Thursday, April 25, 2019

2018 Adjusted Pythagorean Record: Mountain West

Last week we looked at how Mountain West 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 2018 Mountain West 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.
Finally, Mountain West teams are sorted by the difference between their actual number of wins and their expected number of wins according to APR.
Using a game and a half as a line of demarcation, Hawaii was the lone team that saw their actual record differ significantly from their APR. The Warriors also exceeded their expected record relative to their YPP and we discussed a few reasons for this last week. Rather than rehash that, let's delve into another characteristic Hawaii has managed to maintain for the last decade or so.

Consistently Inconsistent
Hawaii’s appearance in the 2018 postseason was shocking. The Warriors entered 2018 fresh off a 3-9 season where they won just a single time in Mountain West play. Yet the Warriors opened the season by winning a road game for the first time in nearly a calendar year. They followed that road upset with a pair of home wins, and following a body-clock loss at Army won three in a row to stand 6-1. Their wins were a little fluky, and the Warriors came back to earth over the second half of the season, losing four of six to finish the regular season 8-5. Still, eight wins was a massive improvement and the Warriors ended up finishing 5-3 in the improved Mountain West. However, perhaps it shouldn’t have been so shocking. Two years previous, the Warriors were starting the 2016 campaign fresh off a winless conference season with a first-year head coach. They managed to grind out four conference wins and a surprise bowl appearance (and victory). In fact, the Warriors have been treating (torturing?) their fans with a year-to-year roller coaster ride for the past decade or so. Based on year-to-year differences in conference victories, Hawaii has been the most inconsistent mid-major team since 2007. To illustrate this, I have charted their number of conference wins below.
From 2007 to 2008, the Warriors went from eight conference wins to five, so their absolute difference in wins was three. From 2008 to 2009 they went from five conference wins to three, so their absolute difference in wins was two. Adding these absolute differences together produces a total absolute difference of five. Using this formula through the 2018 season yields an absolute difference of 33, which equates to an average of three wins per year difference in conference record. Only one other mid-major school has come close to being as inconsistent as Hawaii.
So the Mountain West is home to the most inconsistent team of the last decade. It is also home to the two most consistent teams, at least in terms of average difference between conference wins.
Boise State has consistently finished near the top of whatever conference they happen to be a part of, never winning fewer than five conference games since 1998. Nevada has consistently finished within one game of .500 in conference play with a few exceptions sprinkled in when Colin Kaepernick was under center (or more precisely in The Pistol). Heading into 2019, expect Boise State and Nevada to finish within about a game of their 2018 conference record. As for Hawaii, they should either go undefeated and challenge for a playoff spot or finish winless and be in the market for a new head coach.

Thursday, April 18, 2019

2018 Yards Per Play: Mountain West

Our offseason sojourn keeps churning through the Group of Five. This week we are examining the Mountain West.

Here are the Mountain West 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 Mountain West 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. 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 2018 season, which teams in the Mountain West met this threshold? Here are Mountain West teams sorted by performance over what would be expected from their Net YPP numbers.
Boise State and Hawaii were the Mountain West teams that significantly exceeded their expected record based on YPP. The reason for Hawaii’s resurgence was simple: they were great in close games. The Warriors were 4-0 in one-score conference games and their other conference win came by just nine points. Contrast that to their losses which all came by at least eighteen points. For Boise State, no one thing stands out, but coupled together, several small factors helped them exceed their expected record. The Broncos were a solid 2-1 in one-score conference games, had a +6 turnover margin in conference play, and were +3 in non-offensive touchdowns in conference play. One non-offensive touchdown proved to be the difference in them winning the division. A 99-yard interception return against Nevada provided the margin of victory in their win against the Wolfpack and allowed them to steal the division when they beat Utah State in the regular season finale.

A Step-Back for Boise?
After compiling the YPP data for 2018, I was surprised Boise State ranked so low. The Broncos were not a bad team, but they did not have the profile of a dominant team either. Yet on the first weekend in December, they were once again hosting the Mountain West Championship Game for the second year in a row and third time in the game’s six year existence. To contextualize just how much Boise State’s actual record exceeded their expected record based on YPP, I went back and looked at all mid-major (non-BCS and Group of Five) teams since 2005 that exceeded their expected record by at least .300 (remember, I consider a difference of .200 significant) and finished with a conference winning percentage of at least .875 (7-1 in an eight game conference schedule). I added this extra qualifier because I wanted to look at teams that not only exceeded their expected record, but were also in contention for their conference title. In other words, I wanted average teams that finished with great records instead of bad teams that finished with average records. So how unique was Boise State in 2018? They were only the sixth mid-major team to exceed their expected record by at least .300 and win at least seven of their eight conference games. Boise and the other five teams are listed below along with their conference records the following season.
Obviously, this reeks of small sample size, but the results should not be encouraging for Boise State fans. The other five teams on this list all declined by at least two games in conference play the next season and the average decline was over three games (3.1). The Broncos do have a better and longer track record than these five teams. All five were coming off losing seasons when they significantly exceeded their expected record, with UCF fresh off a winless campaign when they surprised the nation by nearly winning Conference USA in 2005. Meanwhile, the Broncos have not won fewer than eight games in two decades. Still, they are losing a four-year starter at quarterback and their defensive YPP was the worst (in terms of actual yards per play and conference rank) it has been in the fourteen years I have been tracking YPP data. The Broncos will likely open as the Mountain West favorite when odds are released in the coming months, and they will be an even more prohibitive favorite to win the Mountain Division. I wouldn’t place a bet on one of the other teams in the division (Air Force, Colorado State, New Mexico, Utah State, Wyoming) to win it individually, but collectively, I would take the field over the Broncos to represent the Mountain half of the conference in the championship game.

Thursday, April 11, 2019

2018 Adjusted Pythagorean Record: MAC

Last week we looked at how MAC 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 2018 MAC 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.
Finally, MAC 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 measure to determine if a team over or under-performed relative to their APR. By that standard, in 2018, no MAC team significantly over or under-performed so we'll move on to a more interesting conversation.

The Dean of MAC Coaches
Mid-major (Group of Five) coaches at the FBS level typically follow one of two career paths. They succeed and move up to a bigger job (another mid-major or Power Five) or they fail and get fired and return to their previous life as an assistant coach. There are exceptions to the rule of course. Sometimes these coaches leave of their own accord to become assistants in the NFL or at a better college job, but for the most part these guys are promoted or fired depending on their success at the mid-major school. For this reason, you rarely see ‘lifers’ at mid-major programs. But in the closing act of Frank Solich’s career, he has become a mi-major lifer.

Beginning a half-century ago, Frank Solich certainly appeared to be a lifer. A Nebraska lifer. Solich played for legendary Nebraska head coach Bob Devaney in the 1960s, coached high school football in Nebraska until the late 1970s, and then became an assistant under another legendary Nebraska coach, Tom Osborne. When Osborne retired following the 1997 season, after winning three national titles in his four years, Solich succeeded him as head coach. Despite three-top ten finishes and a conference title in six seasons, Solich was Gene Bartowed following a 9-3 regular season in 2003. After a gap year, Solich returned to coaching in 2005 at Ohio. The Bobcats had gone just 11-35 in four seasons under Brian Knorr, but Solich had them in the MAC Championship Game in just his second season. The Bobcats dipped a bit in his third and fourth seasons (combined 10-14 record), but beginning in 2009, Solich has had the Bobcats bowl-eligible each of the past ten years. In that span, the Bobcats have made three additional appearances in the MAC Championship Game, won their first bowl game in school history (plus three more), and briefly appeared in the AP Poll for the first time since 1968.

While 2005 may not seem like that long ago, Solich is very close to becoming the longest-tenured MAC coach ever. Only four other men in history have lasted more than ten seasons as head coach at a MAC school. They are listed below.
First off, Ohio fans, don’t @ me. I know Bill Hess was the coach at Ohio from 1958-1977. However, in those first four seasons, Ohio was not an FBS school and the MAC was not an FBS conference. Northern Illinois fans, don’t @ me either. The Huskies were an FBS independent in Joe Novak’s first season (1996). Solich has been Ohio’s head coach for fourteen seasons, so if he is still the head coach in December of 2020, he will tie Hess and Herb Deromedi from Central Michigan as the longest tenured MAC coach. In addition, thanks to the twelve-game regular season, potential conference championship games, and the proliferation of bowl games, Solich already owns the MAC record for games coached and is just four wins shy of tying Deromedi for most wins all time as a MAC coach. However, thanks to Deromedi’s phenomenal conference record, Solich would probably need to coach at least four more seasons to tie or break the record for most MAC conference wins (he would be 78 in December 2022).

Solich has yet to reach the MAC mountaintop at Ohio, and realistically, he is probably running out of time to win Ohio’s first MAC title since 1968. I’d love to see the Bobcats win the MAC in the next year or two while simultaneously making Solich the dean of MAC coaches, not just in the present, but of all time.

Thursday, April 04, 2019

2018 Yards Per Play: MAC

After a two week break where we talked about basketball, this blog makes its triumphant return to the only sport that really matters - football. This week, we will be examining the MAC.

Here are the MAC 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 MAC 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. 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 2018 season, which teams in the MAC met this threshold? Here are MAC teams sorted by performance over what would be expected from their Net YPP numbers.
Buffalo was the lone MAC team to significantly exceed their expected record based on YPP. The Bulls were not exceptionally lucky (1-0 in one-score games), nor were they fueled by a fluky turnover margin (+1 in conference play). They simply were not quite as dominant as we might expect from a team that lost just once in eight conference games. On the flip side, three MAC teams significantly under-performed relative to their expected record based on YPP. Bowling Green, Central Michigan, and Kent State finished a combined 1-8 in one-score conference games and the lone victory came in a game played between these teams (Kent State edged Bowling Green in October). Those three teams were bad, they just weren’t quite as bad as their respective records.

The MAC LEast
While the MAC East champion lost in the MAC title game for the third straight year and seventh time in the last ten seasons, the division as a whole finally enjoyed a modicum of success against the MAC West. For the first time since 2009, the MAC East did not post a losing record against the MAC West in inter-division play.
Which MAC East teams have been most responsible for this ghastly record over the past nine seasons? It really has been an equal opportunity performance them. Between 2010 and 2018, eight teams spent time in the MAC East. Akron, Bowling Green, Buffalo, Kent State, Miami, and Ohio were fixtures in the division, but Massachusetts (2012-2015) and Temple (2010-2011) also vacationed there. How did those teams do against the MAC West?
No MAC East team posted a winning record against the MAC West. Ohio carried water best for the MAC East squads (although their bucket was littered with holes) as they almost won half their games against the MAC West. I find it interesting that Bowling Green is the lone MAC East team with multiple conference titles since 2010 (which obviously means they were able to beat a MAC West team in the conference title game), yet they have the worst record against the MAC West. 

Was 2018 an outlier or will the MAC East continue to close the gap on the MAC West and usher in a new era of competitive inter-division play? Stay tuned to ESPN2 on November weeknights to find out.

Thursday, March 28, 2019

Rumors of the Pac-12's Demise Have (Probably) Been Greatly Exaggerated

The Pac-12 has been the butt of a lot of jokes lately. Their football championship game had attendance numbers commensurate with a middle school JV game in Iowa, their football champion has missed the College Football Playoff two years running, their network appears to be going the way of the Prevue Channel (or the go-round channel as my mother calls it), and they may have been in real danger of sending a single team to the NCAA tournament this year. Fortunately, the Pac-12 was able to coax three bids to the tourney after Oregon won four games in four days to grab the automatic bid. I’m not here to suggest ways the Pac-12 can improve its championship game (move it out of Santa Clara), opine on when the conference will land a playoff bid, or offer advice on how to improve the Pac-12 Network, but I will give Pac-12 hoops fans reasons for optimism heading into 2020.

While college football has five ‘Power Conferences’, college basketball has six (and another two pseudo power conferences). Read my post from two years ago for some more background and justification on this (its really good!). Those six conferences (ACC, Big East, Big 10, Big 12, Pac-12, and SEC) account for the majority of at-large bids to the NCAA tournament. How many bids has each conference produced over the past 19 years? Glad you asked. The total number of NCAA tournament bids (at-large and automatic) for each power conference are listed below along with an average number of bids per season.
The Big East leads all power conferences with an average of six and a half tourney bids per season. The Big 10, Big 12, and ACC are bunched closely together and have received almost six bids per season, while the SEC has received about five per season and the Pac-12 is last at about four and a half. Alright, so things aren’t looking great for the Pac-12 thus far, but let’s come at this from another angle. Obviously a team that receives a number one seed in the NCAA tournament has been more successful than a team that receives a six seed. In the interest of rewarding better seeds, we’ll award what I have deemed 'seed points' for each tournament bid. A one seed receives sixteen seed points, a two seed receives fifteen seed points, and so on until the bottom of the bracket, where a sixteen seed receives one seed point. Using this methodology, how do the power conferences stack up?
Once again, the Big East is on top, receiving about 75 seed points on average per year. The ACC is a notch below with 70, the Big 12 and Big 10 are close together with 67 and 66 respectively, the SEC is further down the list with 55, and the Pac-12 again brings up the rear with 47. This just seems to further illustrate the point that the ostensible 'conference of champions' has struggled for quite some time. However, I think we need to make one more adjustment. Conference composition has changed a great deal since 2001. In 2001, the ACC had nine basketball playing members. Today, the conference has fifteen. That represents a 67% increase in membership size. To adjust for fluctuating membership size, let’s divide the total number of NCAA tournament bids for each conference by the average size of each conference over this 19 year period. Doing this gives us an average percentage of teams that qualify for the NCAA tournament for each conference.
When we make this adjustment, the Big 12 reigns supreme as on average, more than half of its member schools qualify for the NCAA tournament. The Big 10, Big East, and ACC are all just below 50%, and while the Pac-12 still rates low in this category, a new conference brings up the rear. A similar thing happens when we divide seed points by the average membership size for each conference.
The Big 12 is again tops by this measure with the ACC, Big East, and Big 10 finishing close together. The SEC and Pac-12 are virtually tied at the bottom of the standings. However, as I mentioned in the opening paragraph, there is hope for Pac-12 fans, and they can look to their neighbors as the bottom of this ranking, the SEC, for proof. Scroll back up and look at the first and second table. I’ll wait. Between 2013 and 2016, the SEC received a total of 14 NCAA tournament bids. In three of those seasons (2013, 2014, and 2016), they received just three bids. They also received the fewest seed points of any power conference in three of those four seasons. Yet, in the past two seasons, the SEC has secured fifteen bids to the NCAA tournament while ranking second and third in total seed points. How were they able to do this? They decided they wanted to be good and basketball. They threw money at the problem and hired good coaches. Since the end of the 2013-2014 season, six SEC schools have hired coaches with previous (successful) power conference head coaching experience and currently seven of the fourteen schools have a head coach with previous power conference head coaching experience (and this number doesn’t include John Calipari at Kentucky). If you allocate a few additional resources, success seems to follow. For further proof that the SEC is serious about basketball, look no further than the unemployment line where moderately successful head coaches just lost their jobs. Bryce Drew was fired at Vanderbilt after three seasons and one NCAA appearance. Maybe scratch him from the moderately successful list since the Commodores did finish 0-18 in SEC play this season. Mike Anderson was fired at Arkansas after three NCAA appearances in the last five seasons. Avery Johnson, a successful former NBA coach (and player) was let go by Alabama one season after making the NCAA tournament. Billy Kennedy was fired by Texas A&M despite two Sweet 16 appearances in the past four seasons. The SEC is serious about basketball now.

The ACC will never be the SEC’s equal in football, but the conference has gotten serious about competing in the sport. Clemson has won two of the last three college football national titles (and the conference as a whole has won three of six). Similarly, the SEC will probably not become the preeminent college basketball conference, but they are in much better shape than they were five years ago. Circling back around to the Pac-12, UCLA (arguably the top job in the conference) will hire a head coach in the coming weeks. A return to relevancy can be as simple as making the right hire in Los Angeles. The harder places to win in the conference (Colorado, Oregon State, and Utah) already have good coaches. Imagine if the school with a great location and history did the same?

Monday, March 18, 2019

2019 Bracket Advice: Don't Make Houston or Texas Tech Your Final Four Sleeper

While this is primarily a college football blog, around March or April over the past two years, I have deviated from the offseason script and penned a few college basketball pieces. Two years ago, I looked at how football expansion was making the March experience worse for mid-majors and last year, I examined college basketball tiers. Those posts both came out around Final Four weekend, but this year, I want to help you fill out your bracket. This post will point out some teams you should probably avoid christening as Cinderellas when you make your bracket picks. Next week, I’ll take a look at the ebbs and flows of at-large bids in the power conferences and hopefully talk Pac-12 fans down from bridges and overpasses. After that, we’ll resume football-centric posts and continue our slog through the ten FBS conferences with a look back at the MAC. As always, thanks for reading.

Which college basketball poll is more accurate: the preseason AP poll or the one put out mere days before the tournament starts? If you guessed the most recent poll, you are a moron you would be incorrect. Despite having a season’s worth of data, the final poll of the season is less accurate than the preseason poll in terms of identifying your national title contenders. Other researchers have written about this phenomenon and why it exists, so I won’t steal their thunder. However, I did conduct some research of my own in an attempt to identify teams that are more likely to be upset in the early rounds of the tournament. Read on for my analysis.

In 2001, the NCAA tournament expanded to 65 teams by adding a play-in game on the Tuesday before the real tournament starts. The play-in game pitted the two lowest rated teams in the field with the winner advancing to face prohibitive odds against a number one seed. While I dislike the play-in game and its successors, three other play-in games that were added when the field expanded to 68 teams in 2011, it is a nice arbitrary demarcation date. With that arbitrary date in mind, I decided to examine every tournament beginning with 2001. As the year is currently 2019, that means eighteen tournaments have been played since the field expanded. In those eighteen tournaments, there have been 171 teams to take the court while ranked in the most recent AP poll (and final as the AP does not vote on teams after the tournament) that were not ranked in the preseason AP poll. To determine how those teams fared in the tournament, I conducted a rudimentary examination of their expected number of wins based on their seed. For example, a number one seed is expected to win their quadrant of the bracket, so their expected number of wins would be four. Were they to lose in the first round (as Virginia famously did last season), they would finish four wins short of expectations. If they lost in the second round, they would finish three wins short and so on. As I mentioned, this analysis is very basic and ignores things like quality of opponent. When Virginia was vanquished in the first round, it made things much easier for the other teams in their quadrant, but that is not taken into account here. In addition, this analysis is only concerned with winning a quadrant (or region) of the bracket and advancing to the Final Four.

So how did those teams perform? Collectively, those 171 teams won about four tenths of a game less than would be expected based on their seed. While four tenths of a win may not sound like much, it means those teams are not living up to their seed in the aggregate. In addition, nearly half of those teams (85) finished at least one game worse (i.e. lost at least one round early) than would be expected based on their seed. Around 30% of those 171 teams (52) won exactly as many game as expected based on their seed and only 20% (34) performed above what would be expected. Those results are summarized in the table.
Still not convinced? Wait, there’s more. Since 2001, 91 teams seeded sixth or better have lost in the first round (about five per tournament on average). Of those 91 teams, 36 (nearly 40%) were ranked in the final AP poll despite not being ranked in the preseason AP poll. Those 36 teams are listed chronologically below. Relive your best (or worst March memories).

I’m not done yet. Since 2001, 56 teams seeded third or better have lost in the second round (about three per tournament). Of those 56 teams, 21 (about 38%) were ranked in the final AP poll despite not being ranked in the preseason AP poll. Once again, here they are. 
I try to be transparent around here, so I will point out the success stories of teams that entered the tournament ranked despite not appearing in the preseason poll. Six teams (out of 171) advanced to the Final Four. They are listed below along with the number of future NBA players on their respective teams.
Every team had at least two future NBA players (Michigan has an asterisk as some additional players from their 2018 runner-up squad may eventually play in the NBA) on their roster and some (looking at you Florida) had an embarrassment of riches. Obviously, the exception to this rule is if a team has a decent amount of future NBA talent, the fact they were not ranked in the preseason poll may not matter as much.

Hopefully I’ve convinced you that currently ranked teams that did not appear in the preseason poll are not good bets to go a long way in March (provided they don’t have a lot of NBA talent). But who fits the bill this year? Here are all the teams ranked in the most recent poll that were not ranked in the preseason poll along with their seed.
Texas Tech is the lone member of this group with a player likely to be drafted in June. Wofford's Fletcher Magee, Houston's Corey Davis, and Wisconsin's Ethan Happ may wind up on an NBA roster, but this group of teams is not brimming with NBA talent. I'm not implying they are all going to lose in the first round, but I would advise against predicting them to exceed the expectations associated with their seed.

Hopefully this analysis helps you win your bracket pool. If so, feel free to send me a percentage of your winnings. If not, well you shouldn’t believe everything you read on the internet.

Thursday, March 14, 2019

2018 Adjusted Pythagorean Record: Conference USA

Last week we looked at how Conference USA 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 2018 Conference USA standings.
And here are the APR standings sorted by division with conference rank in offensive touchdowns, touchdowns allowed, and APR in parentheses. This includes conference games only with the championship game excluded.
Finally, Conference USA 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 a line of demarcation and by that standard, no team significantly exceeded their APR. Two teams did finish at least a game and a half below their APR, but Florida Atlantic and UTEP also failed to live up to their expected record based on YPP and we went over some reasons why last week.

The Shifting Membership of Group of Five Conferences
Over the past decade or so, conference affiliation has been fluid. Massive realignment shifts at the top level of college football have caused significant aftershocks at the lower levels of the sport. Look no further than Conference USA for evidence of this. Despite barely being old enough to drink (the league began playing football in 1996), only one of the original six founding members is still in the conference. With that in mind, I decided to see just how stable conference membership is at the Group of Five level. Is Conference USA unique in its shifting membership or is this par for the course for the other mid-major conferences? To find out, I looked at the current (football playing) makeup of the four senior Group of Five conferences (I ignored the American as it was only founded in 2013) and calculated how many years the average team had been a continuous member as well as how many of the original members were remaining. I’ll go through each conference one by one starting with the most stable (defined as having the longest tenured average member).
Founded in 1962, the MAC is easily the oldest mid-major conference and has by far the longest-tenured average member. Nine of the twelve teams have been in the conference for at least forty years and all twelve have spent at least two decades as a member of the Big 10’s little brother. Note that the average membership length would be even greater if I included both tenures of Northern Illinois. The Huskies joined the MAC in 1975 and played for eleven seasons before leaving after the 1985 campaign. They were vagabonds for another eleven season, playing as an independent and as a member of the Big West before rejoining the MAC in 1997. Also note that six of the original seven MAC teams are still in the conference. Marshall was a founding member of the conference in 1962 and stayed for eight season before becoming an independent in 1970. The Herd later rejoined (and dominated) the MAC when they made the jump from I-AA (now FCS) in 1997.
Finishing with an average membership length merely three decades short of the MAC’s, the Mountain West is up next. The conference will be celebrating its 20th anniversary in 2019 and six of the original eight founding members are still there. The other two, TCU and Utah, moved up to Power Five conferences several years ago. Outside of those original six, the other half of the conference is filled with relative newcomers. Boise joined in 2011, Fresno State, Hawaii, and Nevada joined in 2012, and San Jose State and Utah State rounded out the membership in 2013.
The Sun Belt began sponsoring football in 2001, but only three of the original seven members are still in the conference. Idaho now plays football at the FCS level while New Mexico State has been banished to the waste land of independent football and Middle Tennessee and North Texas are now members of Conference USA. The Sun Belt has served as a landing spot for teams looking to get a slice of that FBS money. Appalachian State, Coastal Carolina, Georgia Southern, Georgia State, South Alabama, and Texas State have all either moved up from the FCS or started their football programs from scratch in the past decade.
Southern Miss is the elder statesman of Conference USA. The other five founding members are all in the ACC (Louisville) or the American (Cincinnati, Houston, Memphis, and Tulane). Outside of Southern Miss, only four other teams have been in the conference more than a decade (Marshall, Rice, UAB, and UTEP). The other members either graduated from the Sun Belt (FAU, FIU, Middle Tennessee, North Texas, and Western Kentucky), escaped the WAC before it imploded (Louisiana Tech and UTSA), or started their football programs in the last decade (Charlotte, Old Dominion, and UTSA again).

Despite being the second-oldest surviving mid-major conference (RIP WAC), Conference USA is the second youngest in terms of average length of membership (behind the six-year old American). Will the conference survive another round of realignment or become a footnote in history like its spiritual predecessor, The Metro? Only time will tell.

Thursday, March 07, 2019

2018 Yards Per Play: Conference USA

After six weeks of Power Five action, we return to the Group of Five with Conference USA.

Here are the Conference USA 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 Conference USA 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. 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 2018 season, which teams in Conference USA met this threshold? Here are Conference USA teams sorted by performance over what would be expected from their Net YPP numbers.
Middle Tennessee State significantly exceeded their expected YPP record and qualified for the championship game for the first time since joining Conference USA. The Blue Raiders nearly won their first outright conference championship as an FBS program but lost to UAB thanks in part to having too many men on the field during a punt return. Middle Tennessee was a solid 2-1 in one-score conference games (prior to their close loss in the title game), but turnovers and non-offensive touchdowns were the reason they were able to exceed their expected record. The Blue Raiders had a league-best in-conference turnover margin of +10 and they also scored three non-offensive touchdowns while allowing none. UTEP and Florida Atlantic were the two Conference USA teams that failed to live up to their YPP numbers. UTEP had a league-worst in-conference turnover margin of -12, finished 1-2 in one-score conference games, and allowed four non-offensive touchdowns while scoring none of their own. For Florida Atlantic, the culprit was close game performance. Of their five conference losses, four came by a touchdown or less. Those four close losses came by a combined fifteen points, while each of their three conference wins came by at least nineteen points.

From the Penthouse to the Outhouse
Going in to the 2018 season, Florida Atlantic was a prohibitive favorite in the east division of Conference USA. And rightfully so. The Owls dominated the league in 2017, winning all eight of their regular season games, with seven coming by double digits, drubbing North Texas in the Conference USA Championship Game, and embarrassing Akron in their bowl game on the way to a school-record eleven victories. Thanks to the, shall we say, interesting career path of their head coach, they even got to hang on to him for another year. A lot of things can go wrong over the course of a football season, but worst case scenario for 2018 seemed to be a second consecutive bowl game. Yet, when December rolled around, the Owls were home for the holidays, having posted a 5-7 overall record (just 3-5 in Conference USA). While it probably wasn’t what their fans were expecting, Florida Atlantic did put together a historic season in 2018. They became just the tenth team since 1998 (the beginning of the BCS era) to finish with a losing conference record one season after going unbeaten in league play. The others are listed below.
Florida Atlantic is even more unique as they, along with BYU, were the lone mid-major (non-BCS or Group of Five) team to post a losing record despite retaining their head coach. The other mid-major teams on this list (Ball State, Central Michigan, Kent State, and Tulane) lost their head coaches to better jobs after their undefeated campaigns or in the case of TCU, moved to a power conference.

Can Florida Atlantic expect to rebound to their previous status in 2019? To find out, I looked at the follow-up conference records for those nine teams. They are summarized below.
Six of the nine teams saw their conference win total improve after their collapse, two declined, and one finished with the same conference record. On average, the teams improved by about six tenths of a conference win the next season. However, if we remove TCU from the equation, as the circumstances surrounding their decline involved an uptick in competition, the results are a little better.
Six of the eight teams saw their conference win total improve, one declined, and one stayed the same. On average the teams improved by nearly one win in conference play the next season. Were I a Florida Atlantic fan, I would feel pretty good heading into 2019. Based on the admittedly small sample here, FAU can expect to improve, albeit probably not to the dizzying heights they reached in 2017. Still, a small improvement in conference play would get them back to a bowl and have them in contention for the division.

Thursday, February 28, 2019

2018 Adjusted Pythagorean Record: Big 12

Last week we looked at how Big 12 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 2018 Big 12 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.
Finally, the Big 12 teams are sorted by the difference between their actual number of wins and their expected number of wins according to APR.
Using one and a half games as a benchmark for under or over-performing relative to APR, TCU and Oklahoma were the Big 12 teams that significantly exceeded their expected record in 2018. Texas almost hit the mark and they were the lone team to significantly exceed their expected record based on YPP. We touched on a few reasons why last week, so I won't rehash that here. For TCU and Oklahoma, their problems were on opposite sides of the ball. TCU, as they always do under Gary Patterson, had an above-average Big 12 defense, but the worst offense (touchdown-wise) in the conference. Meanwhile, Oklahoma moved the ball at will, but could not stop anyone, scoring and allowing the most touchdowns in Big 12 play. Both teams were able to exceed their APR thanks to solid records in close games. TCU went 4-2 in one score-conference games while Oklahoma was 3-1 in such contests.

The Hrned Frgs Lost Their 'O'
I had to look twice while calculating the Big 12 APR data to make sure I hadn’t made a mistake. I could not believe that TCU ranked ninth in the conference and just ahead of lowly Kansas. But, as usual, my math checked out and if you watched any of the Cheez-It Bowl, you know why TCU ranked so low. The Horned Frogs scored just seventeen offensive touchdowns in Big 12 play. That means they averaged less than two offensive touchdowns per game over the course of their conference campaign. The Big 12 has a reputation as a defense optional conference so, how bad was TCU’s offensive performance in a historical Big 12 context? Glad you asked. The table below lists every Big 12 offense that averaged two or less offensive touchdowns per game in conference play since 2005. Spoiler alert: You’ll find a lot of Kansas entries on this list.
Since 2005, seventeen Big 12 teams have averaged two or less offensive touchdowns per game and eight of those teams have been headquartered in Lawrence, Kansas. While it is not quite as impressive as their Big 12 title streak in basketball, the Jayhawks did somehow average two or less offensive touchdowns for six straight seasons of Big 12 play before finding a modicum of offensive success in 2018.

Looking at touchdowns per game is one way of ranking these poor offenses, but we can also look at how consistent they were and what their best offensive performance was. I have recreated the previous table with two additional columns. The first lists the number of conference games where the team was held to one or zero offensive touchdowns and the second lists the most offensive touchdowns the team managed in any conference game that season.
I would like to direct your attention to two teams that really stand out: Baylor circa 2009 and Kansas in 2010. In 2009, sophomore Robert Griffin began the season at quarterback for Baylor. He was injured in their third game (a victory against Northwestern State), but even in his absence, the Bears finished non-conference play 3-1 and had designs on their first bowl bid since 1994. However, the Bears lost their first four conference games while scoring just 34 total points. Then miraculously, their offense exploded in a road trip at Missouri. The Bears scored 40 points and managed five offensive touchdowns in a shocking win. Alas, their offensive firepower was a mirage and the Bears scored just 30 total points in losing their last three games. Outside of that offensive explosion against Missouri, Baylor did not score more than one offensive touchdown in any of their other seven conference games.

Kansas opened the Turner Gill era in 2010 by scoring three points in a loss at home to North Dakota State. They rebounded to beat a ranked Georgia Tech team the next weekend and split their remaining two non-conference games against Southern Miss and New Mexico State. At 2-2, a bowl game was an unlikely proposition, but the Jayhawks had posted at least five wins in six of the seven previous seasons, so few could have predicted the depths to which the program would sink. Kansas opened conference play losing their first four games while failing to score more than 16 points in any one. They looked to be well on their way to a fifth straight loss as they trailed Colorado 45-17 at home early in the fourth quarter. Then suddenly, without warning, the offense found its rhythm. The Jayhawks scored five touchdowns in twelve minutes (four offensive and one fumble return) to shock the Buffaloes. The Jayhawks used the momentum they generated there to score 24 total points over their last three games. The Colorado game, especially the fourth quarter was an amazing outlier performance. In that fourth quarter, the Jayhawks scored four offensive touchdowns. In the other 31 quarters they played against Big 12 opponents that season, the Jayhawks scored nine offensive touchdowns.

Doing research, uncovering surprising statistics, and recalling events that had escaped my memory are the reasons I continue to run this blog. Baylor and Kansas were two bad and unremarkable teams, but their performances in one game and one quarter respectively were truly amazing, especially considering how those teams played the rest of the season. I’m not sure if he was the first to say it, as he may have only popularized it during his stint on ESPN, but Lou Holtz was fond of the adage ‘The kids on the other team are on scholarship too’. Baylor and Kansas are stark reminders that even bad teams have gifted athletes that are capable of remarkable performances.

Thursday, February 21, 2019

2018 Yards Per Play: Big 12

Three conferences down, seven to go. Next up, the Big 12.

Here are the Big 12 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 Big 12 team. This includes conference play only, with the championship game not included. The teams are sorted by Net YPP with conference rank in parentheses.
College football teams play either eight or nine conference games. 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 2018 season, which teams in the Big 12 met this threshold? Here are Big 12 teams sorted by performance over what would be expected from their Net YPP numbers.
Texas was the only Big 12 team that saw their actual record differ significantly from their expected record based on YPP. The Longhorns did have the best in-conference turnover margin in the Big 12, but it was hardly historic (+9). Texas was able to qualify for the Big 12 Championship Game thanks to some good fortune in close games. Seven of their nine conference games were decided by a touchdown or less, and the Longhorns won five of them. They won close games against good teams (Oklahoma), bad teams (Kansas), and everything in between (TCU).

Miles to Go
To no one’s surprise, Kansas was recently again in the market for a new football coach. To the surprise of some, the Jayhawks were able to coax former Oklahoma State and LSU coach Les Miles to take the job. Miles may not be the most tactically gifted football coach in history, but he does have a national championship on his resume. For a school that has spent the last decade as the worst BCS/Power Five program, that was quite a (public relations) coup. But what can Kansas fans reasonably expect from Miles? Have other national championship winning coaches been able to duplicate their success at a new locale? To answer that question, I looked at all FBS coaching hires since 1984 (what I deem the ‘modern era’) where the coach had at least one national championship at the FBS level. This means we are excluding coaches like Craig Bohl and Lance Leipold who won titles at lower levels before becoming FBS head coaches. By my count there are twelve head coaches (including Miles) that have been hired to lead FBS programs while having a national title on their resume. The other eleven are listed below along with their record at their new jobs.
That’s a pretty eclectic group of coaches. Let’s go through them one-by-one.

Larry Coker – Coker won the 2001 national title at Miami and nearly won a second in 2002. Diminishing returns led to his ouster after the 2006 season, but he returned to coaching at Texas-San Antonio and helped start the program from scratch. In four seasons of play at the FBS level, Coker guided the Roadrunners to two winning seasons and a 22-26 overall record.

Dennis Erickson – Erickson started his college head coaching career out west, leading Idaho, Wyoming, and Washington State before getting the Miami job. He won or shared two national titles at Miami before heading back to the west coast to take the Seattle Seahawks job. His teams in Seattle were the epitome of mediocrity, accumulating a 31-33 record over four seasons. He returned to college in 1999, finishing the rebuild Mike Riley began at Oregon State. The Beavers made three bowl appearances in his four seasons and reached unprecedented heights in 2000. The NFL came calling again and Erickson joined the 49ers in 2003. He only lasted two years this time, as his 2004 team bottomed out at 2-14. He returned to college in 2006, coaching Idaho for one season before leaving for Arizona State. His first Arizona State team won ten game and was briefly ranked in the top-ten. However, his final four teams in the desert went just 21-28.

Jimbo Fisher – Jimbo won the final championship of the BCS era at Florida State in 2013, but after a 6-6 season in 2017, he abruptly left the program for Texas A&M. He produced a 9-4 record in his first season at the helm.

Danny Ford – Revered at Clemson for winning the 1981 national title, scandal and differences with the administration caused Ford to resign after the 1989 season. He returned to coaching at Arkansas in 1993 and while he guided the Hogs to their first SEC West title in 1995, he had three losing seasons in five years after having none in more than a decade at Clemson.

Lou Holtz – Holtz guided the Fighting Irish to the national title in 1988 and stuck around for eight more seasons in South Bend. After a brief two-year retirement, he took the South Carolina job and had the Gamecocks bowling in his second season. However, despite winning seventeen games between 2000 and 2001, his teams could not maintain that level of play, going just 16-19 over his final three seasons.

Urban Meyer – Meyer won the 2006 and 2008 national titles at Florida. He resigned amid health concerns after the 2010 season, but returned to coaching at Ohio State in 2012. He won the first title of the playoff era with the Buckeyes in 2014 and continued to have great success on the field. However, off the field problems likely contributed to his retirement following the 2018 season.

John Robinson – Robinson led Southern Cal to a share of the 1978 national title and although he left the school following the 1982 season, he stayed in Los Angeles to coach the Rams. He guided the Rams for nine full seasons, making the playoffs in six of his first seven years. The Rams made the NFC Championship Game twice, but were soundly defeated both times (by the Bears and 49ers). He returned to Southern Cal in 1993 and while the sequel was not quite as good as the original, the Trojans did win the Rose Bowl in January of 1996 and never posted a losing season under his watch. After a one-season retirement, he took on the biggest challenge of his career when he became the head coach of UNLV. The Rebels did qualify for and win a bowl game in his second season, but that was as good as it got in the desert. His last four teams went 17-29.

Bobby Ross – Ross shared a national title at Georgia Tech in 1990. After one more season in the shadow of another Bobby, he left to coach the San Diego Chargers. Ross guided the Chargers to a Super Bowl berth following the 1994 season and actually made the playoffs three times in five years before leaving to coach the Detroit Lions. Ross coached the Lions for three and a half seasons and while they qualified for the playoffs twice, they failed to win a playoff game. With his military background as an alumnus of VMI and having started his head coaching career at the Citadel, Ross seemed like a fine choice to lead the Army program back to prosperity. Alas, the Black Knights won just nine games in his three seasons at the helm.

Nick Saban – Saban shared the 2003 national title at LSU and left one year later to coach the Miami Dolphins. Not sure what happened to him after that.

Howard Schnellenberger – Schnellenberger guided the Miami Hurricanes to their first national title in 1983, but left after the season to coach The Spirit of Miami of the USFL. His timing was not great as the team decided against relocating to Miami and did not retain Schnellenberger as head coach. He took the Louisville job in 1985 and while his overall record was south of .500, he did lead the Cardinals to a pair of top-25 finishes. He coached Oklahoma for one season in 1995 and later laid the blue print for Larry Coker’s second act as he helped start the Florida Atlantic football program.

Steve Spurrier – Spurrier won the 1996 national title at Florida, but five seasons later he left the Gators to coach the Washington Redskins. After two seasons and twelve wins, Spurrier resigned and in 2005, followed Lou Holtz at South Carolina. After a mediocre start, his 2010-2013 teams won 42 games, the first division title in school history, and finished in the top-ten three times.

Overall, with the exception of Ross, who was an abject failure at Army, those title winning coaches all experienced at least a modicum of success at their new spots. But that begs the question, what is a modicum of success at Kansas? The Jayhawks have not won more than three games since 2009, so I guess that is where we should set the bar. If Miles can get the Jayhawks to one bowl game in his tenure, he will have done better than the three coaches who have preceded him.