WHEN GOOD COACHES GO BAD (The Howland/Lavin Conundrum)I doubt many people would disagree with these two statements (I wouldn't):
Steve Lavin was a lousy coach at UCLA with an even lousier work ethic.
Ben Howland is a very good coach who works hard to succeed at UCLA.
Yet aside from three outstanding years, comprising a smaller and smaller fraction of his UCLA career, Howland's teams have under performed Lavin's. In those three good years he went 97-17 (85.1%), 13-3 in the NCAA tournament. But in the other five and a half years he's 105-76 (58.0%) with just two tournament wins. By comparison, Lavin had a better record for his worst five years 97-61 (61.4%) and won six tournament games. Want to toss out his first year using "Lavin's players"? (You wouldn't want to if he had a good year with them, would you?) Then he's an identical 94-59 (61.4%).
How is this possible?
Is Howland not as good as people think?
Was Lavin better (horrors!) than he gets credit for?
Or perhaps it's a matter of semantics.
Or something else.
There is more to coaching than the traditional aspects like teaching fundamentals and game preparation and strategy. Those who are strong in these areas are considered to be the "good" coaches. Howland is in this category and rightly so. But there's more to winning than that. You have to be a GM, cheerleader and PR guy, too, not Howland’s strengths. If you can't recruit the right players or motivate them or hire a good staff, then all the good practice and strategy won't win games. You have to build a team with the right combination of experience and talent that fits your system. If you can't create a buzz around the program, people won't show up to the games. If a coach does nothing but recruit good players, roll a ball out for practice and hires great assistants to run it and wins, then he's doing a good job, even though he might not be considered a good coach. The problem is that once a coach gets a reputation as good or bad, that perception stays with him no matter how good or poorly a job he does in the future. Fans pay a lot of money to buy tickets and make donations and want to see the team win games and be entertaining to watch, not “fundamentally sound” and have a mediocre record.
Let’s look at some of the factors people believe make a good or bad coach and how they pertain to the perception of Lavin and Howland:
TRACK RECORD – Of course Lavin had no track record as a head coach before he fell into a big-time job, which, in hindsight, was a legitimate concern. But neither did some of the most successful coaches around like Jim Boeheim, Steve Fisher and Roy Williams, so that didn’t automatically mean he couldn’t succeed. In fact Boeheim had almost the identical amount of experience, was the same age and was the default choice when his predecessor unexpectedly left. I'm sure a lot of people weren't happy with his hiring. Clearly, Howland’s record of turning around programs qualifies him as a “good” coach. However, sometimes that perception is magnified if a coach follows someone who was doing a poor job. If you succeeded a highly successful popular coach, no matter how well you do might not be good enough in the eyes of the fans (e.g. Bartow). On the other hand, if you follow an unpopular coach, fans seem to have a greater tolerance for mediocrity. That, for example, is why Nikki Caldwell was overrated to the point where some fans though UCLA should have paid nearly $1 million to keep her here. CNC was a good coach who won a lot of games with a semi-soft schedule, but she was only 2-15 against ranked teams, overemphasized defense (where have we heard that before?), couldn't keep the top local high school talent from leaving the area and didn’t have much success in the NCAA tournament (two wins in three years). However, she followed IMO, one of the worst coaches ever to work in Westwood in any sport and whose teams horribly underachieved for many years, considering the great success of almost all of the other UCLA women's sports. So not only does Howland look pretty good following Lav at UCLA, he had the good fortune to follow the inept Harold Merritt at NAU, who won only 32% of his games in four years there after he was somehow hired after winning 38% of his games at Morris Brown. But at Pitt Howland followed Ralph Willard, who only had one winning season in five years there, but won big in jobs before and after, so CBH looks pretty good in that job.
STYLE OF PLAY - Lavin's teams were often criticized for playing "streetball", code for letting his players freelance and not running a structured offense. However, the flip side of this is enforcing too much control over talented players and not letting them use their natural talents. Calipari wouldn't be doing a very good job if he forced his lottery picks to slow the game down and run a patterned offense every time. But the coach who makes his team run a fundamentally sound offense will always get credit for being the "good" coach even though it might not fit his personnel. But certainly it's better to err on the side of discipline. An undisciplined offense will tend to take bad shots and turn the ball over more often. Offensive efficiency, the number of points per possession, is one way to measure the effectiveness of an offense. (Note: Pomeroy's adjusted efficiency only goes back to 2003, so we used raw efficiency which yields similar numbers.) So whose offense was more efficient? Lavin's streetball or Howland's fundamentally sound system?
(Note: There are approximately 330 DI teams, so a ranking of about 165 is average.)
Average national rank OE (offensive efficiency):
Howland - 93.9 (220-131-74-36-11-6-182-131-53)
Lavin - 84.9 ( 36-39-122-74-83-66-174)
Throw out Howland's first year and the average is 78. Not much difference.
One way to look at shot selection is the effective field goal percentage (fg% weighted for 3-point attempts). If a coach is running a patterned offense, the team should get open shots. If the team is running more, they should get easy shots that way. If a team is taking a lot of bad shots, the percentage should drop.
Average national rank eFG%:
Howland - 69.2 (once in top-40, three times over 100)
Lavin - 60.1 (four times in top-40, twice over 100)
So Lavin's teams managed to get as many or more higher percentage shots with his offense than Howland's (or he just recruited better shooters). But the negative of a wide open, undisciplined style is turnovers. We can adjust for the discrepancy in tempo between the two styles by using turnover percentage (turnovers/possessions).
Average national rank TO% (higher number is better):
Howland - 180.1 (four times 270-312, five times under 120)
Lavin - 149.6 (five times 141-191, twice under 120)
So Howland's teams were either excellent at taking care of the ball or below average. Lavin's teams were usually average. Ironically, this year's team is Howland's best at taking care of the ball.
Since Howland's offense seems to be no better statistically than Lavin's, the difference must be on the defensive end:
Average national rank opponents OE (higher number is better):
Howland - 200.2 (64-171-313-323-335-87-251-101-157)
Lavin - 144.0 (153-54-266-107-199-88-51)
The offensive and defensive numbers suggest that Howland needs his teams to play outstanding defense in order to be successful. Even the relatively high offensive efficiency in 2009 and 2012 didn't get him very far.
There is, of course a large downside when you look at the big picture. Many recruits will not come if they have the impression they won't be able to play a running style. Many fans will not come to the games if the team is not playing an entertaining brand of ball, especially when the team isn't winning. That leads to an atmosphere at the games which doesn't help the team and isn't appealing to recruits. It's a vicious cycle.
RECRUITING - When Lavin first started pulling in highly rated recruiting classes, the standard line was UCLA recruits itself . . . anyone can do that here. Well, Howland has proven over the last few years that isn't true. His recruiting difficulties have been well documented: head cases, busts, bad fits, inability to get a top high school point guard, etc. Then, later, Lavin was knocked for poor scouting, picking players from the published rankings without regard to position or specific skills or abilities. The beginning of the end came when he started recruiting highly rated but flawed players like T.J. Cummings, Andre Patterson and Cedric Bozeman. But Howland is having the same problems with an unbalanced roster. The quick players he recruited must be so fast, no one can see them and his three-point shooters (258th nationally in 3's made) must have so much range, they aren't even in the gym.
HIRING STAFF - Some have the opinion that one of Lavin's biggest mistakes was to hire his high school cronies as assistants instead of older experienced coaches. The comparison:
Of the five who served as Lavin's assistants, three had little or no D1 coaching experience (Holton, Spencer, Madkins). The other two had at least seven years at major schools. Of Howland's seven assistants, two had over 10 years D1 assistant coaching experience and two had been D1 head coaches (Daniels, Matthews). The other three had one to four five years D1 experience.
RESPECT & DISCIPLINE - Lavin's teams had a reputation of lacking discipline, on and off the court, and there was a strong impression that his players didn't respect him. Baron Davis' offhanded comment about being the first team to make the tournament without a coach is often cited. However, lack of respect for Howland exists, but is much more subtle. The massive amount of players who have left the program is telling. While each individual case can be attributed to something like lack of playing time or wanting to play in the NBA, Howland wanted almost all of them to stay. He made his case, but they all left anyway, a sign of disrespect. His discipline of players hasn't always been strong. The wishy-washy treatment of Reeves Nelson, letting him stay after repeated problems and allowing an out-of-shape Josh Smith play to the detriment of the team. Off the court Howland's players have had their share of legal problems, too, although none quite matches Rico Hines' attempted murder of Matt Barnes (I was there).
CONSISTENCY - A big knock on Lavin was that his teams were wildly inconsistent. They played great one game and terrible the next. One can't measure that by wins and losses because of the varying strength of opponents from game to game. However, if we use a rough estimate of the expected performance, the point spread, we can get a measure of volatility, the percent of games which had the opposite result of the previous one, i.e. beating the point spread (playing better than expected) or not beating the point spread (playing worse than expected). So if a team alternated between the two every game of the season, their volatility would be 100%.
Looking at this number for all of Lavin's years compared to Howland's years, there is quite a difference: Howland's team's, especially in recent years, were much more inconsistent than Lavin's. For Lavin the numbers are 25-45-32-45-47-48-45. Howland's are 39-48-41-56-54-40-72-62-52. So not only were Lavin's teams more consistent within seasons, they were more consistent from season to season. And what is more inconsistent than making the Final Four one year and having a losing season two years later?
Now how is this possible, given that many people have a strong impression of Lavin's teams inconsistency? Sometimes one or two incidents of extreme swings in important games leave such a strong impression on people, they generalize it to an entire year or career. For those of you technically oriented, imagine the team's game to game performance as a sine wave. Lavin's has the greater amplitude, while Howland's the greater frequency.
IMPROVEMENT OVER THE SEASON - Howland has a reputation for his teams improving over the course of the season. But how do we measure this? The variation of schedule strength from the non-conference to conference seasons makes this difficult. One way is just to look at the first half of conference play vs the second half. You play the same nine teams in both.
Howland - first time around 50-22 (69.4%)
second time 47-25 (65.3%)
Lavin - first time around 38-25 (60.3%)
second time 42-21 (66.7%)
Another way is to look at the whole season against the point spread:
Howland - first half 55.7%
second half 54.5%
Lavin - first half 42.9%
second half 51.0%
Better to bet on Howland, but Lavin's teams improved more.
Those are just rough ways to look at it. A more precise way would be to look at offensive and defensive efficiency for each half, but you run into strength of schedule problems. The methods above adjust for that.
PLAYERS INTO THE NBA - This subject has been discussed quite a bit, but it's one of the more overrated aspects of college coaching. The majority of players who make the NBA would have gotten there playing under most coaches. The league is full of players who had little or no college experience and/or played for mediocre coaches. They learn more about playing in the NBA from summer pickup games against Russell Westbrook than college practices scrimmaging against Tyler Trapani. There are a few exceptions. Probably Luc owes it to Howland since his whole game is based on defense, but there are no other Howland players who one could give him full credit. Hollins? Maybe. Afflalo? He came to UCLA as a prolific scorer with a reputation for toughness and hard work. Westbrook? Howland didn't teach him that crazy athleticism. Collison? If you want to give him credit for any four year player who wasn't an NBA prospect coming in, then that's a slippery slope since you also have to give Calipari credit for Josh Harrellson and Lavin for Earl Watson. Certainly not Ariza, Holiday and Love. And you can't count Lee and Honeycutt since Howland himself said they weren't ready. Now that's not to say he didn't make them into better NBA players. He certainly did that because he's a good teacher and coach. But he didn't make them into NBA players. That's a product mostly of natural talent and hard work.
But there's a big downside, too. If a coach gets lots of players into the NBA, fan expectation is higher. Why couldn't Howland win a national championship with four future NBA starters, two of them all-stars? Plus, the exponential improvement of Love, Westbrook and Holiday once they left UCLA has some wondering if their talents were held back here. And does he also take the blame that Josh Smith's NBA stock has probably dropped since he arrived on campus?
While it's nice to have a lot of NBA players from your school, it can also be a negative. Aside from early exits destroying the continuity of the program, a reputation for producing a lot of pros might attract the type of recruits who are more interested in showing off their skills to NBA scouts rather than playing winning team ball. Mike Kryzewski once had a reputation for not turning out many successful NBA players at a time when his teams were dominating college basketball.
20 WIN SEASONS - This is the most overrated statistic in college basketball. In Wooden's day, teams were only allowed to play a maximum of 26 regular season and four playoff games. That's when it meant something. Now teams routinely play 35 or more a season. Last season 116 Division I schools won 20 or more games and a couple dozen more were close. The year before it was 118. But even just the win total is misleading. Howland has been notorious for padding his non-conference schedule with weak teams. At Pitt, his strength of schedule was never higher than 65th and as high as 97th, despite playing in the Big East. At UCLA, it has only been in the top 20 once and under 50th just three times. On the other hand, Lavin never won more than 24 games, but his SOS was in the top-20 four times and worse than 32 only once. The conference was also stronger during the Lavin years, other Pac-10 teams winning 46 NCAA tournament games in his seven years compared to 31 in eight Howland years.
VS USC - An extremely important stat for Bruin fans, Lavin was 10-4 against the cross-town rivals while Howland is a rather pedestrian 11-8 (soon to be 12-8 with a gimmie win over a pathetic Trojan team).
NCAA TOURNAMENT RESULTS One of the big knocks on Lavin was that he "only" made Sweet 16's. Some derided him as "Stevie 16". Well Howland is finding out it's not that easy. He's won exactly two tournament games the last three years (Lavin never won less than four in any three year span), by a total of 3 points, both against double digit seeds and almost blew a 23 point lead in one of them. Odds are he won't win one this year. How long can one rest on three Final Fours? The only other coach to make three consecutive Final Fours without winning one was Guy Lewis and Lewis once went 14 years between Final Four appearances. (Would you want to wait that long?) But three Final Fours takes some luck, too. Actually, Lavin would have made one Final Four if it wasn't for a cheating Minnesota team and Howland might have made only one if it weren't for a choking (and crying) Gonzaga team and an official's non-call against Texas A&M (Shipp: "They didn't call it, so I'm not thinking about it," he said of the Texas A&M non-call. "It's over.").
Steve Lavin reached at least the Sweet Sixteen as many times in the six years between 1997-2002 as Ben Howland has in his entire 18 year coaching career.
Some fans consider Jim Harrick a great coach, simply because he won a championship here. But were it not for Tyus Edney's miracle play, his entire legacy at UCLA would be five embarrassing NCAA tournament losses in eight years. In fact, one could make a case that he's one of the worst NCAA tournament coaches ever. Even before and after UCLA his teams folded time after time in March Madness games:
1982 - lost to Oregon St by 19
1983 - blew a six point lead with 24 seconds left against NC St in a game which could have changed the course of basketball history
1985 - lost to Duke by 13
1991 - lost to #12 Penn State in first round
1992 - lost to Indiana by 27 in the regional final with arguably his most talented UCLA team (considering future NBA players)
1993 - blew 19 point lead and lost to Michigan
1994 - lost to #12 seed Tulsa in first round giving up 112 points
1996 - lost to #13 Princeton in first round
1998 - blew a 6 point lead with less than a minute left against Stanford
1999 - outscored 16-5 in overtime period by Charlotte
2001 - started game with Missouri 0-15
2002 - win over Murray St vacated
HALFTIME ADJUSTMENTS - A good coach will make the proper adjustments at halftime if they are behind and come back to win. Poor coaches will be outcoached and blow halftime leads.
Howland - 37 wins after trailing at halftime
21 losses after leading at halftime
Lavin - 37 wins after trailing at halftime
23 losses after leading at halftime
BLOWOUT LOSSES - Oner of the biggest knocks on Lavin: all of those blowout losses. Lavin had 14 losses of 20 points or more, Howland only seven. Some fans call them embarrassing losses. But an embarrassing loss is one that should embarrass the players and the coach . . . not the fans. Except that the players seem to get over them the next day, but some fans haven't gotten over some of those games in years.
Many of Lavin's blowouts are explainable if not excusable. The three worst were his first three:
It was the Maples Massacre very early in his career that first set many Bruin fans against Lavin. It turned the questioners into doubters and the doubters into haters. But if the team used that as motivation for beating Stanford by 19 in the rematch and winning the Pac-10 championship, then it's a positive. The next was a 41 point beatdown by UNC. But that was a UNC team that made the Final Four and featured three future 1st round draft choices including that year's college POY, Antawn Jamison and the next year's NBA Rookie of the Year, Vince Carter, while UCLA was starting two freshmen guards in their first ever college game and without two key players who were suspended. The same year they lost to Duke by 36. That was on a trip to the east coast late in the conference season (preceding an important trip to the Washington schools) to a Duke team with five future top-20 draft choices who won fifteen games that year by 30 or more points including several against ACC opponents.
While all but three of Lavin's blowout losses came against teams that went deep into the NCAA tournament (the others were against above average Pac-10 teams) and only one at home, two of Howlands were against mid-majors and three were at home. But the worst was the mother of all embarrassing losses, a home loss to a mediocre USC team which makes the Maples Massacre look like a Maples Slap on the Wrist by comparison. It was the largest losing margin in the series since 1945. ("I'm embarrassed for our team," UCLA Coach Ben Howland said. "I feel embarrassed for the program, for the former players and coaches. That's all you can say, it was embarrassing.").
UPSET LOSSES - So we're glad to have Howland. We don't have to suffer through so many of those games we have no business losing. But are you sure about that?
Games lost when UCLA was a 5 point or greater favorite - 18 in seven years.
Games lost when UCLA was a 8 point or greater favorite - 10 in seven years.
Games lost when UCLA was a 10 point or greater favorite - 7 in seven years.
Games lost when UCLA was a 5 point or greater favorite - 21 in eight and a half years.
Games lost when UCLA was a 8 point or greater favorite - 11 in eight and a half years.
Games lost when UCLA was a 10 point or greater favorite - 9 in eight and a half years.
Not much difference, is there? Yet this is something that is always associated with Lavin, not Howland.
If you want to make the argument that Howland's teams may have been a double-digit favorite more often, then you are admitting he inflated his record with his penchant for playing easy non-conference games. Even if they were big favorites more often because he had better teams, that is also irrelevant. These are games you shouldn't lose if you play one or one hundred.
Now look at the flip side of the inconsistency.
Lavin's record in games they were 8 or more point underdogs: 8-17 (32%).
Howland's record in games they were 8 or more point underdogs: 3-16 (16%).
So under Howland, anytime UCLA is a big underdog, don't count on an upset. With Lavin, there was a reasonable chance.
OVERTIME GAMES Winning them is a product of strategy, conditioning and motivation. Surely a good coach should win the majority of those games. The records:
Howland: 18-17 (career).
Lavin: 12-4 including a streak of eight in a row, just three short of the NCAA record.
TOURNAMENT LOSSES - Last March, UCLA lost a Pac-10 tournament game to an Oregon team which had lost their last four regular season games by 17 points. A well known local columnist for a major media outlet called it "a Steve Lavin sort of loss." Actually, it had Howland written all over it. Until UCLA lost it's NCAA tournament game last season by eight points, all six of Howland's NCAA eliminations were by double digits as well as his last three conference tournament losses.
THE EYE TEST - Many will say that what doesn't show up in the statistics is clear when you watch the teams of each coach play. Sorry, but basketball is not a subjective sport. There are no judges to decide who wins after the game is over. The team with the most points wins. Period.
Now if you've gotten to this point and are thinking, This post is garbage. I don't care what the numbers say, Lavin sucks and Howland is a great coach, then you haven't been paying attention. Go back and read the first few sentences. This is not a defense of Lavin, nor an indictment of Howland. This is all about the fans' perception. The perception of what characteristics make up a "good" coach and how some of them are actually negatives in the long run. It is how a coach who is acknowledged as "good" can be no better or even worse in some important categories than a "bad" coach. If I had cited the stats blindly as for Coach A and Coach B, most would be hard pressed to identify which belonged to Lavin and which belonged to Howland for most except defensive efficiency in the three Final Four years. But people tend to disregard the negatives of the coaches they like and magnify them for the coaches they dislike. If something goes wrong, they blame the "bad" coach in one case, but blame the "good" coaches' players in the same circumstances. Although that may be true in some cases, the simple answer is that a coach may not be at fault for everything that happens under his watch, he is certainly responsible for it. People are reluctant to criticize someone who replaces an unpopular coach since that would make the "bad" guy they just trashed look better by comparison.
Unfortunately, most fans choose to ignore the elephant in the room. Although Howland is clearly the better coach, except for the three "Golden" years, he isn't doing a better job than Lavin, the benchmark of a bad coach, did by almost any measure. If you say the numbers don't matter, that Howland has been doing a good job the last few years, then you are proving my point. You need to produce some tangible statistical evidence rather than just the perception that since he's a good coach he must be doing a good job.