The Dean
“Great minds think alike, small minds rarely differ” -Unknown
I’m sure at some point of your life you and a friend stumbled into the same idea, and you referenced the first part of the idiom I stated above. “Great minds think alike!” My experience is quite the contrary. This may be true in well-studied and understood areas of reasoning, but it seems to me that the greatest minds tend to call the general consensus into question. In return, we gain a greater understanding of the world around us.
It seems to be that only those who ask, “Why do we do it that way?” ever come to a real understanding of what the “rule” is, when it applies, and when it can/should/will be broken. Bill James (one of the founding fathers of the Moneyball movement in baseball) did this when he began the process that brought sports analytics to the forefront of the modern sports conversation.
For me, my learning about analytics in basketball lead me to bump heads with the Dean. Dean Oliver that is.
Paper Basketballs
Dean Oliver is a former assistant coach for the Washington Wizards of the NBA and also the author of “Basketball On Paper: Rules and Tools for Performance Analysis”. He has been one of the influential figures that has pushed forward the use of analytics to evaluate basketball teams and make decisions. As I have worked to understand the game of basketball better, I came across Dean’s book and it arrived on my doorstep Sunday, April 7th.
In the beginning of Dean’s book he makes a case for the best historical offenses and defenses in the NBA (Using data from 1974 - 2004 when the book was published). These teams are generally classified by their Offensive Rating (ORtg) and their Defensive Rating (DRtg). A team’s Offensive Rating for a season is an estimate of the number of points that team scores when given 100 offensive possessions1. The defensive rating of the team is an estimate of the number of points that team will ALLOW to be scored when given 100 defensive possessions2.
Dean’s Tall Take
After looking over the top offenses and defenses Dean decided to look at the average heights of these teams compared to the rest of the NBA. After comparing the roster height of top offenses to the rest of the league, Dean notes:
“…So height should help an offense.
In the numbers, the story isn’t perfectly clear but supports this assertion, as shown in table 3.6.
The greatest offensive teams were generally taller than average…
The bottom line is confirmation of traditional lore: height is good, but there are other factors.”
Table 3.6 as referenced can be seen here:
The only problem with these statements is that he is….sort of right. Sure more of the top teams were a little taller than the league average that year, but not by much. I went ahead and tried to correlate Average Height above league average to their offensive rating and this was the result:
The NBA is just not the place to be looking at this kind of data. By the time you get to the NBA, 99.9% of us 5’ 10” guys are on our couches eating potato chips dreaming of what could have been if coach would have just seen our potential. We could have been a 5’ 10” MJ! Just give us a chance coach! (The 1982 Denver Team we see on the far left of this graphic may very well be a team worth looking out for short people advice on basketball.)
Giving Dean a Second Chance
So Dean certainly didn’t give us much to work with there on offense. Maybe his defensive results gave us some more solid evidence. Does height affect a team’s defensive ability? Dean says:
“As with offense, yes it does….Interestingly, though, the best defensive teams are on average a little shorter than the best offensive teams (see table 3.13).
Table 3.13 can be seen here:
Once again, Dean is only kind of right. Team height and defensive rating looks like this:
Wah, wah, waaaaah. Keep in mind, on this graphic, a lower number on the Y-axis is better. It means your team allowed MUCH fewer power points per possession than the league average. However, both sides of the height spectrum have teams above and below our line.
The 1975 Washington Bullets are the team here that interests me. They were nearly 1” shorter than the rest of the league but were 6.4 points better on defense per 100 possessions.
Fail Forward
There were some interesting “Nuggets” in Dean’s analysis (pun intended), and I will have to chase a stray “Bullet” or two (again, pun intended) but I think what Dean has shown me is that the NBA is going to be a difficult place to find short teams to learn from.
My NCAA data that I showed you this past week hasn’t seen much cleaning. I tried to work with the data some today, but Microsoft BI seems to move pretty slow. Hopefully I can have some more analysis of that data in the near future. For now, enjoy this GIF of a chihuahua dunking a basketball:
Training Update
Training these days is load management and working on my lower body strength. My shoulders continue to hurt and it doesn’t feel like there was much improvement. It is hard for me to figure out a balance between activity and rest. I have always been a very active person, and never experienced significant injuries while I was growing up. I also have never been disciplined with my sleep. I don’t know that there has ever been a time in my adult life where I have averaged 8 hours of sleep per night.
The downside of not having dealt with injuries is that I never learned how to manage recovery. I am used to going as hard as I can at all times rather than patiently working into higher levels of activity. I believe I was cautious with my shins and shoulders this week but only additional time and experimentation will tell me whether that is true.
The focus of my workouts this week were as follows:
Monday
Clean Pull
Barbell Deep Squats
Posterior chain superset
Hip Thrusts
Calf Raises
Tuesday
General Strength Day
Wednesday
Dumbbell Squat Jumps
Barbell Back Squat
Posterior Chain Superset
Nordic Curls
Calf Raises
Thursday
General Strength Day
Friday
Snatch Pull (I substituted some wide-grip clean pulls here to try and protect my shoulders)
Dumbbell Walking Lunges
Plank Walkouts (I did barbell roll outs with my arms close to my body on this one because of my shoulders as well)
Saturday
Jump Day
I don’t know that I was jumping higher, but I was jumping! I will take that! Happy jumping!
Are these "average roster heights" weighted by minutes played?
Does keeping a 7'0" or a 5'10" guy on the bench all season long affect the average roster height? What if a team trades a 5'10" guy for a 7'0" guy midway through the season?
I wonder if a better comparison might be a comparison of 5 man lineups DRTG and ORTG compared to that lineup's height.
The "small ball" warriors had Javale McGee, but he wasn't part of the death lineups.