In the Wins Produced Primer we spell out how much value we give each box score stat. But the basic breakdown is – a point is worth, well a point. A possession is worth a point. All stats can either be placed a possession or a point. Dave has broken basketball stats down into a currency and explained how to trade these in for wins. The secret? Get points or get possessions and do this better than your opponent.
Now in this currency, a shot is worth a possession AND a turnover is worth a possession and this rubs some the wrong way. After all, how can I say when Russell Westbrook fires up a shot it’s the same as when he just throws the ball to the opposing team. The answer is, it’s not! Let’s explain.
Gotta Spend Money to Make Money
Basketball is entirely the process of turning possessions into points. When I shoot the ball, I spend a possession. When I turn the ball over, I spend a possession.
Now, here’s the difference. When I shoot the ball, some percentage of the time my shot goes in. If this percentage is good, then I’ve made a good use of the possession. When I shoot the ball, some percentage of the time I miss. In these events two things happen. Either my team, via an offensive rebound, gets the ball back, which is good. Or the opposing team, via a rebound or the ball going out of bound, gets the ball which is bad.
When I turn the ball over, I get no such luxury, the opponent gets the ball. Unless the opponent does something incredibly silly like shoot into your hoop, you have no chance of getting any points.
In this regard we see the difference. The COST of these two actions is the same but my potential outcome is much different.
But What About Rebounds?
In recent talks a theory that a shot is really only worth 0.7 possessions because about 30% of the time the shot misses, it’s rebounded by the shooting team. Alright, first, anyone with access to basketball-reference, shouldn’t make such claims. Let’s examine two stats. The first is Offensive Rebound %, which tells us which percent of misses the team gets back via rebounding. The next is Effective Field Goal %, which tells us how well teams shoot (adjusting for the three point line)
The complicated way to explain it is the thin blue line explains the range of values, and the big blue box explains the range where one standard deviation falls. Another way to say this is all 30 teams each year fall on the skinny blue line and 18 teams fall within the big blue box. Stats lesson over!
First off, this 0.7 number is a bit interesting. Most teams don’t hit that mark yearly. What’s more we’ll notice that the range for all teams in offensive rebounds is between 9% and 12% and even in the “clustered” range ignoring outliers, this is still a difference of 5% – 6%.
And when we line this up with shooting data, we notice that shooting percentages among teams is in fact more clustered than rebounding!
The answer to this riddle actually lies in Stumbling on Wins. Rebounding is a skill! It’s not a given that when a team misses that their team will get it back. Here’s actually a breakdown of how consistent the basic box score stats are.
A good rebounding player is valuable. Not just because they get the ball but because they have one of the most consistent stats across sports! In fact, it turns out that players rebounding ability is much more consistent than players shooting skills — moreso than shooting in fact. Yes, passing is also very consistent, but as we’ve noted, teams seem to realize the value in passing.
Tangent / Side Note
And I want to add one other complication! Alex Konkel at Sports Skeptic discussed offensive rebounds a while back. Basically, it’s possible to have lower offensive rebounds by doing things like shooting better, or turning the ball over more. Now the good news for all the scenarios is that we can give value to these actions.
The end story to all of this is as follows. Basketball is trying to turn possessions into points. And a way to help this is to get the ball back. However, in digging into the stats, it’s not as simple as just assuming 30% of misses go back to the shooting team, nor assuming that offensive rebounds tell the whole story. The way we handle this is to record the actions of players and give them their corresponding value in possessions or points. We can then get a good estimate for a player’s value. So is a shot more valuable than a turnover? Yes! But they both have the same upfront cost. Which players will bring the actual value to the team will vary based on what actually happens and the end story is that it’s more than a bit sketch to give much, if any credit to the shooter.