Sunday, June 27, 2010

Coaching Myths and Truths

If you ask most college basketball fans they would probably tell you that it is much harder to be a college basketball coach today than it was in the past. The days of Gene Keady and Tom Davis coaching forever seem like a distant memory. Today’s coaches either win now or find another line of work.

Three statements certainly seem true:
1) Coaching turnover is worse than it has ever been.
2) BCS coaches face much worse pressure than non-BCS coaches.
3) You need to win to make the NCAA tournament.

But do the historical data really support those statements? To answer these questions, I built a database of coaches in the 64+ team NCAA tournament era, 1984-2010. Coaching data is readily available from a number of sources including ESPN’s College Basketball Encyclopedia,,, and Wikipedia. Sadly all are incomplete in some way, and there are various inconsistencies between the various databases. If you want more details on the data, see a very boring section at the end of this post.

Let’s start by looking at whether turnover is worse than it was in the past. Table 1 graphs the percentage of coaching turnover each year. Table 1 also breaks turnover into its two components, BCS jobs and non-BCS jobs.* Of the 347 schools that participated in D1 last season, 51, or 14.7% will have new coaches this fall.**

Table 1

The first trend in Table 1 is that coaching turnover has been roughly stable over time. Perhaps due to the recession it tapered off substantially last summer, but the last five years are very consistent with the historical levels of turnover.

Table 1 also breaks the turnover into the BCS and non-BCS component. Obviously more non-BCS coaches lose their job every year because there are more non-BCS jobs. But do non-BCS coaches have a higher probability of being fired? Table 2 looks to answer this question. Table 2 separately shows the amount of BCS turnover as a percentage of BCS jobs and the amount of non-BCS turnover as a percentage of non-BCS jobs.

Table 2
A few trends are apparent in Table 2.
-2009 had the lowest non-BCS turnover rate on record. Thanks to the recession, barely 10% of non-BCS coaches were replaced last summer.
-In the late eighties and early nineties, the amount of BCS turnover briefly reached even lower levels.
-Turnover is very similar between BCS and non-BCS positions. If anything, BCS turnover rates are minimally lower.

Sadly I don’t have documentation on all the coaching changes back to 1985. Thus I don’t know how many of these changes were voluntary and how many coaches were forced out for poor performance. But other details in the data can tell us something about why the job change happened.

Of the 1272 coaches to change jobs,
-946 did not start another head coaching job immediately
-326 jumped to a new Division 1 head coaching position

Of those without another job, some like Tom Brennan of Vermont voluntarily retired. But I think we can assume that the vast majority of these 946 changes are situations where the coach was forced to leave.***

Of the 326 coaches to move to a new job
-37 moved from a BCS job to another BCS job

-148 moved from a non-BCS job to another non-BCS job
Besides the BCS conferences, the most common conferences to hire D1 coaches away from other conferences include the MWC, A10, MVC, CUSA, and the CAA. Other elite leagues like the WAC, Horizon, and WCC have not traditionally hired many coaches away from other schools.

-115 moved from a non-BCS job to a BCS job
These are the yearly success stories.

-26 moved from a BCS job to a non-BCS job
These are another type of failure. Certainly some non-BCS jobs are very prestigious today. Think Memphis, Gonzaga, ect. But I’ve looked over the 26 moves from BCS to non-BCS jobs and none of them appear to be a move to a more prestigious position. Some like Greg McDermott moving to Creighton from Iowa St. might have been a mutual agreement, but I would classify all 26 of these moves as a situation where the coach was forced to leave.

Tenure at the Time of Termination

I’d next like to document some facts about the coaches that were forced out. (These are the 26 coaches who moved down to a non-BCS conference and the 946 coaches who lost their job and did not have another D1 coaching job immediately.)

First, I’m going to calculate the tenure of each coach. Sadly, I do not have clean data prior to 1984-85, so I do not know the tenure of all coaches in the early years of my data. But starting in 1994, I can make at least a 10-year tenure calculation.

Of the 650 coaches forced out starting in 1994,
55 had 1 year tenure
45 had 2 years tenure
64 had 3 years tenure
85 had 4 years tenure
81 had 5 years tenure
72 had 6 years tenure
46 had 7 years tenure
33 had 8 years tenure
35 had 9 years tenure
134 had 10 or more years of tenure

A few notes:
-Year 1 is somewhat high because of the large number of interim head coaches in my sample.
-The data also includes many assistants who take over when a long-term coach leaves the school. It appears that these assistants make up a disproportionate fraction of the coaches released in Year 2. Perhaps these coaches get a shorter leash because they were expected to recruit while they were an assistant.
-The peak in terminations appears to be year 4, 5, and 6. But this is a bit misleading. The denominator, the number of jobs that still exist in year 7 and 8 is much smaller. Even though the number of terminations is smaller, that’s also because the number of coaches who make it this far is smaller. As an alternative, let’s estimate the probability a coach will survive to various points in time.

I re-define my exercise as follows: Assume you were hired after 1984-85 to a D1 team****. What is the probability you would survive to various points in the tenure clock?

One advantage of asking the question this way is that I do not have to wait to start using data until 1994. Table 3 shows a step-function with the probability you would survive to various points in the tenure clock.

Table 3

Obviously the odds you survive 10 years are fairly low. But what is the probability you get fired at any single point in time? Table 4 graphs the probability you are fired as a smooth curve.

Table 4

Table 4 estimates a smooth curve and that curve suggests that job separations do not begin to decline until after year 7.

Now clearly at this point, a number of factors can influence whether a coach keeps his or her job. Some of them are easily quantifiable, such as wins. Others are harder to quantify, such as expectations. (Expectations may depend on school prestige, recruiting success, and the interest level of the Athletic Director, which is certainly hard to measure.)

But for today I’m going to stick to a simple question. Does making the NCAA tournament improve a coach’s chances of keeping his or her job?

I estimate a model based on the percentage of seasons the coach has made the NCAA tournament at any point in time. I could obviously graph a number of lines, but to keep the graphs easy to read, I’m just going to show the estimates for two situations. In blue, I show the probability the coach is fired if he never makes the NCAA tournament. In red, I show the probability the coach is fired if he makes the NCAA tournament 50% of the time.

I’m going to look at non-BCS coaches and BCS coaches separately. Tables 5 and 6 show the non-BCS coaches. Tables 7 and 8 show the BCS coaches.
Table 5 and 6: Non-BCS Coaches

Table 5 and 6 shows that making the NCAA tournament half the time in a non-BCS job ensures almost perfect job security. The red line is close to zero in Table 6.

Table 7 and 8: BCS Coaches

Table 7 shows that if you never make the NCAA tournament in a BCS job, the probability you survive a decade is extremely low. (There are exceptions. See Bill Carmody.)

Table 8 also shows that simply making the NCAA tournament half the time does not provide as much job security in a BCS job. It certainly is a lot better to make the tournament half the time instead of never at all, but many coaches make the tournament half the time and still get fired.

Notice that the odds of getting fired from a BCS job always appear higher. But this is conditional on how often you make the NCAA tournament. And because BCS coaches make the tournament much more frequently, BCS coaches are much closer to their red line and non-BCS coaches are much closer to their blue line. As we saw in Table 2, the overall level of turnover is roughly equivalent.

You may wonder whether survival probabilities would reach 100% if the coach made the tournament every season. The answer is clearly no. Just ask recent Arizona head coaches Kevin O’Neil and Russ Pennel. Or ask Jim Harrick who made the tournament every year at UCLA but was fired for lying to NCAA investigators. Making the tournament matters, but it is not the only thing you have to do to keep your job.

Nonetheless, Tables 5-8 show that the data support the popular perception. Making the NCAA tournament is very important for keeping your job. And this is particularly true if you coach in a BCS conference.

A Few Extremely Boring Notes

* The BCS didn’t exist in 1984-1985, but I count ACC, Big 10, Big East, SEC, Big 12, and Pac-10 members as BCS schools all the way back to 1984-85. Some schools joined a BCS conference in the middle of the database. See Miami FL, Penn St., Rutgers, Virginia Tech, ect. I list these schools in a BCS conference only after they joined a BCS conference. For the Big 12, I had a choice to make about whether to count the Big 8 or SWC as a historical BCS conference. Because many of the SWC schools are not in a BCS conference today, I count only Big 8 schools as being in a BCS school historically and not SWC schools.

**In my merged database, if an interim coach served most of the season, I count two coaching changes. For example, when Minnesota coach Dan Monson was replaced by Jim Molanari early in the 2006-2007 season, I credit that season to Molinari and count a second coaching change when Tubby Smith took over in 2007-08. I do this because when trying to understand coaching styles (tempo, bench utilization, ect.), I want each season to count towards the coach who was actually managing the team.

Besides the obvious impact on volume, this can also alter the timing of some job changes. For example, Jerome Allen actually took over at Pennsylvania as an interim head coach during the 2009-2010 season. (And I use the 2009-2010 data to try to understand his style-of-play.) But most sources do not credit him as officially becoming head coach until this summer. Since he took over early last year, I refer to 51 coaching changes this summer instead of the 52 reported elsewhere.

***You might think I could identify the Tom Brennan voluntary seperations by finding coaches with winning records or some other metric. But that is not as clean as it first appears. Due to NCAA violation issues, graduation issues, and unrealistic expectations, even winning coaches can get forced out.

****I also include coaches whose teams moved to D1, but only after they move to D1. For example, Doug Knoll took over at IPFW during the 1999-2000 season and the school began playing D1 hoops in 2001-2002. Thus I do not include him in the database until 2001-2002, but I do count him as a third year head coach in 2001-2002.

"Mutter Ineffectually" is Usually my Policy

(That's a nod to John Gasaway for those of you who don't recognize it.) I maintain this blog in my spare time and I would like to stay a happily married man so I’m trying to avoid joining Twitter and I rarely respond to email promptly.

But one thing that makes me sad is that I do not have enough time to link to other great statistical articles on college basketball. Linking is very important and I sincerely thank everyone who has linked to me over the years. (A single Luke Winn Top 25 can keep me motivated for weeks.) Thus if you have something statistical that you think would interest me, please send me an email to DLHANNER at GMAIL dot COM.

Recently Villanova By the Numbers managed to contact me and remind me that I’m not the only one writing about the coach hiring process. Do you know how often coaching changes happen after the spring signing period? Did you know that sometimes D1 coaches take assistant coaching positions voluntarily? Here’s Part 1, Part 2, and Part 3 of the VTBN series.

Sunday, June 20, 2010

The State of Coaching (Summer 2010)

Four years ago, Indiana had a chance to hire former Hoosier player Steve Alford as head coach. They passed on the opportunity and Alford eventually moved on to New Mexico. At New Mexico, Alford has won a ton of games and posted fantastic efficiency numbers along the way. Meanwhile, Kelvin Sampson imploded the Indiana program. In retrospect, it seems like Indiana made the wrong decision.

But I still think Indiana made the right decision. As evidence, I point to NC State which hired former player Sidney Lowe in that same summer four years ago. Lowe was supposed to restore the tradition of NC State’s 1983 national title. But while he’s brought back some nice memories for NC State fans, he certainly hasn’t added any new ones.

NC State has played some exciting games against North Carolina, but otherwise remained irrelevant in the ACC race. The team has never finished better than 6-12 in four seasons. And of the remaining four-year coaches, Sidney Lowe has the worst efficiency margin in the ACC.

And yet, NC State is not quite ready to fire their inconsistent coach. Some claim a new recruiting class will save the day, but I doubt it. Lowe has seemed over his head from day one. But I suspect that because Lowe is a former star player, the university boosters are hesitant to make a move. It’s easy to get rid of an inconsistent outsider with failed promises. It is hard to get rid of a respected part of the university tradition.

This year Iowa St. heads down the same road, hiring former star shooter Fred Hoiberg. Hoiberg played in the NBA for 10 seasons, and no doubt knows a lot about basketball. But does he know enough to win in a BCS league right away? My extended thoughts about the new hires are listed below.

Inside The Numbers

Before I make comments on various coaches, I’m going to update the efficiency stats. The following tables list the number of years at the current school (YRS), the average adjusted offensive efficiency (AAOE), the average adjusted defensive efficiency (AADE), and the national rank of the efficiency margin (RK). I list the numbers with the “current school” and the average over “all schools” in the last seven years. Averages are calculated using data available on

Keep in mind context when interpreting any number. Steve Donahue’s numbers are not great nationally, but they were great for the Ivy League. The best remaining Ivy League coach is Tommy Amaker whose efficiency margin at Harvard is -5.9 points per possession. On the flip side, Sidney Lowe’s numbers seem pretty good. But in the ACC, they’ve been near the bottom every season.

Also keep in mind that this is a “what have you done for me lately” world. So the simple average is not always the best way to evaluate which coaches are hot and which coaches are in trouble. This past year several coaches were hired to new jobs after making positive changes with their teams. Jeff Bzdelik (Colorado to Wake Forest), Tad Boyle (Northern Colorado to Colorado), and Bob Marlin (Sam Houston St. to Louisiana-Lafayette) were all hired to new positions based on dramatic offensive improvements. Similarly, Tony Barbee (UTEP to Auburn), Ed Conroy (The Citadel to Tulane), and Milan Brown (Mt. St. Mary’s to Holy Cross) were all hired to new jobs based on tremendous defensive improvements. And after Steve Donahue turned Cornell around on offense and defense, he became the hot hire for Boston College.

Because the recent trend matters, I highlight in green and red coaches that were 5 points better or worse than their historical average last season. The point here is to illustrate that not all coaches low on a conference list are necessarily the worst. Scott Drew had a negative efficiency margin in his first two seasons at Baylor. But Drew brought his team to the Elite Eight last year, and the recent numbers have been much better.

Hot Seat

We’ll revisit this topic in-season, but the hot seat candidates are pretty obvious. Sidney Lowe probably won’t get to see his current recruiting class graduate. Ed DeChellis bought some time with his NIT run in 2009, but he and Bill Carmody have done little to warrant serious job security. Doc Sadler might not survive to see Nebraska head to the Big Ten. And while Pat Knight may coach a few more games than Sean Sutton did after replacing his father, it isn’t clear that he’ll coach a lot more games. Mick Cronin is in serious trouble after Lance Stephenson didn’t deliver an NCAA bid. And luckily everyone looks safe in the Pac-10, but that’s only because they are all so new. (It is hard to believe that Herb Sendek is already the third most tenured coach in the Pac-10.) Finally, Andy Kennedy’s off-court activities haven’t helped his job security in the SEC.

New Hires

Now that Tom Izzo is not heading to the NBA, I think we can safely say this was the most boring off-season for coaching changes in some time. I can’t even generate a decent coaching chain.

"Oliver Purnell – Brad Brownell – Billy Donlon" just doesn’t look that exciting on paper.

Sure, there were 51 jobs that changed hands (and one at Chicago St. that remains open). But with Butler’s Brad Stevens staying put, and with none of the elite programs making a change, I’ll do my best to make these comments interesting.

ACC New Hires

-I’m not showing pace in any of these tables today, but I think it is worth emphasizing that the ACC may be getting a lot slower. Last year the ACC added slow-paced Tony Bennett, and this year the league adds three coaches with a reputation for working the shot clock. Jeff Bzdelik and Brad Brownell’s squads have ranked over 300th in tempo on numerous occasions, and Steve Donahue’s Cornell team ranked a not-so-speedy 245th in the nation last year in tempo.

This is probably smart strategy. Most teams won’t consistently perform at Duke and North Carolina’s level. And if you want to upset Duke and Carolina, you want to minimize the possessions in the game. But for fans used to bragging about the entertainment value of track meet ACC games, this change may not be for the better.

-Steve Donahue’s negative efficiency margin doesn’t look great, but remember that he’s been coaching in the Ivy League. Donahue clearly accomplished something special in building Cornell into a Sweet Sixteen team. I don’t expect Donahue to have Boston College competing for the ACC title next year, but he deserves a chance to build a program. A good projection for Donahue might be Fran Dunphy. Dunphy jumped from the Ivy League to Temple four years ago and after an inconsistent first year, Temple’s efficiency margin has improved three years in a row.

-Jeff Bzdelik’s Colorado team was rated near 300th in the country in 2 Pt FG% defense the last two seasons. So my initial reaction to the Wake Forest hire was that the Demon Deacons were making a mistake. If a team doesn’t play interior defense in the ACC, they are going to get crushed. But when I look at Bzdelik’s overall numbers, and remember what he did at Air Force, the move starts to make more sense. Bzdelik probably has the best efficiency margin of anyone Wake Forest could have hired at this point. And assuming the problem with the interior defense was a lack of quality post players at Colorado, it may be an aberration. In the end, Wake Forest gets a star offensive coach, and Bzdelik resets his tenure clock with a more prestigious BCS team.

-Did Oliver Purnell bolt Clemson because of a lack of commitment from the university? Or does he just like a challenge? There’s no question Clemson experienced serious heart-break under Purnell, but there’s also no question he built the program to levels not experienced since Rick Barnes left. Under Purnell, Clemson was no longer the team projected for the cellar or near the cellar every year. And with Purnell gone, the team adds Wright St.’s Brad Brownell. Brownell isn’t the sexy pick based on big NCAA tournament upsets, star recruits, or even a recent league title. But all he has done is win. In 8 seasons in the mid-major CAA and Horizon league, his teams have finished lower than 3rd only once. The numbers suggest he’ll have a tough-minded defensive team. And if you don’t have the most talent in the league, defense is the great equalizer.

Big Ten New Hires

-Fran McCaffery pulled the ultimate turnaround at Siena. He took over a team that was 273rd in the Pomeroy rankings and raised them to a team that was 59th nationally. His team went from last in the MAAC in 2005, to three straight MAAC titles. Sure, his average efficiency numbers aren’t phenomenal. But that’s only because of where his teams started. They’ve gotten steadily better since he took over. Iowa’s basketball team is a huge rebuilding project, and that is why McCaffery is the perfect fit. He’s rebuilt a disaster before, now he just has to do it again.

Big 12 New Hires

-Besides my concerns about how quickly Iowa St. can fire Fred Hoiberg if he fails, I’m also really concerned about his lack of coaching experience. Can he really motivate and communicate with players in the heat of a dreadful season? While Fran McCaffery can look at his Iowa team and say, “I’ve been here before guys. I know it is bad now, but it will get better.” I’m not sure Hoiberg can say the same thing. But my pessimism is lessened somewhat by Hoiberg’s decision to hire former Charlotte coach Bobby Lutz as an assistant. I never felt like Lutz got a fair shake. He built UNC-Charlotte into a consistent NCAA contender in CUSA, and then when the Big East raided CUSA, he lost all his rivalries with elite teams. His team was put in the A10 and was never able to reform an identity. A lot of people talk about how Memphis was left behind when CUSA reorganized, but Charlotte really got left behind. Hopefully this is the opportunity for Lutz to show what he can do once again, even if he isn’t officially the head coach.

-For this year at least, Tad Boyle is a Big 12 coach. And Tad Boyle should know how to recruit in Colorado, after coaching at Northern Colorado. And his team did win 25 games last year. But the positive vibes end there. Boyle’s teams have never been particularly dominant, and even last year’s 25 win club wasn’t great. They finished second in the Big Sky, and only picked up 25 wins because of an incredibly weak non-conference schedule. I’ll be the first to admit that the stats are not everything. But there isn’t anything in Boyle’s resume that makes me think Colorado won’t be hiring again in a few years.

Big East New Hires

-Sadly Steve Lavin’s last year at UCLA came before we started tracking tempo-free stats, but based on my memory, I have questions about what he can accomplish. He had a much better winning environment at UCLA and his teams often underachieved. But I have effusive praise for this move because of the timing. Lavin inherits a team of seniors with a chance to win right now. And, he has a full year to bring in what should be a critical recruiting class to St. John’s. We should know by next spring if Lavin can recruit St. John’s back to glory.

-Oliver Purnell is the dream hire for DePaul. He brings an excitement to the program thanks to his full-court pressure and attacking system. And this isn’t just Keno Davis seven-seconds-or-less revolving door at the basket. Purnell’s average defensive efficiency in the last seven years is 90.8. That’s equal to the average defensive efficiency of Jay Wright, and nearly as good at the 90.4 posted by Jamie Dixon. This is a team that will play defense, and be fun to watch. But Purnell brings more than just occassional full-court pressure, Purnell brings instant credibility. Instead of a mid-major coach trying to convince recruits that DePaul may win someday, Purnell knows he has a system that can win immediately. Just like Purnell did when he took over at Clemson, and just like Tubby Smith did when he took over a down-trodden Minnesota program, I expect DePaul will be better immediately. The ultimate question for Purnell is not whether he will win, but whether he can eventually compete for a Big East title. I’m skeptical he can get DePaul to that level. But after the last couple of years, even a .500 team sounds fantastic to DePaul fans.

-Does the “almost” upset count for anything? Robert Morris “almost” upset Villanova in the NCAA tournament this year. Which means by now most of you forgot about it. Just like you may have forgotten that Fred Hill stepped down at Rutgers and that Mike Rice Jr. was hired from Robert Morris. Hey, maybe you never even read that headline. But that’s the problem. Rutgers isn’t hiring a well-known coach. Rutgers isn’t hiring a proven winner, a proven offensive or defensive genius from the mid-major ranks. Rutgers is rolling the dice outside the normal hiring pattern. Robert Morris is hiring the coach of the three-time NEC champ, a league that as recently as two years ago was the third weakest conference in the country. And that’s fine. I’m actually a pretty big advocate for trying out winners from different levels. But if you are going to take a risk on an unproven commodity, couldn’t he at least have a signature victory? Is he supposed to sell Rutgers recruits on the fact that he once “almost” beat Villanova?

-Kevin Willard did little at Iona that would merit getting hired as a coach in the Big East. His team finished 7th, 7th, and 3rd in the MAAC. But Kevin is the son of a former coach who has lots of connections in the business. And while it is easy to criticize this type of hire for the lack of supporting “data”, the truth is that college basketball coaches need connections to succeed. Fair or not, Kevin Willard has access to a much bigger network of recruits than Mike Rice Jr. He’s also spent years on the sidelines next to incredibly smart basketball minds. And even though I can’t point to a single piece of data to tell you why Kevin Willard will succeed at Seton Hall, he has something a lot of mid-major coaches would dream of… an opportunity.

Pac-10 New Hires

-Certain coaches are off limits in the mid-major ranks. Mark Few is not leaving Gonzaga anytime soon. Jim Larranaga is happy at George Mason. And for a long time, Dana Altman has been in that category. But somehow, Oregon made the hire. But for some reason, people want to find negative things to say about this hire. Why didn’t he go to Arkansas a few years ago? Why hasn’t he won in the NCAA tournament? Why has his team struggled lately? Why couldn’t Oregon find someone better after such an extensive search? I find all of these comments ludicrous. No BCS team (other than DePaul with Oliver Purnell) got a coach with a better efficiency margin than Dana Altman. He knows how to coach offense and defense, and recruit to a league that gets multiple NCAA tournament bids. On paper, he is the perfect hire.

SEC New Hires

-The post John Calipari CUSA was supposed to give multiple teams a chance to be good. But probably no one seized the opportunity more than Tony Barbee’s UTEP squad which made the NCAA tournament as an at-large selection. But while part of me wanted to see a little bit more out of UTEP before I was willing to anoint Barbee a top mid-major coach, the truth is in the four year numbers. UTEP’s efficiency margin has improved three years in a row. Barbee is ready for the challenge of a BCS league.

Who's Next?

I suppose the natural question is what coach will be next to jump to a BCS gig. But the MWC and A10 are often as much a destination as a way to get a better job. Steve Alford isn’t getting paid peanuts at New Mexico. Fran Dunphy will gladly tell you he can win a ton of games in the A10 at Temple. But many of the coaches with the best numbers, from Lon Kruger to Brian Gregory, will go elsewhere if the price is right.

-Chris Lowery was once the hot coach with the best numbers, but he’s been trending in the wrong direction. Still, if any of those top MVC coaches has a good year, they could easily be on the move.

-Other people have said this, but it is worth pointing out again. CUSA is definitely a league that likes to recycle former BCS coaches:

Mike Davis – Indiana
Larry Eustachy – Iowa St.
Matt Doherty – North Carolina
Ben Braun – California
Jeff Lebo – Auburn
James Dickey – Texas Tech
Tim Floyd – USC

-I’m expecting Tad Doyle will make me eat my words now, but if Colorado has to make an odd hire, why not someone like Stew Morrill? I know he isn’t young. And maybe he cannot be lured away from a safe, low stress environment. But I find it sad that only people living in the mountains ever get to see how good Stew Morrill is at coaching offensive basketball.

-I’m also surprised Old Dominion’s Blaine Taylor didn’t get a little more press for some of the openings. Defensive coaching will translate to any league.

-Notice that the mid-major coaches with the best numbers tend to be relatively new. They inherited great programs (think Chris Mack and Josh Pastner), and don’t have enough track record to get a BCS job yet.

This last group includes the rest of the coaches with positive efficiency margins. I also list the coach with the worst efficiency margin who has kept his job for 7 years and the coach with the worst efficiency margin overall.