Saturday, July 17, 2021

First Half Point Differential: The 2020 Group of 5

Last year around these parts, I introduced a very complicated advanced stat: First Half Point Differential (1HPD). It is a difficult concept to explain concisely, but I'll give it a shot. It is the number of points by which a team outscores its opponent in the first half of a game. The theory behind using it being that consistently relying on second half comebacks is not a good long term strategy and can potentially help us identify regression candidates. Similarly, teams that rack up solid first half differentials, but ultimately wind up losing more game than we might expect actually have the bones of a solid team and might be a good candidate to bounce back the next season. In our YPP and APR offseason recaps, we go through each conference individually, but since the season will (hopefully) be here in about six weeks, we'll tackle the Group of Five this week and in two weeks examine the Power Five. 

American Athletic Conference
The 2020 AAC standings. 
1HPD in AAC play (title game excluded with numbers on a per game basis). 
I expected Cincinnati to come out on top here, but UCF actually led by double digits in seven of their eight conference games. If you just looked at these numbers, you would have expected East Carolina and Tulane to have much better seasons as they outscored their opponents in the first half while Houston, Memphis, and SMU all finished with at least a .500 conference record despite being outscored in the first half of their conference games. 

Conference USA
The 2020 Conference USA standings. 
1HPD in CUSA play (previous disclaimers apply).
Teams in CUSA played a wildly disparate number of games (as many as seven and as few as three) so I would not put a whole lot of stock in these numbers. 

Mid-American 
The 2020 MAC standings.
1HPD in MAC play (ditto). 
The top team (Buffalo) and bottom team (Bowling Green) really stood out in 2020. Buffalo led at halftime in each of their five regular season conference games by at least a touchdown and on average by more than two touchdowns. Meanwhile, Bowling Green trailed by at least two touchdowns at halftime in all five of their MAC contests. 

Mountain West
The 2020 Mountain West standings.
1HPD in Mountain West play (ditto).
The bottom of the MAC was bad in 2020 and the bottom of the Mountain West was equally horrific. Three teams (Colorado State, Utah State, and UNLV) trailed by more than ten points on average at halftime of their conference games with Utah State and UNLV never holding halftime leads. 

Sun Belt
The 2020 Sun Belt standings.
1HPD in Sun Belt play (ditto).
Nearly every team in the East division had a positive differential at halftime, with Georgia State the lone holdout (Panthers were outscored by a grand total of one point in the first half). 

In the YPP and APR offseason recaps, I sort the teams in each conference by how much they over or under-performed relative to their expected record. Since all the Group of Five conferences are grouped together here, I am only going to list those that significantly over or under-performed (a difference of at least .200). We'll start with the overachievers.
In the AAC, Memphis trailed in six of their eight conference games at the half! That they ended up winning five games was a minor miracle. Meanwhile, Tulsa trailed by double digits in half of their six regular season conference games, but managed to come back and win each one. Take UAB's Conference USA numbers with a grain of salt. They played only four conference games and had a slightly positive margin in them. However, UTSA is a different story. The Roadrunners nearly played a full league schedule (seven games), and were outscored in the first half despite winning five times. Ball State made a surprise run to the MAC Championship Game with a decent 1HPD, but were tied or trailing at halftime in half of their conference schedule. San Jose State also made a surprise run to their conference title game despite middling 1HPD numbers. 

And now the underachievers. 
I briefly touched on East Caroline and Tulane finishing with losing records despite outscoring their conference opponents in the first half, but I'll add that Tulane only trailed at the half once in their eight conference games. I also touched on UCF dominating their opponents in the first half. Against both Memphis and Tulsa they blew huge leads in games they easily could have won. The two Conference USA teams on this list finished winless in short seasons and probably should have broken through for at least one or two wins based on their numbers. Eastern Michigan won just a third of their MAC games despite holding halftime leads four times. Meanwhile, Northern Illinois went winless in MAC play, and while the Huskies never held a halftime lead, they were tied twice. Finally, Troy finished with a losing Sun Belt record despite holding halftime leads in five of their seven conference games, with three of those leads being double digits. 

Will these over and underachievers regress or progress to the mean in 2021 or were these numbers products of a truly unusual season with little to no predictive value going forward? Hopefully, we'll get some answers starting in late August. As always, thanks for reading and we'll be back in two weeks with the Power Five numbers. 

Thursday, July 08, 2021

YPP Throwback: The 2000-2003 Big East (the final years of the old Big East)

In this edition of our YPP Throwback, we are actually going to examine four seasons (and a few Jersey Boys) of the early aughts Big East. These four seasons encompass the supernova run of the Miami Hurricanes when they reemerged from sanctions to become a national player for the first time in a decade only to see their potential dynasty abate just as they left the Big East to join the ACC. 

We start, as we always do, with the standings. This time, its the Big East in 2000. 
Miami and Virginia Tech combined to go 22-2 in 2000 with both teams finishing in the top six of the final AP Poll. Miami's only loss came in the second game of the season by five points on a cross-country trip to Washington. That Washington team was pretty damn good, finishing 11-1 and winning the Rose Bowl. Meanwhile, in Michael Vick's final season in Blacksburg, the Hokies lone defeat came at Miami (in a game Vick hardly played). Outside of Miami and Virginia Tech, the only notable aspect of the Big East in 2000 was the final season of legendary West Virginia coach Don Nehlen. Nehlen capped his career by winning his first bowl game in sixteen seasons against Ole Miss in the Music City Bowl

And here are the YPP numbers for the 2000 Big East.
While Virginia Tech finished a game behind Miami in the Big East standings, they were clearly several notches below Miami once we look at the YPP numbers. The Hokies were second on offense and defense, but the gap between the Hokies and Hurricanes was enormous. Miami was about 0.78 yards per play better than the Hokies on offense. That was larger than the gap between Virginia Tech and the sixth best offense (West Virginia). The gap on defense was similar. Miami was about 0.80 yards per play better than the Hokies on that side of the ball. That was greater than the difference between Virginia Tech and the fifth best defense (Temple). On the other end of the spectrum, Rutgers finished with the worst YPP Net in the Big East. Despite making a great hire, in the short term, things would get much worse. 

Here are the 2001 Big East standings. 
Miami once again finished with an unblemished record, and while they were often regarded as the best team of this century until the recent iterations of Clemson, LSU, and Alabama destroyed the curve, they were challenged a few times in conference play. Boston College was driving with a chance to win late until a tipped pass led to an Ed Reed touchdown and a misleading final score. In their regular season finale, Miami avoided the late season upset that plagued highly ranked teams in 2001 by stopping a two point conversion against Virginia Tech that could have sent the game to overtime. Elsewhere in the conference, Syracuse shook off an 0-2 start to win ten of their last eleven games and finish in the top-fifteen of the final AP Poll for the first time in nearly ten years. And at the bottom of the conference, as you'll soon see, things got much worse for Rutgers. 

Here are YPP numbers for the 2001 Big East. 
Miami was again clearly a cut above the rest of the conference although Boston College, Pittsburgh, and Virginia Tech all finished with a YPP Net greater than 1.00. I was surprised that Syracuse, which finished all alone in second place in the conference standings, and had the Big East Defensive Player of the Year in Dwight Freeney finished with the second worst per play defense and was actually underwater in terms of YPP Net. Finally, at the bottom of the YPP rankings, Rutgers may have boasted the worst BCS/Power 5 conference team of the 21st century. I have YPP data for every conference going back to 2005, and Rutgers outpaces all the other BCS/Power 5 teams by a significant margin. 
Most of that is due to an impossibly bad offense. The Scarlet Knights scored three touchdowns in seven conference games, were shut out three times, scored five points against Temple, and averaged just north of five points per game in conference play. Their per play offensive numbers were also significantly worse than any BCS/Power Five conference team between 2005 and 2020. 
Here are the 2002 Big East standings. 
For the third year in a row, Miami finished Big East play with an unblemished record and for the third year in a row, a different team finished in second place with a 6-1 record. West Virginia, under second year coach Rich Rodriguez rebounded from a rough non-conference start (lost to Wisconsin and Maryland by a combined 48 points) to finish with a winning conference record for the first time since 1998. Rutgers once again finished all alone in last place as they extended their conference losing streak to 22 games. 

Here are the YPP numbers for the 2002 Big East. 
Miami was once again head and shoulders above the rest of the Big East. Their YPP Net was more than double that of the second place team (Pittsburgh). In an odd turn of events, Temple, despite a 2-5 conference record, finished with a solid per play defense, ranking fourth in the Big East behind Defensive Player of the Year Dan Klecko. On the other hand, for the first time in a long time, Virginia Tech put a bad defense on the field, ranking second to last in the Big East in yards allowed per play. The Hokies lost two conference games where they scored 42 and 45 points respectively. And while Rutgers improved in Net YPP, their offensive YPP was even worse than the year before!

Here are the 2003 Big East standings. 
2003 was the year a pair of streaks were broken. After winning 27 consecutive Big East games, Miami was surprisingly blown out in Blacksburg on November 1st. The Hurricanes turned the ball over four times and were down 31-0 entering the fourth quarter. It marked Miami's first conference loss since 1999 when they also lost in blowout fashion during a trip to Blacksburg. To test your knowledge of late 90's Big East football, do you know the last Big East team other than Virginia Tech to defeat Miami? I'll reveal the answer at the end of this post. Elsewhere in the Big East, Rutgers broke their 25 game conference losing streak the week before Miami's winning streak was snapped. The Scarlet Knights blew out Temple and capped their year with a second Big East victory against Syracuse. 

Here are the YPP numbers for the 2003 Big East. 
For the fourth year in a row, Miami was tops in YPP Net, but they were not nearly as dominant on offense. The defense was still national championship worthy (in fact it was better than the 2002 version), but the offense fell off a cliff with the loss of Ken Dorsey, Willis McGahee, and Andre Johnson. Quarterback Brock Berlin took my breath away by throwing 17 interceptions (against just 12 touchdowns) and despite the presence of future NFL skill players Frank Gore, Jarrett Payton, Roscoe Parrish, and Kellen Winslow Jr., the team finished 51st nationally in points per game. 

Just how much of a problem was the Miami offense in 2003? The following table lists the number of offensive touchdowns Miami scored and allowed in league play between 2000 and 2003 as well as the number of turnovers the Hurricanes committed. 
With Ken Dorsey at the helm, the Hurricanes averaged at least four offensive touchdowns per game each season, maxing out at a little over five per game during his senior season. They also scored two or fewer offensive touchdowns just twice (including the infamous Boston College game where they did not manage a single touchdown until Ed Reed's return). In 2003, they scored two or fewer offensive touchdowns in four of their seven conference games. Their cumulative touchdown total would look even worse if we ignored the Temple game. The Hurricanes dominated the Owls, scoring seven offensive touchdowns, but managed just 13 in their other six Big East games. Their turnover numbers were also much worse once Dorsey departed, as the team committed twice as many turnovers in 2003 as they did in 2002. Football is a team game, but I'm good putting this all at the feet of Brock Berlin. 

That's how Miami's run in the Big East came to an end. They entered the 2004 season as co-favorites in the ACC with Florida State, but lost three games, including two at home en route to a disappointing Peach Bowl appearance. The decline continued the next two seasons with another Peach Bowl appearance (this one a shellacking at the hands of LSU) and finally a losing conference record in 2006 that got Larry Coker canned. Considering Alabama's run of dominance is going into its 14th season, it really puts the length of Miami's run into perspective. Depending on whether you consider 2004 part of the run, it really only lasted four or five years, or roughly the same length of time that a superstar in another sport turned South Beach into a mini-dynasty. 

And before we leave Miami, lets answer that trivia question from before. Other than Virginia Tech, the last Big East team to beat Miami was...
Syracuse. The week before Miami ended UCLA's run to the inaugural BCS Championship, the Hurricanes lost in blowout fashion to Donovan McNabb and the Orange in the Carrier Dome. 

Saturday, June 19, 2021

YPP Throwback: The 2001 SEC

One of the forgotten greats teams of the 21st century is the 2001 Florida Gators. While the Gators didn't even win their division, much less the SEC, I believe they were the only team from that season that could have given Miami a run for their money. I have YPP data going back to 2005, but thanks to the wonderful folks at College Football Reference, it took me less than an hour to calculate it for the 2001 SEC. With that season now twenty years in the rearview, I thought now would be a great time to take a look back at arguably the best (college) football team Steve Spurrier ever put on the field. 

First, here are the obligatory SEC standings. 
As I mentioned in the introduction, Florida didn't even win their division in 2001. The Gators lost at home to Tennessee in a game that was rescheduled due to the 9/11 terrorist attacks. The loss put the Volunteers into the SEC Championship Game where they lost to an LSU team they had beaten earlier in the season. LSU's victory gave a long forgotten coach named Nick Saban his first SEC title and marked the last time the SEC champion finished the season with three conference losses. 

And here are the Yards Per Play (YPP), Yards Per Play Allowed (YPA), and Net Yards Per Play (Net) numbers for each SEC team. This includes conference play only with the championship game not included. The teams are sorted by division by Net YPP with conference rank in parentheses. 
Based on the YPP numbers, Florida was head and shoulders above the rest of the SEC in 2001. The Gators averaged over seven yards per play on offense and were nearly a full yard per play ahead of the second most prolific offense (LSU). The Gators also boasted the conference's top defense and their Net rating of +2.79 was nearly a full two yards per play better than Tennessee. 

This is where I usually post a table showing which teams over or underperformed relative to their expected record based on YPP, but instead of pointing out some close game records, I wanted to try and put the Florida offense into some historical context. How do those per play numbers compare to all SEC offenses this century? Eleven SEC teams have averaged at least seven yards per play over the course of a conference season since 2000. Seven have occurred in the past four seasons and the other ten outside of the 2001 Gators have all happened since 2010. In other words, before 2018, this Florida offense was the best on a per play basis this century and likely in SEC history. Here are all eleven sorted from highest to lowest. 
Florida also did a pretty good job of turning all those yards into points. The Gators scored 41 offensive touchdowns in SEC play making them just the eighth SEC team this century to average at least five offensive touchdowns per game in league play. And six of the other seven instances have come in the past three seasons. Here are all eight sorted from highest to lowest. 
If you just went by the yards and points, you may have confused the SEC for the Big 12 in 2020. Three offenses averaged at least five offensive touchdowns per game and over seven yards per play! But back to Florida. That the 2001 Gators still rank as one of the SEC's best offenses twenty years later considering the modern offensive environment is amazing. Steve Spurrier really was ahead of his time, especially considering this offense took the field the same year that men like the atavistic Lou Holtz and Houston Nutt were coaching SEC programs. 

So how did this Florida team manage a silver medal in the SEC East? The offense was great, but it had a bit of a turnover problem. Florida turned the ball over twenty times in eight conference games, including five in an upset loss against an Auburn team that lost five games (more on that later). My abacus tells me that is exactly 2.5 turnovers per game. Compared to the other great per play offenses, the Gators were the worst at turning the ball over. 
This table illustrates how efficient offenses have become. Eventual national champions LSU in 2019 and Alabama in 2020 combined to average nearly eight yards per play and turn the ball over just once per game! In fact, no other offense that broke the seven yards per play barrier averaged more than two turnovers per game. 

Going back to that Auburn game, that loss, not the home loss to Tennessee, and by extension, all those turnovers, may have been what cost Florida a shot at the national title. Let's consider an alternate 2001 campaign where the Gators don't play their best, but manage to edge Auburn in mid-October instead of losing to them. Since they were ranked first in the AP Poll at the time of the game, they are likely still number one heading into their rescheduled game against Tennessee. The Gators drop that game to the Volunteers and in the process, Tennessee still wins the division and advances to the SEC Championship Game. However, late in the 2001 regular season, the best thing you could do was not play. Unbeaten Nebraska dropped their regular season finale to Colorado and the Buffaloes advanced to the Big 12 Championship Game. In that game, they avenged an earlier regular season loss to Texas and eliminated the Longhorns from BCS title game contention. Similarly, LSU's upset of Tennessee in the SEC Championship Game gave the Volunteers their second loss and eliminated them from BCS title contention. When the Championship Week smoke cleared, Miami was a clear cut number one in the final BCS standings and their were three contenders to face them in the Rose Bowl: Nebraska, Colorado, and Oregon. Despite a head to head loss to Colorado, Nebraska edged the Buffaloes by five one hundredths of a point to finish second in the BCS standings. For what its worth, despite owning the aforementioned loss to Auburn, Florida still finished fifth in the final BCS standings. 
Take away the turnover fueled loss to Auburn, and I think Florida is the non-champion that is able to back their way into the Rose Bowl. While the 'it just means more' SEC narrative did not exist in 2001, the Gators regular season dominance probably would have propelled them to their first ever Rose Bowl. 

Before we close the book on the 2001 Gators, I thought I would point out the relative dearth of NFL stars on such a dominant college football team. The 2001 team produced an impressive 22 players who spent time on NFL rosters, but Lito Sheppard was the only one to ever make a Pro Bowl. And the best pro from the offensive side of the ball? That's either Rex Grossman or Jabar Gaffney. The fact that these players didn't pan out as pros shouldn't take away from their accomplishments in college, but it is surprising such an explosive college offense (especially in the SEC) didn't produce a single Pro Bowl offensive player.

Thursday, June 03, 2021

2020 Adjusted Pythagorean Record: Sun Belt

Last week we looked at how Sun Belt teams fared in terms of yards per play. This week, we turn our attention to how the season played out in terms of the Adjusted Pythagorean Record, or APR. For an in-depth look at APR, click here. If you didn’t feel like clicking, here is the Reader’s Digest version. APR looks at how well a team scores and prevents touchdowns. Non-offensive touchdowns, field goals, extra points, and safeties are excluded. The ratio of offensive touchdowns to touchdowns allowed is converted into a winning percentage. Pretty simple actually.

Once again, here are the 2020 Sun Belt standings.
And here are the APR standings with conference rank in offensive touchdowns, touchdowns allowed, and APR in parentheses. This includes conference games only with the championship game excluded. Since teams played a varied number of games (the only scheduled conference game that was not played was between Louisiana-Monroe and Troy), the rankings are on a per game basis, not raw totals. 
Finally, Sun Belt teams are sorted by the difference between their actual number of wins and their expected number of wins according to APR.
I use a game and a half as an arbitrary line of demarcation to determine if a team significantly over or under-performed relative to their APR. By that standard, no team saw their record differ significantly from their APR. 

East Vs West
The Sun Belt is new to divisional play, having split into separate divisions in 2018. In that short time frame, the East has dominated the West, going 39-20 in interdivision play. The East was especially dominant in 2020, posting a 15-4 record against West division opponents, with Louisiana-Lafayette accounting for three of the West's quartet of victories. 
While the East won nearly 80% of interdivision games in 2020, that doesn't do justice to the extent they dominated their rivals in the West. Their four losses all came by a touchdown or less and their total margin of defeat in those four games was just fifteen points. If we look at per game scoring margin, East teams beat West teams by nearly two touchdowns per game on average!
While we missed out on a Sun Belt Championship Game in 2020, the East is 2-0 in the game's brief history. In that span, Louisiana-Lafayette has won the West each season. Perhaps not surprisingly, they are the only West division team with a winning record against the East in that span. If the West is to make up any ground in this nonexistent challenge, the other four teams in the division must start pulling their weight.  

That concludes our ten conference, twenty post slog through the beginning of the offseason. We are a little more than twelve weeks from some hot Week Zero action. Between now and then, posts will be more sporadic (think every two or three weeks instead of every week), but look for some First Half Point Differential posts on the 2020 college football season as well as a few YPP and APR throwbacks. As always, thanks for reading and enjoy your summer. 

Thursday, May 27, 2021

2020 Yards Per Play: Sun Belt

We finally made it to the last of the YPP recaps. This week we look back at the Sun Belt.   

Here are the 2020 Sun Belt standings.
So we know what each team achieved, but how did they perform? To answer that, here are the Yards Per Play (YPP), Yards Per Play Allowed (YPA) and Net Yards Per Play (Net) numbers for each Sun Belt team. This includes conference play only, with the championship game not included. The teams are sorted by division by Net YPP with conference rank in parentheses.
College football teams play either eight or nine conference games (typically fewer in 2020, although the Sun Belt almost played all their conference games in 2020). Consequently, their record in such a small sample may not be indicative of their quality of play. A few fortuitous bounces here or there can be the difference between another ho-hum campaign or a special season. Randomness and other factors outside of our perception play a role in determining the standings. It would be fantastic if college football teams played 100 or even 1000 games. Then we could have a better idea about which teams were really the best. Alas, players would miss too much class time, their bodies would be battered beyond recognition, and I would never leave the couch. As it is, we have to make do with the handful of games teams do play. In those games, we can learn a lot from a team’s YPP. Since 2005, I have collected YPP data for every conference. I use conference games only because teams play such divergent non-conference schedules and the teams within a conference tend to be of similar quality. By running a regression analysis between a team’s Net YPP (the difference between their Yards Per Play and Yards Per Play Allowed) and their conference winning percentage, we can see if Net YPP is a decent predictor of a team’s record. Spoiler alert. It is. For the statistically inclined, the correlation coefficient between a team’s Net YPP in conference play and their conference record is around .66. Since Net YPP is a solid predictor of a team’s conference record, we can use it to identify which teams had a significant disparity between their conference record as predicted by Net YPP and their actual conference record. I used a difference of .200 between predicted and actual winning percentage as the threshold for ‘significant’. Why .200? It is a little arbitrary, but .200 corresponds to a difference of 1.6 games over an eight game conference schedule and 1.8 games over a nine game one. Over or under-performing by more than a game and a half in a small sample seems significant to me. In the 2020 season, which teams in the Sun Belt met this threshold? Here are Sun Belt teams sorted by performance over what would be expected from their Net YPP numbers.
Using the .200 standard, Coastal Carolina and South Alabama overachieved relative to their YPP numbers. However, both teams went about it quite differently. Coastal was a good team that ended up with a great record thanks to winning both of their one-score conference games. Meanwhile, South Alabama was a bad team that finished with a middling record. The Jaguars were not especially lucky, finishing with an even 1-1 record in one-score games, but they really wheezed to the finish line after winning their first two conference games. South Alabama was far from dominant in those first two victories, being outgained by about .10 yards per play. However, over their final six contests they were outgained by more than two yards per play. Their overall performance was more indicative of a team that won a single conference game versus one that won thrice as many. The administration at South Alabama was not fooled as they relieved head coach Steve Campbell of his duties after the season was over. 

What are the Chants of a Repeat?
First off, apologies for the bad pun. As I mentioned earlier, the Sun Belt almost played a full conference season in 2020. The only regular season game that did not get played was between Louisiana-Monroe and Troy. Of course, just before the Sun Belt Championship Game, some positive tests and contact tracing prevented it from being played. As a result, we did not get to see a rematch of Costal Carolina and Louisiana-Lafayette. For Louisiana-Lafayette, 2020 was the third consecutive year they won the West division of the Sun Belt. For Coastal Carolina, it marked their best season as an FBS program and cemented their status as one of the most improved teams in all of college football. The Chanticleers improved their conference win total by six full games. In 2019, the Chanticleers won two Sun Belt games, beating Texas State and Troy by a combined four points. In 2020, they went 8-0, becoming the first Sun Belt team to roll through conference play unbeaten since Georgia Southern in 2014. Their six-win improvement is tied for the second largest increase among non-BCS/Group of Five conference teams since 2005 (behind Fresno State's amazing seven win-improvement in 2017). However, at this blog, we don't live in the past. Time marches on, and so do we. What can we expect from Coastal Carolina this season? To help answer that question, I looked at all non-BCS/Group of Five teams that improved their conference record by at least five games from one season to the next. Sixteen teams fit this criteria (an average of about one per season). They are listed in the table below along with their performance the following season. 
With a few exceptions, most notably Air Force in 2015, these teams tended to decline the following season. This makes sense. For teams that don't recruit elite talent, regression's pull is hard to avoid. These sixteen teams declined on average by a little more than two conference games the next season. Take heart Coastal fans. Even if the Chanticleers suffer a similar decline, they would still be in contention for the Sun Belt East title in 2021. In addition, while it would be impossible for Coastal to improve on their 2020 conference win total, perhaps they can join the other four teams that managed to hold their win total steady. Coastal Carolina was a great story in 2020 and they should contend for another division title in 2021, but were I betting on the division, I would probably take a look at a team that plays at a higher elevation.  

Thursday, May 20, 2021

2020 Adjusted Pythagorean Record: SEC

Last week we looked at how SEC teams fared in terms of yards per play. This week, we turn our attention to how the season played out in terms of the Adjusted Pythagorean Record, or APR. For an in-depth look at APR, click here. If you didn’t feel like clicking, here is the Reader’s Digest version. APR looks at how well a team scores and prevents touchdowns. Non-offensive touchdowns, field goals, extra points, and safeties are excluded. The ratio of offensive touchdowns to touchdowns allowed is converted into a winning percentage. Pretty simple actually.

Once again, here are the 2020 SEC standings.
And here are the APR standings with conference rank in offensive touchdowns, touchdowns allowed, and APR in parentheses. This includes conference games only with the championship game excluded. Since teams played a varied number of games (every team in the SEC played at least nine and most of the teams played ten), the rankings are on a per game basis, not raw totals. 
Finally, SEC teams are sorted by the difference between their actual number of wins and their expected number of wins according to APR.
I use a game and a half as an arbitrary line of demarcation to determine if a team significantly over or under-performed relative to their APR. By that standard, a pair of Tigers in Auburn and Missouri significantly overachieved. Auburn was only 2-1 in one-score conference games, but they also scored three non-offensive touchdowns in SEC action while allowing none. Two of those non-offensive touchdowns proved decisive in helping the Tigers snatch victory from the jaws of defeat. A blocked punt that resulted in a touchdown provided the winning margin against Arkansas, while a long interception return touchdown resulted in a likely fourteen-point swing in their thirteen-point win against Tennessee. Meanwhile Missouri managed a 3-0 record in one score conference games to eke out a .500 record despite allowing ten more touchdowns than they scored in SEC play.  

Did the SEC Become the Big 12 (But With Better Players)?
One season after LSU made their a case as the best college football team of the new millennium and perhaps by extension, all time, and two seasons after Clemson did the same, Alabama fielded arguably the best college football team ever in 2020. With all the testing, disjointed schedules, canceled games, and other assorted issues that go along with contesting a college football season in the midst of a global pandemic, its debatable how closely 2020 can be compared to previous seasons. However, you would have to be willfully ignorant to downplay how dominant Alabama was in 2020. The Crimson Tide won twelve of their thirteen games by double digits and those thirteen opponents all happened to be Power Five teams, so there were no cupcakes on the schedule. In the process of winning yet another national title, Alabama became the thirteenth Power Five/BCS conference team to average at least six offensive touchdowns per game in conference play in the APR era (since 2005). The Crimson Tide also scored the most offensive touchdowns of any Power Five/BCS conference team in league play, but since they got to play ten conference games, I'll do my best Ford Frick impression and sully their accomplishment with an asterisk.
If you take a look back at the APR standings a few lines up, you might notice Alabama was not the only SEC offense that was humming in 2020. Ole Miss and Florida also averaged five or more offensive touchdowns per game and five SEC defenses allowed at least four touchdowns per game. In fact, for the first time since I have been tracking APR data (since 2005), SEC teams actually scored more offensive touchdowns in conference play than Big 12 teams. 
Is 2020 an outlier, the result of an odd season played in challenging circumstances, or is this the new normal for the SEC? The SEC may not always outpace the Big 12 in scoring (2020 was their lowest offensive touchdown production in a decade), but SEC teams did average an extra touchdown per game compared to their output in 2005. With the conference consistently getting the best players and the rules of the game tilting ever more in the favor of offense, its hard to envision things going back to the relative Dead-ball Era the SEC experienced in the late aughts. 

Thursday, May 13, 2021

2020 Yards Per Play: SEC

We are hitting the home stretch of our YPP reviews. This week, we examine the SEC.   

Here are the 2020 SEC standings.
So we know what each team achieved, but how did they perform? To answer that, here are the Yards Per Play (YPP), Yards Per Play Allowed (YPA) and Net Yards Per Play (Net) numbers for each SEC team. This includes conference play only, with the championship game not included. The teams are sorted by division by Net YPP with conference rank in parentheses.
College football teams play either eight or nine conference games (typically fewer in 2020, although the SEC bucked that trend). Consequently, their record in such a small sample may not be indicative of their quality of play. A few fortuitous bounces here or there can be the difference between another ho-hum campaign or a special season. Randomness and other factors outside of our perception play a role in determining the standings. It would be fantastic if college football teams played 100 or even 1000 games. Then we could have a better idea about which teams were really the best. Alas, players would miss too much class time, their bodies would be battered beyond recognition, and I would never leave the couch. As it is, we have to make do with the handful of games teams do play. In those games, we can learn a lot from a team’s YPP. Since 2005, I have collected YPP data for every conference. I use conference games only because teams play such divergent non-conference schedules and the teams within a conference tend to be of similar quality. By running a regression analysis between a team’s Net YPP (the difference between their Yards Per Play and Yards Per Play Allowed) and their conference winning percentage, we can see if Net YPP is a decent predictor of a team’s record. Spoiler alert. It is. For the statistically inclined, the correlation coefficient between a team’s Net YPP in conference play and their conference record is around .66. Since Net YPP is a solid predictor of a team’s conference record, we can use it to identify which teams had a significant disparity between their conference record as predicted by Net YPP and their actual conference record. I used a difference of .200 between predicted and actual winning percentage as the threshold for ‘significant’. Why .200? It is a little arbitrary, but .200 corresponds to a difference of 1.6 games over an eight game conference schedule and 1.8 games over a nine game one. Over or under-performing by more than a game and a half in a small sample seems significant to me. In the 2020 season, which teams in the SEC met this threshold? Here are SEC teams sorted by performance over what would be expected from their Net YPP numbers.
LSU and Texas A&M significantly exceeded their expected record based on YPP while Arkansas significantly under-achieved. LSU and Texas A&M combined for a 5-1 record in one-score conference games, leading to their better than expected records. Meanwhile, Arkansas finished 1-3 in one-score conference games, narrowly missing out on program building wins against Auburn, LSU, and Missouri. Before we move on, I wanted to note how bad LSU's defense was and how amazing it is that they managed to finish .500 in conference play. The Tigers allowed over seven yards per play (and they weren't even the worst defense in the conference) and this wasn't the result of a bad game or two skewing their numbers. The Tigers allowed at least seven yards per play in eight of their ten conference games! Despite being routinely gashed, the defense did force a league-best 22 turnovers in SEC play, including nine in their last two regular season games, helping them scrape past two teams they probably had no business beating (Florida and Ole Miss). 

Is the SEC Championship Game Loser an Auto-Fade in Bowl Season?
Did you watch the Cotton Bowl this year? I went to bed at halftime. And I didn't miss much. Oklahoma led Florida 31-13 at the break and ended up winning the game by five touchdowns. Conventional wisdom posits this is one of the irrefutable facts of bowl season: The SEC Championship Game loser lays an egg in their bowl game. Despite the fact that I live in South Carolina, we're into facts at this blog, so I did a deep dive on the SEC silver medalists. 

The SEC Championship Game has been around since 1992, but I limited my analysis to the BCS and College Football Playoff eras (since 1998). We'll start with the BCS era (1998-2013). Here is how those sixteen SEC Championship Game losers fared in their bowl games. 
No matter which way you parse the results (either straight up or against the spread), they don't appear to be a great deal different from flipping a coin. However, there is one thing this generic analysis leaves out. The reason SEC Championship Game losers have a reputation for under-performing in their bowl game is because they are disappointed to not be playing for a national championship. Having to play in the Capital One Bowl instead of the BCS Championship Game is a recipe for a flat performance, or so the theory goes. What if we only looked at SEC Championship Game losers that had legitimate national title aspirations going into the SEC Championship Game? By my count, there were only five such teams in the BCS era: Tennessee in 2001, Alabama in 2008, Florida in 2009, Georgia in 2012, and Missouri in 2013. Outside of Alabama, those teams did surprisingly well in their bowl games. 
SEC Championship Game losers don't appear to have performed more poorly than expected in the BCS era. Maybe things have changed in the Playoff era. With four teams now able to compete for a national title in bowl season, perhaps the SEC Championship Game loser is more despondent and not motivated for their bowl game. In the Playoff era, here is how the seven SEC Championship Game losers have fared in their bowl games. 
As in the BCS era, the overall results don't appear to be much different than what we might find if we flipped a coin to determine the outcomes. Again though, not all SEC Championship Game losers are created equal. What if we limit our analysis to those teams with legitimate national championship aspirations entering the SEC Championship Game. There have been four such teams in the Playoff era: Auburn in 2017, Georgia in 2018, Georgia in 2019, and Florida in 2020. Their bowl performance has been mixed at best.
In the Playoff era, SEC Championship Game losers with national title aspirations have not fared well. However, four games is not nearly a large enough sample to offer any definitive proclamations. Instead of auto-fading the SEC Championship Game loser, you would be much better served reading the tea leaves before their bowl game. Florida had a host of opt-outs prior to the Cotton Bowl and the betting line in the game reflected that. At full strength on a neutral field, Florida would have likely been a slight favorite against Oklahoma. As the spread was more than a touchdown in the other direction, its obvious something a little stronger than motivation was influencing the number.