The Myth Of The Stretch 5
Real Eyes, Realize, Stretch Lies
The Center who can both anchor a defense and stroke the three has become a hallmark of the modern NBA. Brook Lopez, Myles Turner, Serge Ibaka and countless other Bigs have been dubbed “Stretch 5s” for their perceived ability to bend a defense via the threat of their three point shot. In theory, these players “stretch” the floor by dragging their defender (often a rim protector) out of the paint and toward the perimeter to create more space and efficient opportunities for teammates to score around the basket.
But in reality, defenses are happy to leave Stretch 5s open and dare them to shoot.
If you browse through the three-point shot logs for just about any Stretch 5 you’re more likely to land on a clip that shows them open than one that shows them guarded. That’s because Centers — even the sweetest-shooting ones — are left wide open on an inordinate number of their three point attempts.
The table below summarizes the shooting volume and efficiency of every player who has shot at least 75 threes while also playing at least a third of their minutes at the Center position this season. In addition to showing each player’s overall three point shooting numbers, I’ve included columns for each player’s volume and efficiency on uncontested and contested threes.1
The column to pay attention to is the colored one on the right side of the table, which shows the frequency with which a player takes contested threes. This value is a reasonable proxy for estimating how much defenses respect and react to the threat of a player’s outside shooting ability. In this column, higher percentages indicate that defenses are reticent to leave the player open, while lower percentages indicate a lack of concern when the shooter is unguarded. It follows then that the higher the percentage, the more a player stretches the defense and visa versa. For context, Kevin Durant leads the league in the share of threes taken that are contested at 91 percent. Meanwhile, Marc Gasol is last in the league at 8 percent.2
The top of the table features Danilo Gallinari and Carmelo Anthony. Both guys take a hefty share of contested threes but only play the 5 in spot minutes. Although they aren’t traditional Centers, I think it’s useful to include them here to show the disparity between real three point threats and mild ones.
The only players on this table that I might consider Stretch 5s are Nikola Jokic, Kristaps Porzingis, Bobby Portis, DeMarcus Cousins, Chris Boucher, and Karl-Anthony Towns. All six players take at least 40 percent (my own subjective Mendoza Line for determining whether a player is a floor spacer or not) of their threes when the defense is within six feet and each one spends the plurality of their minutes at the Center position. But the difference in how frequently these guys take contested threes compared to a small ball 5 like Gallinari is massive.
Every other player on this table not already mentioned usually only takes threes when no one is around them. Which is to say they’re hardly stretching the defense, if at all.
It’s not enough to just look at a player’s overall volume and efficiency on threes to determine whether they can stretch the floor. For example, even though Joel Embiid is having a career year from three, defenses rarely guard him beyond the arc. Fewer than 20 percent of Embiid’s threes are contested. That’s not an accident. It’s by design. A defense would rather have Embiid take an open three because it means he isn’t wrecking shop in the paint.
In other words, most Centers take wide open threes because defenses want them to take wide open threes. Or as we say in Soviet Russia, “In modern NBA, defense stretches you!”
To see the lack of spacing most 5s provide, it helps to show the frequency with which they take contested threes relative to other positions. That’s what I’ve depicted in the chart below. Note here I’ve assigned players a position based on where they spend the plurality of their minutes so as not to double count anyone (i.e., Danilo Gallinari is shown as Power Forward (PF) here as opposed to a Center (C) because he plays 53 percent of his minutes at the 4).
Compared to other positions, hardly any Centers take a meaningful share of contested threes. Even Jokic — arguably the stretchiest of the Stretch 5s — is guarded on threes far less frequently than other players who have taken at least 75 attempts. For reference, Jokic ranks 80th in terms of how frequently his threes are contested while a small ball 5 like Gallinari ranks 21st.
Once you get past the first tier of Stretch 5s ending with Boucher and Towns, the stretchiness of the remaining Centers on the chart fades quickly.
Veteran Bigs like Nikola Vucevic, Al Horford, and Serge Ibaka may camp out on the perimeter, but defenses contest their outside shots less than 40 percent of the time. As a comparison, historically shaky three pointers like Russell Westbrook and Ricky Rubio are guarded at around the same rate. Meanwhile, protomodern Bigs like P.J. Washington, Lauri Markkanen, and Myles Turner are contested on their threes at even lower rates.
This isn’t to say that there’s no value in having Centers position themselves behind the three point line. Even if a defender isn’t closely guarding a Stretch 5, they’re probably standing further away from the action than they would be if the Stretch 5 was battling for post position. Additionally, even a mediocre Stretch 5 can act as a safety release for a driver and turn an empty possession into an average one.
But these things are all relative. As we’ve seen again and again in the playoffs, teams will downsize their lineups in order to put more shooters on the court that defenses have to respect. In a seven-game series, when opposing teams have time to game plan, pseudo-Stretch 5s like Brook Lopez and Myles Turner will be goaded into shooting even more often than they are now. Unless they’re willing to fire away, it will be difficult to justify having them on the floor.
As an example, come crunch time in the playoffs, I think we’re more likely to see the Clippers play Marcus Morris at the 5 even though Ibaka’s main selling point in free agency was his floor spacing ability. At the end of the day, defenses respect Morris’ shot a lot more than Ibaka’s.
I’m not suggesting teams should stop playing their Stretch 5s entirely. But I think it’s worth asking if they could dial it back. After all, how valuable is it really to have a Center permanently stationed out on the perimeter taking wide open threes if that’s what the defense wants them to do anyway.
The New Orleans Pelicans have found success on offense in spite of the lack of spacing between Zion Williamson and Steven Adams. Neither player settles for open threes and yet they have one of the leagues most potent offenses. Meanwhile, although it’s a small sample, the Philadelphia 76ers are 5-1 since inserting the non-shooting Tony Bradley into the starting lineup in place of the injured Embiid.
I’m just saying if a Center is going to get played off the floor eventually anyway, maybe teams should adopt the philosophy of the Pelicans and Embiid-less 76ers by spending more offensive possessions doing things defenses don’t want them to do.
On stats.nba.com, shot attempts are grouped into one of four categories based how much space is between the nearest defender and the shooter when the shot is released. Those categories are:
Very tight (closest defender between 0 and 2 feet)
Tight (closest defender between 2 and 4 feet)
Open (closest defender between 4 and 6 feet)
Wide open (closest defender more than 6 feet away)
Friend of The F5, Seth Partnow, has written extensively about the problems with these naming conventions. The tl;dr of it is that it’s more accurate to describe any shot where the defender is within 6 feet as “Contested” and any other shot as “Uncontested.” In other words, “Open” shots aren’t really all that open. So when I say a shot was “uncontested” I’m only referring to shots that were labeled as “Wide open” on stats.nba.com. Todd Whitehead has a great Twitter thread showing visual examples of each type of shot category.
Here’s a table showing the top 10 and bottom 10 players in contested threes taken, irrespective of position.