Basketball Analytics - 5 Things I Didn't Realise
I will start out by saying I am definitely not an analytics expert, or even novice by any stretch. I am a complete beginner. So please forgive me if I make any silly statements. But this book is definitely worth a look, as well as Basketball on Paper by Dean Oliver as two great starting points for a greater understanding on the numbers.
Here are my 7 things I didn't realise from Basketball Analytics by Stephen Shea. Some I had a bit of a gut feel or speculations about, but nothing I knew for certain.
1. We should strive for discomfort from a deep understanding of the complexity of the analytics.
I always had this bias that, when people talked about analytics it was the answer.
So I wrote a lot of things off.
Analytics should help test and challenge your biases, opinions and create better questions.
What it doesn't do is tell you the direct answer to a problem.
"We don't want analytics; we want analytical people." - RC Buford
2. Plus Minus is Fairly Garbage
Consider who you are playing, if you are on the end of a big win your plus minus is looking pretty good. If you cop a big loss it's looking the other way.
Quality of opponent determines production.
Consider who you play with, if you playing with the entire bench against their starters you are in trouble.
Chalmers/Battier in 2013 if they played with Lebron their +/- was through the roof, if not it was way down.
Consider times of the game, in junk time does that really add much value to plus minus?
It over rates role players on good teams.
Here are Danny Ainge's thoughts on plus minus.
3. The Box Score Stinks
"Someone created the box score, and he should be shot." - Darryl Morey
Four Factors: eFG%, TOV%, RB%, FTR from Dean Oliver's book will give you a greater understanding of your box score.
The amount of rebounds, shots, free throws etc changes due to the context of the game so we shouldn't look at the numbers in a vacuum. It's all relative.
Assists per minute = not just a reflection of how he can pass but how he's being used in the offence.
Example: usage of Goran Dragic playing with or without Eric Bledsoe in 2013 was very different.
4. Things Change in the Playoffs
Pace slows down 3-4 possessions per game.
Attempted pull up jump shots go up, catch and shoots go down.
Assuming higher attention to detail defensively, as every possession matters a little bit more.
Corner 3's become more important.
When n=1, the last play of the game. It doesn't matter where you take the shot from if you can make it.
5. Player Evaluation
Our eyes overvalue quantity and scoring.
Analytics don't check the eye test. That's why they're worthwhile.
Check efficiency versus quantity, helps to question how they go in the next level.
What's the appropriate analytic for that type of player?
Eg. individual playmaking efficiency for point guards
Recognise the context - who are they playing with and who are they playing against?
If you have any further thoughts on analytics that can help my understanding, please message me on Twitter, comment or shoot me an e-mail at email@example.com.