Switch 5 (Part 3): Does This Stuff Matter in the Regular Season?
Part 3 of a 4 part series. Read Part 1 here and Part 2 here.
Takeaways so far:
We’ve established that switching has become significantly more popular, it happens much more often in the playoffs, teams likely started switching more to mitigate the advantages that ball screens created, and while switching may result in giving up better shots, teams are having a hard time playing any defense successfully in the NBA.
And:
Switch 5 lineups are: Maybe better at guarding mid-range and above-the-break 3’s (but shooting percentages can be so fluky its hard to draw conclusions), not as good at rebounding (but the difference on the defensive glass is pretty small), and better at forcing turnovers.
NBA teams have concluded there are more efficient play types than the post-up, and while small lineups are not as good at guarding the post-up as ones with traditional bigs, the advantage gained isn’t enough to “expose” these types of lineups.
Why Don’t Teams Switch More in the Regular Season?
In Part 1, we saw that the gap between switching frequency in the regular season and playoffs was massive:
If teams have concluded that switching is the better defensive strategy, why do teams not switch nearly as much in the regular season? Let’s examine a couple of potential reasons:
Switching is Hard
Second Spectrum provides a handful of fitness stats, including speed and load for each player, along with distance traveled during games. Per the Second Spectrum Definitions Handbook:
Speed: Magnitude of the two dimensional velocity of the player in feet per second
Total Load: The total acceleration in feet/s² integrated over the chance incorporating changes in speed and changes in direction
Speed Zones: Fast = greater than 14 ft/s, Medium = between 8 ft/s and 14 ft/s, Slow = less than 8 ft/s
I don’t know anything about biomechanics, but wanted to visualize some of this data in case anything popped out. First I gathered the data of all the centers who played in the playoffs since the beginning of the tracking era. Next, I divided them into two groups based on switch frequency. Here’s how their fitness stats compared:
Second Spectrum fitness stats show a slight increase in distance traveled per 36 minutes for players with high switch rates. Additionally, players with low switch rates spend much more time in the “Slow Load” category as opposed to the medium or fast ones.
Not a massive difference here, but keep in mind that the players in the “High Switch Frequency” category still only switch *at most* 50% of the time.
This is not surprising — especially in the era of load management. Coaches don’t want their bigs flying around the court every possession and putting more stress on their bodies if not all possessions are created equal.
Law of Averages
Statistically, we can conclude that switching in the regular season does not lead to a better defensive rating:
Not much of a correlation here. In the regular season, great defenses are not necessarily built on switching — they can be — but they don’t need to be.
Let’s take a look at the top 10 defenses by defensive rating since 2017:
A lot of teams here with traditional centers and a conservative, rim-protecting scheme. The 2020–21 Milwaukee Bucks, who went on to win the title, are further down on this list because they “experimented” with different defensive styles during the regular season (shoutout Eric Nehm), rather than sticking to their drop-back-and-shut-off-the-rim style that was so successful in regular seasons of yore.
Why does the traditional rim protection/drop/less aggressive strategy work in the regular season? One potential answer is the competition. The best shotmakers in the NBA make the playoffs (typically) and actually play in those games (typically). In the playoffs, you won’t find load management or a game against a team that is tanking.
Another reason is sample size. If you give up shots that are typically harder to make, it has a greater chance of evening out over 82 games than over a 7 game or less playoff series.
Some great analytical work has been done to increase our knowledge in the”Quality of Shot” space— check out Cleaning the Glass’s Location eFG%, Second Spectrum’s Shot Quality metric, or ShotQuality.com.
The correlation between defensive rating and Second Spectrum’s Shot Quality surrendered in the regular season is just over 60%.
82 games is a lot. This isn’t March Madness. In the playoffs, not only are you going up against the better shot makers, but there’s also less of a sample for it to even out. There is more of a reason for coaches to feel uneasy about giving up a pull-up jumper to a really good player even if it’s statistically a tough shot to make.
Where we go from here
The difference between the regular season and playoffs is as stark as it has ever been. We see teams build rosters specifically for playoff basketball, while teams that put up great numbers in the regular season may fizzle out in the postseason. A portion of that difference is defensive strategy.
We established two potential reasons that switching occurs significantly more in the playoffs: (1) Switching is physically more taxing and (2) Switching doesn’t necessarily give you leg up in the regular season because a larger sample typically means shotmaking regresses to the mean.
Next up, let’s see how the switching trend has impacted roster building — stay tuned for Part 4.