How Good Are the Irish? A Year-End Offensive Statistical Review
It may be head coach Brian Kelly’s first year at the helm, but positive change is already tangible.
While the 2008 and 2009 Irish squads finished 1-8 in November, this year’s unit went 3-0. It isn’t time to anoint Kelly the next Lou Holtz or Ara Parseghian, but Notre Dame played its best football down the stretch without several critical players such as quarterback Dayne Crist, running back Armando Allen, wide receiver Theo Riddick, tight end Kyle Rudolph, inside linebacker Carlo Calabrese, and nose guard Ian Williams who all missed at least one game due to injury.
But Kelly’s troops weren’t just undefeated, they were dominant. The Irish displayed a toughness and physicality led to a November winning streak where Notre Dame outscored their three opponents by a margin of 75-22.
If that wasn’t enough, the decisive victory over the Miami Hurricanes in the Sun Bowl was the explanation point on the season. The Irish dominated through the three meaningful quarters of play with an advantage in plays (20), yards (116—including an 85-yard advantage on the ground), first downs (6), turnovers (+4), and explosive gains (4 big gains, 116 big play yards). The result was a 30-3 score at the end of the third period, a far more accurate representation game’s outcome than the 33-17 final score.
The upward team trajectory combined with the return of a host of starters on both sides of the ball has many Irish fans excited about 2011. But before looking ahead, a review of the past is in order. This is the first of a two-part installment detailing the offensive and defensive performance of the Irish. This offensive analysis appropriately benchmarks the statistical production, providing value and ranking information on several pertinent metrics, and outlining underlying trends that contributed to the overall performance.
Benchmarking Notre Dame’s Performance
Consistent with previous assessments (2009 year-end offensive and defensive analyses, 2009 mid-year offensive and defensive analyses, and 2008 year-end offensive and defensive analyses), the statistical performance detailed here will be benchmarked to the competition.
There are a host of reasons benchmarking is valuable, but the primary one is quite simple: an examination of a team’s performance, without consideration of the competition against which that performance was generated, excludes a huge source of variance.
Here, performance ratios (or PR’s) will be used to account for the quality of competition (additional information on PR’s can be found here). Performance ratios are not entirely perfect, but do provide insight into how the opposition can skew a team’s performance and form the basis for the Team Performance Ratio (TPR) of the AVR.
A Tabular Description
The tables below contain 36 statistical metrics divided into five categories: miscellaneous, offensive efficiency, and total, rushing, and passing offense. This data will be supplemented with more detailed information aimed at understanding underlying factors that contributed to the high-level performance.
The pertinent statistics for each category are presented in tabular form and include the numbers for Notre Dame and the averages of their 13 opponents.
Each table has six columns (the miscellaneous table is void of PR’s):
- Statistic — The statistical metric
- Notre Dame — The value of 1. for the Irish offense
- Notre Dame Rank — The rank of 1. for the Irish offense
- Opponent Average — The average defensive value of 1. for Irish opponents
- Opponent Average Rank — The average defensive rank of 1. for Irish opponents
- PR — The PR of 1. for Notre Dame—values greater than zero indicate the Irish are performing above the average level of their competition, values less than zero are indicative of below average performance (Notre Dame ranked 64th in offensive TPR in 2010).
The stats presented here are from the NCAA statistics website and are accurate as of January 10, 2011.
Definitions and Disclaimers
The following disclaimers and definitions are pertinent to this discussion:
- Definition: An explosive—or big—gain is a run of 15 or more yards or a pass of 20 or more yards.
- Definition: Down and distance situations are defined as follows: short—less than three yards for a first down/touchdown, medium—between three and seven yards, long—greater than seven yards
- Definition: Open downs are those where a run is equally as viable as a pass, e.g. 1st and 10, 2nd and medium, and 3rd/4th and short. Closed downs are those where a pass play is “needed” for a chance to convert a first down/touchdown, e.g. 2nd/3rd/4th and long.
- Definition: A successful play is one that achieves the following gains: 1st down—40 percent or more yards needed for a first down/touchdown, 2nd down—60 percent or more yards, 3rd/4th down—100 percent of yards. Expressed differently, a succesful play is one that maintains an open down situation on the following play on play series that begin with 1st and 10.
- Definition: A meaningful possession is one that occurs when the outcome of the game is still in doubt (defined by score differential and number of remaining possessions), and excludes drives where the quarterback takes a knee to preserve a victory.
For reference purposes, the following are links to the statistical recaps for each 2010 Irish opponent (Purdue and Miami in absentia): Michigan, Michigan State, Stanford, Boston College, Pittsburgh, Western Michigan, Navy, Tulsa, Utah, Army and USC.
Possession, Penalties and Turnovers—One Out of Three Ain’t As Bad As It Seems
Time of possession stunk, but is it a meaningful metric? The (penalty) discipline was good. Turnovers must improve for the Irish to reach their ceiling.
|Statistic||Notre Dame||Notre Dame Rank||Opponent Average||Opponent Average Rank|
|Time of possession||27:56||105||30:24||54|
There was a lot of talk prior to the season about Kelly’s offensive approach and how it didn’t seek to manage the game via ball control. Many were fearful that operating at a fast pace and scoring quickly would leave the defense on the field far too long.
True to his form, Kelly didn’t maximize possession. The Irish ranked in the bottom 15 of the country with 27:56 in time of possession. Furthermore, Notre Dame only enjoyed a possession advantage against WMU, Army and Miami, the last being the most skewed time of possession game of the entire season.
But, in reality, losing the time of possession battle is only important if a team is also behind in play differential, and Notre Dame ran more plays than their opponent in seven of 13 games (1.4 more plays per game). Additionally, the Irish offense averaged 2.47 plays per minute while opposing offenses averaged 2.18. In other words, the Irish defense had more time to rest between plays than did their opponents.
Former head coach Charlie Weis’ teams were never terribly undisciplined, but Kelly’s squad really minimized penalties. The Irish ranked 7th in penalties per game and 15th in penalty yards per game. Not only was Notre Dame not racking up infractions with any frequency, they also weren’t committing many penalties of the 10 and/or 15-yard variety.
Unfortunately, however, the discipline didn’t translate to the turnover department where the Irish were only +1 on the season (51st in the country). Notre Dame won the turnover battle seven times in 2010, but were on the other side of the ledger in several close losses including Michigan (-3), Michigan State (-2), and Tulsa (-2). Most of the damage came in the form of interceptions, Irish ball carriers only lost eight fumbles on the season (25th ranked), while Crist, Nate Montana, and Tommy Rees combined for 16 interceptions (99).
The most damaging aspect of the turnovers was the points opposing offenses were able to generate from them. The offense surrendered 24 turnovers that resulted in average starting field position near midfield and resulted in 69 points (2.9 per turnover, 5.3 per game). The defense was able to force 25 takeaways and turn them into 64 points (2.6, 4.9), but eight turnovers and 27 points off them came against the Broncos and Hurricanes. Excluding these two games the Irish faced a huge turnover deficit, forcing only 17 takeaways that turned into 37 points (2.2, 3.4) compared to allowing 69 points off 23 turnovers (3, 6.3), i.e. almost a touchdown per game surrendered off turnovers.
Regression In the Efficiency Department(s)
Down efficiencies regressed. Red zone opportunities are down. Red zone efficiency was eerily similar to 2009.
|Statistic||Notre Dame||Notre Dame Rank||Opponent Average||Opponent Average Rank||PR|
|3rd down efficiency||38.3||73||40||64||-0.04|
|4th down efficiency||45.4||82||52.9||65||-0.14|
|Red zone appearances/game||3.5||71||3.3||48||0.06|
|Red zone efficiency||83||49||80.4||52||0.03|
|Red zone TD efficiency||56.5||87||56.9||55||0|
Third down efficiency declined slightly from 41 percent in 2009 to 38.3 percent this season. A breakdown of the situational third down efficiency (percentage of third downs/conversion rate):
- 3rd and short—23.9/62.8
- 3rd and medium—18.9/52.9
- 3rd and long—57.2/23.3
- 3rd and more than three yards—76.1/30.7
Compare the values above to the situational third down efficiency for the Irish defense (36.5 percent conversion rate, 28th ranked):
- 3rd and short—33/62.7
- 3rd and medium—26.1/34
- 3rd and long—40.9/16.9
- 3rd and more than three yards—67/23.5
In other words, converting in unfavorable situations wasn’t the problem—the Irish offense moved the chains at a higher rate than their counterparts in every situational category. But facing 3rd and medium-to-long distances occurred far too often, something that speaks to poor first down execution (see below for more detail on this).
Red zone efficiency was almost identical to last season, despite Kelly’s strong track record at Cincinnati. To compound the problem, red zone appearances per game were also down from 4.2 in 2009 to 3.5 this year. Virtually the same red zone efficiency coupled with fewer opportunities at least partially explains the dip in scoring production.
Goal-to-goal play, however, did improve. The Irish offense scored on 20 of 26 (76.9 percent) goal-to-go opportunities in 2010 compared to 15 of 25 in 2009 (60). Kelly’s squad notched eight goal-to-go rushing touchdowns (out of 11 total) and 12 passing scores to combine for 20 of 39 offensive touchdowns in goal-to-go situations. Part of being a strong play-caller is creating and taking advantage of high-probability scoring opportunities, and Kelly certainly improved over Weis in this area.
Perhaps more impressive was the execution improvement in these situations. Over half of the goal-to-go plays in 2009 went for zero or negative yardage compared to only 38.9 percent in 2010—with three new starters in the front five. Additionally, Kelly’s quarterbacks completed 60.7 percent of their passes in goal-to-go situations compared to 37.5 percent last season—with a first-year starter (Crist) and true freshman (Rees) under center, not the uber-accurate Jimmy Clausen.
Part of the reason for the improved performance lies in the play-calling. While Weis’ shifted from a pass-happy, spread attack in the open field to a run-first, heavy formation offense inside the 10-yard line, Kelly’s play-calling approach didn’t change. Counting sacks as pass plays, the Irish run/pass split overall (44/56), outside the red zone (44.4/55.6), and in goal-to-go situations (44.4/55.6) was virtually identical. There is something to be said for consistency in approach, and, in this case, it translated into solid goal-to-go scoring production.
A Fairly Pedestrian (Total) Offense
Rankings fall in the middle of the pack with PR’s around zero. Failing to capitalize on early scoring opportunities and poor execution on first down were the primary culprits.
|Statistic||Notre Dame||Notre Dame Rank||Opponent Average||Opponent Average Rank||PR|
Notre Dame, from a total offensive production standpoint, was a mediocre unit. A quick glance at the rankings in the table above shows a range from 53rd to 76th, with a mean value of 62. Moreover, all of the PR’s are just over zero, indicating that the Irish produced near the rate opposing defenses typically allowed.
From a construction standpoint, Kelly largely favored the pass over the run (more on this below), with 33.3 percent of the total yards coming on the ground (96th in the country) compared to 66.6 percent of total yardage via the air (25). The first down and scoring trends are the same. Just over 34 percent of first downs came on the ground (102) compared to 58.2 percent through the air (20), while 28.2 percent of offensive touchdowns were rushing scores (117) compared to 71.8 percent passing (4). And, as noted above, eight of those 11 rushing touchdowns came in goal-to-go situations when calling a run play is virtually a no-brainer.
But being a pass-heavy offense doesn’t necessarily translate into average production. So, why where the Irish so pedestrian? For one, replacing over 70 percent of the scoring and yardage output—i.e Clausen and Golden Tate—isn’t easy. But there were certainly other issues, namely failing to take advantage of early scoring opportunities and poor execution on first down.
Notre Dame received the opening kickoff in 12 of their 13 games (USC was the lone exception), but only scored four times—all touchdowns—to open the game. On an aggregate basis, the Irish only generated a third of the possible opening possession points. For a unit that puts such a premium on receiving the opening kickoff, that is a relatively low return.
Moving the ball wasn’t really the problem, mistakes were. Including only meaningful drives, the Irish offense gained 42.6 percent of their available yards, but surpassed that mark on seven of 12 opening drives. Five opening drives, however, were punts (average 21.4 percent of available yards), one ended in an interception (93.2), and two others ended in a turnover on downs (62.5). In other words, three Irish possessions gained an average of 72.7 percent of available yards but failed to generate a single point.
Slow starts aside, the bigger problem was first down play. The Irish averaged 5.7 yards per first down snap (64th in the country) including a paltry four yards per carry (87) and 7.5 yards per pass attempt (66). This clearly had a negative impact on the poor third down efficiency detailed in the section above, but it also impacted the ability to sustain drives and put points on the board.
Including only meaningful drives, Notre Dame averaged 8.4 first down yards on touchdown drives and eight yards on scoring drives (touchdown and field goal), but only 3.2 yards on non-scoring possessions. Minimal first down gains lead to unfavorable down and distances on third down, and, as noted above, the Irish struggled mightily with situational third down offense.
And the problem was not one of play-calling, it was one of execution. A breakout of the first down production by possession type (run/pass/total/average yards to go on third down):
Additionally, the first down successful play percentages for touchdown (57.5 percent), scoring (55.4), and non-scoring (36.7) possessions further reinforce the contribution effective first down execution played on the overall success of the offense.
The Offensive Focus Was Not Running the Ball
The rushing attack wasn’t the focus of the offense, but (when called upon) performed quite well…although explosive rushing plays continue to be problematic.
|Statistic||Notre Dame||Notre Dame Rank||Opponent Average||Opponent Average Rank||PR|
Despite playing with two young and/or inexperienced quarterbacks, Kelly’s play-calling focus was not running the football, and the numbers back it up. The run/pass production for yards, first downs, and scoring mentioned above highlight a trend that is further supported by the rankings for rush attempts (109), yards (92), first downs (97), and touchdowns (109) per game, all of which fall in the bottom quarter of the country.
Additionally, the overall run/pass split (46.3/53.7) ranked 104th as the offense was run-heavy in only five games, and three of them—Utah, Army and Miami—came with Rees at quarterback. In fact, Kelly practically reversed his approach with Rees taking snaps. The Irish run/pass split from Purdue through Tulsa was 41.4/58.6 but switched to 58.8/41.2 over the final four games.
The efficiency and effectiveness followed the play-calling. Notre Dame averaged 29.7 attempts, 113.4 yards (3.8 yards per carry), and 6.7 first downs per game on the ground in their first nine outings compared to 36.8 attempts, 156.3 yards (4.3), and 7.3 first downs per game in their last four.
What makes the trends puzzling is the level of production when a run was called. Again, counting sacks as passes, the Irish averaged 4.4 yards per open down carry, 4.4 yards per first down rush, and 4.9 yards per attempt outside of the red zone. What was missing, however, were big plays.
Notre Dame ranked 77th in 10-plus yard rushing gains and 49th in 20-plus yard running plays. Given the relatively few opportunities to reel off a big gain, these numbers appear respectable. But a comparison to last year’s low-production rushing offense shows otherwise.
The 2010 offense generated 26 explosive rushing gains that totaled 614 yards (23.6 yards per carry). Expressed differently, Irish backs ripped off a 15-plus yard gain once every 15.9 attempts. Compare these numbers to 2009 when the offense generated 25 big rushing gains for 494 yards (19.8 yards per carry), i.e. one every 16 attempts. The 2010 per-carry average is nearly four yards higher, but the frequency of explosive rushing gains is virtually identical, as is the per-carry average without big plays—2.7 yards per carry in 2010 compared to 2.8 yards per carry in 2009.
Plenty O’ Passing, But Not A Lot of Efficiency
Kelly was quite clearly a pass-first play-caller, and the front five protected quite well, but the efficiency simply wasn’t there. Big play production took a dip as well.
|Statistic||Notre Dame||Notre Dame Rank||Opponent Average||Opponent Average Rank||PR|
Simply stated, Kelly likes to throw the ball. The 2010 offense attempted 37 passes per game (22nd ranked) compared to 37.3 last year (19). And this with a first-year starter and true freshman taking snaps.
Despite all the passes, the efficiency, and even the effectiveness to a lesser extent, are hardly reminiscent of last season when the Irish fielded one of the nation’s most prolific air attacks. The loss of Clausen and Tate (along with three new offensive line starters) essentially precluded a repeat performance, but ranking 22nd in attempts per game, 75th in yards per attempt, 58th in completion percentage, and 75th in pass efficiency, speaks to relatively poor execution. Stated differently, throwing the ball so often with so little return isn’t a great option.
The front five played extremely well considering the youth and inexperience of first-year starters Zach Martin, Braxston Cave, and Taylor Dever. Notre Dame ranked 38th in sacks allowed per game at 1.5 and 26th in sacks per pass attempt at 24.1, both marks better than last year when the Irish fielded a mostly veteran offensive line. The attempts per sack were also better than the 2008 unit which allowed only 22 sacks.
But while these two sack metrics are impressive in their own right, they loom even larger after accounting for the competition. The PR’s in both categories (0.27, 0.39) indicate that the Irish offensive line limited their opposition to well below their usual sack production. No where was this more on display than in the Sun Bowl. The Hurricanes entered the contest averaging 3.1 sacks per game (6th in the country) but failed to record a single quarterback takedown against Notre Dame.
The Irish also performed well notching scores through the air. Notre Dame ranked 20th in passing touchdowns per game, again performing at a much higher level than their defensive competition (PR of 0.36). Some of the passing touchdown production was the result of the frequency of the throws, but improvement in the red zone also played a big part.
The Clausen-led 2009 offense notched 15 passing scores inside opponents’ 20-yard line while Kelly’s 2010 unit generated 18—primarily due to improved efficiency. Irish quarterbacks completed 62.5 percent of their red zone pass attempts in 2010, compared to 45.2 percent in 2009. Additionally, Crist and Rees combined for a 175.8 passer efficiency while Clausen only managed a 139.8 mark.
In addition to a dip in the efficiency and effectiveness of the passing offense, the big play production also fell. Weis’ 2009 unit notched a big play through the air every 9.3 pass attempts compared to 11.7 in 2010. Additionally, Clausen and company averaged nearly five yards more per big pass play and almost a yard per attempt more excluding explosive gains.
How About Some Possession Metrics?
Stat guru and Notre Dame fan Brian Fremeau (you can follow him on twitter @bcfremeau) of BCF Toys and College Football Outsiders was nice enough to provide some input on the Irish offensive possession-based efficiency using some components of the Fremeau Efficiency Index (FEI). A few highlights (metric, value/ranking—explanation):
- Offensive FEI, 0.122/42—Opponent adjusted success rate of maximizing offensive possessions
- Offensive Efficiency, -0.046/64—Unadjusted value of offensive drive success, i.e. the drive-ending scoring value of the offense divided by the expected value (based on starting field position)
- Offensive 1st Down Rate, 0.655/75—Percentage of offensive possessions resulting in at least one first down
- Offensive Available Yards, 0.44/66—Drive yardage gained divided by yards available
- Offensive Explosive Drives, 0.152/37—Percentage of drives that averaged at least 10 yards per play
- Offensive Methodical Drives, 0.131/64—Percentage of drives with at least 10 plays
- Offensive Strength of Schedule, 0.094/36—Likelihood that an elite offense (two standard deviations better than average) would have an above average (unadjusted) offensive efficiency against every opposing defense on the schedule
- Offensive Points Per Possession, 2.12/62—Points scored per meaningful drive
- Offensive Field Goal Efficiency, 0.751/3—The field goal efficiency measured by the value of a kicker (in points per attempt) over an average kicker accounting for the percentage of field goal points contributed by the offense, drive starting field position, and the kicker.
Overall, the possession-based efficiency numbers stackup very closely with the play-by-play-based metrics above. However, the Offensive Explosive Drives ranking and bipolar first down production on touchdown/scoring vs. non-scoring possessions indicates is a need for consistency. In other words, when the offense clicks, it really clicks. But a lack of consistent execution was the rule, not the exception.
What Is Needed Going Forward
Given the change that occurred last off-season, as well as the injuries to key personnel, Kelly’s first offensive unit performed fairly well. The Irish lost their two most prolific offensive players in recent memory, had three first-year starters on the offensive line, implemented a new scheme with virtually an entirely new coaching staff, and lost several critical offensive players to injury, yet still managed to be an average offensive football team.
Many (myself included) struggle to understand why the game plan was so pass-heavy with a new scheme, inexperienced/young quarterbacks, and new starters in the front five, but, in some regards, the Irish actually threw the ball better than last season when they had one of the most polished quarterbacks and explosive wide receivers in Notre Dame history. There weren’t many areas where the Irish offense excelled, but there was marked increase in passing efficiency in the red zone and in goal-to-go situations. Given the importance of these high-probability scoring situations, this improvement should not be understated.
The off-season will undoubtedly bring about an interesting quarterback competition, with as many four signal callers vying for the starting spot. But the balanced, more conservative offensive approach in the last four games is evidence that Notre Dame—provided the dominant defensive play continues—can beat quality opponents without exceptional production from the quarterback position.
That doesn’t mean there isn’t room for improvement, there certainly is. Consistency is the name of the game moving forward, and it needs to manifest itself in two primary metrics: turnovers and first down.
If Kelly wants a BCS caliber team, improvement in turnover margin is a must. From 2000 to 2009 eight of the 10 BCS champions were +6 or better in turnover margin and nine ranked in the top 30. Moreover, the average turnover margin for these 10 title winners was +14.2. The 2010 Irish defense did their part generating 25 takeaways, but 24 turnovers is far too many and cost the Irish at least one game this season.
Additionally, more production is needed on arguably the most critical down in football. When the players executed Kelly’s spread scheme, Notre Dame consistently produced on first down, faced manageable distances on third down, moved the ball well, and extended drives. Unfortunately, this didn’t happen with enough frequency.
In summary, Kelly’s inaugural offense hardly set the world on fire. But given the situation, a top-flight performance wasn’t a realistic expectation. If the front five continue to play with physicality and the unit continues to gel as a team, the rest will take care of itself. An off-season of fast-paced, rep-heavy practices should improve consistency, and the return of wide receiver Michael Floyd provides a dependable option for any of the potential quarterbacks. Couple that with a vastly improved Irish defense, and you arrive at a recipe for a double-digit win season in 2011.