How Well Does True Shooting Percentage Explain How Many Points A Team Scores?
By Jae Bradley
True shooting percentage is an adjusted form of field goal percentage that takes into account the type of shot a player attempts. For example, take two players — Player A and Player B.
Player A shoots ten layups and makes five of them — Player A’s field goal percentage is 50%.
Player B shoots ten three-pointers and makes five of them — Player B’s field goal percentage is 50%.
However, while Player A and Player B have the same field goal percentage, this doesn’t mean that the relative worth of their shooting was the same — a layup counts for two points (and is easier to make) while a three-pointer counts for three points (duh) and is harder to make than a layup (also, duh). True shooting percentage more accurately measures the shooting efficiency of a player based on the number of free throws, two-point field goals, and three-pointers that a player makes (and misses).
(For the actual true shooting percentage formula see here. Still don’t get it? See here.)
I’m going to switch gears quickly to another statistic, points per 100 possessions. This statistic is exactly what it sounds — how many points a team scores per 100 possessions. The great thing about points per 100 possessions is that it is rate adjusted, which means that the statistic accounts for how fast or how slow a team plays.
Why do I bring up points per 100 possessions?
Because you’d think that a team that scores more (i.e. has a higher points per 100 possessions) would also have a high true shooting percentage (i.e. they shoot very efficiently) — and you’d be right!
But the real question is, how much of a team’s points per 100 possessions can true shooting percentage account for? In other words, how much of a team’s scoring can be explained by it’s shooting efficiency and how much of it’s scoring can be explained by “other stuff”?
It turns out that almost 80% of a team’s points per 100 possessions can be explained by a team’s true shooting percentage!
(can’t see the picture above? click on it!)
Whoa, whoa, whoa — graphs with lines, and dots, and crap — what are you doing here?
Here’s what I did — I took points per 100 possessions and true shooting percentage data for each team from the past three NBA regular seasons (you can get it here from NBA Stats) and I correlated the two variables.
What the hell is correlation?
Simply put, when things increase or decrease together, they’re correlated. When it is rainy out, people tend to use umbrellas. When it is not rainy out, people tend to not use umbrellas. Thus, rain and people using umbrellas are correlated.
However, it is important to remember that correlation does not necessarily imply causation. Umbrella use doesn’t cause rain (though rain might cause umbrella use).
(Still don’t get it? Try this.)
Alright, alright, I get this correlation stuff — but how do you know how much a team’s points per 100 possessions is explained by true shooting percentage?
If you click on the graph, you’ll notice a thing that looks like R^2 (I’m gonna call this R-squared from now on). This R-squared measures how well true shooting percentage explains a team’s points per 100 possessions. R-squared is between 0 and 1, so an R-squared of 0.8193 means that true shooting percentage explains about 82% of a team’s points per 100 possessions.
Ok, so what does this all mean?
The first thing that it means is that how efficiently a team shoots greatly determines how many points it scores.
Now you’re all thinking, “I read through all this freakin’ $*!# to get to that idiotically obvious statement? No $*%@! , Sherlock!”
(See 0:08)
This might be your face right now — and I don’t blame you
But the other main takeaway is that you can look at this from the flipside.
Nearly 20% of a team’s points per 100 possessions can be explained by something other than how well it shoots.
Now that’s a pretty interesting thought — nearly 1/5 of a team’s offense doesn’t directly have to do with how well a team shoots the ball.
So, what are the other factors that help explain the “other 20%”?
I don’t know exactly what they are (probably offensive rebounding, the number of forced turnovers, etc), but until next time, that’s some pretty substantive food for thought.