was Al Groh. If London is to be truly successful at Charlottesville, he will have to fix an offense that has not been better than average since Matt Schaub left after the 2003 season.