Handicapping the Coastal Division

Inside Carolina provides insight into critical statistics for the 2015 season.

In order to obtain an accurate gauge on the season to come, it is necessary to look back on how the 2014 season played out.

In the two charts below, total defensive and offensive outputs are listed along with yards-per-play averages to provide context of the snap differential. For example, Georgia Tech ranked 81st nationally in total defense despite ranking 111th in yards per play. The reason is that the Yellow Jackets defense played 65.1 snaps per game, compared to the 76.2 defensive snaps UNC played per game.

The numbers in the final column represent an advanced metric developed by Bill Connelly at FootballOutsiders.com. The four components of the S&P+ ranking are efficiency, explosiveness, field position and finishing drives. It’s a good gauge for how effective an offense or defense was in a given year beyond the basic stats.

Team Total Defense Yards Per Play S&P+ Rating
Miami 329.6 4.79 113.4 (28th)
Virginia Tech 343.8 5.15 124.9 (11th)
Virginia 353.2 5.14 109.9 (35th)
Pittsburgh 359.8 5.59 98.2 (67th)
Duke 399.7 5.28 97.8 (70th)
Georgia Tech 411.3 6.32 103.2 (46th)
North Carolina 497.8 6.53 88.9 (109th)

Team Total Offense Yards Per Play S&P+ Rating
Georgia Tech 476.5 6.72 128.6 (4th)
Pittsburgh 435.4 6.17 115.4 (19th)
Miami 430.3 6.68 108.9 (35th)
North Carolina 429.8 5.56 100.5 (58th)
Duke 398.3 5.38 100.4 (59th)
Virginia 374.2 5.08 96.1 (78th)
Virginia Tech 365.3 4.93 94.0 (88th)

A closer look at those statistics reveals which teams outperformed and underperformed their expected results.

Football analyst Phil Steele has long relied on an elementary formula to determine which teams overachieved or underachieved in any given year: subtract net close wins (games decided by 7 points or less) from net close losses. Despite the simplistic nature of this research, Steele’s data over a 12-year stretch ending in 2013 indicates that 76.9 percent of teams with three net close wins failed to improve on their win total the following season.

Connelly has added advanced statistics to the mix to determine how many points a team should have scored and how many it should have given up, utilizing factors such as efficiency, turnovers and field position. In doing so, his stats suggest how many games a team should have won in any given year, which he labels second-order wins.

Connelly’s data indicates that UNC, not Georgia Tech or Duke, was the most significant overachiever in the Coastal Division in 2014. Pittsburgh, on the other hand, was the conference’s worst underachiever, which could play into new head coach Pat Narduzzi’s favor.

Team 2014 Record Second-Order Wins Net Close Wins/Losses
North Carolina 6-7 4.9 (-1.1) +3 (4-1)
Georgia Tech 11-3 10.1 (-0.9) 0 (3-3)
Duke 9-4 8.4 (-0.6) +1 (3-2)
Virginia Tech 7-6 7.0 (0.0) -3 (2-5)
Virginia 5-7 6.0 (+1.0) -1 (2-3)
Miami 6-7 7.1 (+1.1) -2 (0-2)
Pittsburgh 6-7 8.3 (+2.3) -4 (1-5)

For the second year in a row, UNC shares the ACC lead in returning the most total starters (18). The Tar Heels also return the most starters (5) and career starts (105) along the offensive line. The chart below provides perspective on the quality of each team's offensive line play in 2014 with sacks allowed and yards-per-carry averages.

Team Starters Returning OL Starters Returning OL Career Starts 2014 Sacks Allowed 2014 Yards Per Carry
North Carolina 18 5 105 28 3.97
Virginia Tech 18 2 42 34 3.79
Pitt 17 4 63 21 5.32
Georgia Tech 16 4 72 11 6.06
Duke 14 3 54 13 4.80
Miami 11 2 27 21 5.26
Virginia 11 3 68 16 3.67

Abundant amounts of experience in the trenches have long been considered a predictor of offensive success but significant supporting statistical data is not readily available.

In 2014, for example, there were 21 teams that returned 100-plus career OL starts, but only 29 percent of those teams improved their yards-per-carry average and reduced their number of sacks allowed from the year before. Ten of the 21 allowed fewer sacks last fall, while only eight increased their yards-per-carry mark.

Conversely, Ohio State entered its national championship season with 21 career OL starts and 11 other bowl participants returned fewer than 40 career OL starts.

For UNC, in particular, experience up front has been a positive development in recent years. The 2015 season marks the fifth time over the past decade that UNC has returned four or more offensive line starters. Improvement has been made in both sacks allowed and yards-per-carry average in three of those four previous years.

In 2011, UNC improved its carry average from 3.71 to 4.08 and cut its sacks allowed total from 37 to 27. The Tar Heels returned four OL starters again in 2012 and raised its yards-per-carry average to 5.09 and gave up 11 sacks in Larry Fedora's first season as head coach.

In order to expand on the experience aspect of the equation, it’s beneficial to look at the amount of production returning on both sides of the ball. Steele tackled that exhaustive challenge in June (here and here).

Team Returning Starters % of Yards Returning % of Tackles Returning
North Carolina 18 93.3 70.9
Virginia Tech 18 92.0 68.4
Pittsburgh 17 87.3 64.4
Georgia Tech 16 29.6 74.7
Duke 14 35.2 69.0
Miami 11 58.7 59.8
Virginia 11 41.7 45.6

Quarterback play should be a strength in the Coastal this fall with only Duke and Virginia having to replace their starting signal callers. Junior Thomas Sirk will replace Anthony Boone at Duke, while UVa junior Matt Johns won the job this spring over incumbent Greyson Lambert, who promptly transferred to Georgia.

Quarterback Total Offense Yards Per Play Rushing/Passing Yards Completion Percentage TD/INT
Marquise Williams, UNC 296.6 6.21 783/3073 63.1 21/9
Brad Kaaya, Mia 235.9 7.57 (-131)/3198 58.5 26/12
Chad Voytik, Pitt 207.6 6.83 466/2233 61.3 16/7
Michael Brewer, VT 206.7 5.16 (-5)/2692 59.4 18/15
Justin Thomas, GT 200.4 7.44 1086/1719 51.3 18/6
Matt Johns, UVa 101.3 6.61 107/1109 54.9 8/5
Thomas Sirk, Duke 25.4 5.00 238/67 71.4 3/0

Kaaya, Thomas and Voytik all ranked in the top-5 of the ACC in total offense yards-per-play in 2014.

Before pairing the aforementioned stats with personnel upgrades (Narduzzi to Pitt as HC, Gene Chizik to UNC as DC, etc) to project the 2015 Coastal Division standings, there is one final piece of the handicapping puzzle: scheduling.

Each team will play six conference games against divisional foes, but the cross-division matchups could grease the tracks for certain teams while constructing additional obstacles for others. In 2013, for example, Duke won the Coastal and played Florida State for the ACC title after beating N.C. State and Wake Forest – the Big Four rivals combined for a 2-14 ACC record that year - in cross-division matchups.

The following chart includes the NCAA’s preseason strength of schedule (based on 2014 opponent records), key ACC road games and cross-division foes.

Team SOS Key Road Games Cross-Division Opponents
Miami 3 FSU, UNC, Pitt FSU, Clemson
Virginia 12 Pitt, UNC, Mia Louisville, Syracuse
Georgia Tech 14 Clemson, Mia Clemson, FSU
Virginia Tech 46 Mia, GT N.C. State, Boston College
North Carolina 52 Pitt, GT, VT N.C. State, Wake Forest
Pittsburgh 52 VT, GT Louisville, Syracuse
Duke 102 VT, UNC Wake Forest, Boston College

Inside Carolina Top Stories