Spike Eskin: Larry Brown Or How I Stopped Worrying And Learned To Love Basketball Analytics
By Spike Eskin
PHILADELPHIA (CBS) – Former Sixers coach Larry Brown joined Anthony Gargano and Glen Macnow on 94WIP yesterday afternoon, and my head almost exploded.
My head almost exploded not from excitement, though I was looking forward to hearing the great coach speak, but from listening to Brown talk about the new Sixers general manager Sam Hinkie, basketball anayltics, and their relation to the success of a basketball team.
Though it shouldn't be surprising that a 72 year-old basketball coach is resistant to a new way of thinking about the sports that he's coached so successfully for so many years, it was how he resisted.
VIEW: Larry Brown Doesn't Believe Analytics Work In Basketball
Brown did what many do when they suggest that analytics aren't helpful, he used a straw man argument. In its simplest terms, a straw man argument is when you misrepresent what the other side of a debate says in order to make your point. You may know it from just about every political debate or discussion about Ryan Howard you've ever taken part in.
So as a once and for all (at least until next time), here is a collection of what Brown and others say about basketball analytics and those who use them, and why they are not true.
Numbers Can't Replace What You See On The Court!
No one who is successful using basketball analytics would ever suggest that you don't pay attention to what's happening on the court. In fact, basketball analytics, even more so than the baseball variety, use film and what is actually happening on the court to gain a better understanding of what makes a team successful.
Analytics don't replace game tape, they use game tape. What basketball analytics hope to do is replace other statistics that are less helpful. Points per possession, or offensive and defensive efficiency is a much better way to measure a team's offense and defense than points per game. Rebound rate is a better way of finding out how well a player rebounds than total rebounds.
These totals that people use, like points per game or rebounds per game are "counting statistics," and not always a very good indication of what's really happening. Think of it this way: if I tell you, "City X had 15 murders in 2012," there's no real way to know what that number means unless you know how many people are in the city. Is it a small town with 200 people? Is it a major city with three million people? The rate is what you want to know, not that total.
GM's Who Use Analytics Don't Watch Basketball, And Aren't Basketball Guys
So, let me get this straight. A man who is choosing to spend 19, 20 hours per day working for a basketball team, analyzing basketball players and basketball teams, isn't a basketball guy and doesn't watch basketball? Doesn't this seem sort of unlikely to you?
Sam Hinkie, as a specific example, is a brilliant guy and is very good with numbers. If he didn't like and know basketball, he could find many other ways to earn money, that would probably make him a lot more of it. Hinkie is known as a guy who believes in traditional scouting and watching players, as well as using analytics.
Here's the thing, guys who use basketball analytics generally have a better understanding of how basketball works than most anyone you know. They watch more basketball than anybody you know. Numbers guys know basketball fall better than non-numbers guys know numbers.
The key is using them both, and you can't use them both unless you know both.
The Oakland A's Did Not Win A World Series Therefore "Moneyball" Is Not A Success
I don't even know when to start here. First of all, there are plenty of baseball traditionalists who have not won World Series' either.
The story of Moneyball, and the amazing thing about what the A's did, was to find a way to compete while being at a huge disadvantage as far as payroll goes. They found a market inefficiency in order to compete.
Theo Epstein did win a World Series while being an advocate of advanced statistics. So, there's that.
Basketball Isn't Baseball, Basketball Statistics Don't Work The Same Way, So They Don't Work In Basketball
This is true, basketball is not baseball. It's also true that basketball statistics are not like baseball statistics.
That's why the advanced statistics for basketball are different than they are for baseball. Basketball analytics are far more concerned with video, and what is happening on the court, while baseball advanced statistics tend to be more results related (what happened rather than why it happened and how it happened).
Numbers Lie, Therefore You Cannot Trust Them
Absolutely, numbers lie. What advanced statistics do is try to sort through the lies of traditional statistics.
"Player A can score 20 points per game, therefore he is a good player." That is a lie. What analytics do is find out how he's scoring those points, how efficiently he's doing it, and if he's as good as we think he is.
Stats Can't Measure Heart
You got me there, statistics cannot measure heart, guts, or want to. They can't measure "will to win" or intestinal fortitude.
Then again, you can't really measure it by looking at the guy either, so it looks like we might have to guess on this.
You Win "If You Play Defense, Rebound, And Share The Ball" - Larry Brown
This was really an amazing straw man moment from Larry Brown. I'm searching and searching but I cannot find an analytics guy who suggests that this isn't true.
In fact, analytics give us a better sense of who is playing good defense and who is rebounding.
Prior to analytics, we always believed blocked shots and steals were great measures of how a player is defensively. Now we know, that sometimes it's the centers who do not get a lot of blocked shots who are the best defenders. Players who are so good at positioning themselves and intimidating that players don't even come into the paint. Players that are good defensively, but are resistant to leaving their man and leaving him open, the way many blocked shots happen.
"The rocket science is this: acquire draft choices, get great players, have good contracts." - Larry Brown
This is another straw man gem from Brown. Again, who is suggesting this isn't true?
What these analytics do is help figure out how to best spend these draft choices, and what is a good contract.
"I think Spencer Hawes is a really underrated center." - Larry Brown
I'm not sure that this is at all related, but Brown really said that. Take it for what it's worth.
Some Reading For You
Now that you know what analytics aren't, it's a good time to find out what they are. Here's a good start:
Read "Lights, Camera, Revolution" by Zach Lowe.
Read "David Lee's Interior Defense" by Kirk Goldsberry
Look, most of the people who are telling you that analytics are no good and not a help are either resistant because they don't feel like learning about them, or afraid that it might threaten the knowledge they thought they had.
Before you put up a wall, and say you don't believe they're valuable, find out what it's all about first. A few years ago, I wasn't a believer. Then I took the time to find out what I was talking about.
When the Sixers introduce Hinkie is "the guy," this afternoon, I won't be excited because I know he will be successful, I will be excited because I know he will have a plan.
for more features.