Assessing The Celtics' Pick-and-Roll Defense, Part I

Assessing the Boston Celtics Pick and Roll Defense: Part One
By Kevin O'Connor  @KevinOConnorNBA on Feb 27 2014, 2:19p 11




How do the Boston Celtics defend the pick and roll? Here, a new way of assessing defense with statistics is introduced, and the Celtics' production defending the pick and roll in February is briefly analyzed.


Basketball is about teamwork, and no play better exemplifies that than the pick and roll. It's a classic: Larry Bird and Kevin McHale dominated with it in the 1980s, John Stockton and Karl Malone popularized it in the 1990s, and Steve Nash and Amare Stoudemire took it to a new level in the 2000s.

Defenses have longed to figure out the best way to stop it, or at least contain it; but there is no preferred way since it's more about the personnel on the team, and not the technique utilized.

That's why the Boston Celtics were such an astounding defensive team the past six seasons. Kevin Garnett is one of the greatest defenders of all-time, and was able to elevate Boston's ability to stop the ever-so-popular pick and roll.

But he's gone, and so is Doc Rivers. For years, Doc and Tom Thibodeau, placed a heavy emphasis on "hedging" the pick and roll with KG's elite defensive abilities. But now, Brad Stevens has taken a new approach to Boston, called "ice."

Throughout the season, the Celtics have had trouble defending the pick and roll. Teams have been able to choose their spots, or use favorable matchups to their advantage.

But it's nearly impossible to assign "credit" or "blame" on defense with one simple statistic. On offense, if the ball-handler makes a bad pass, it's labeled as a turnover, which clearly signifies where the blame should be placed. Or, if a shot is made, he gets credit for the points.

But on defense, even with programs like Synergy Sports, points can be pinned against a certain player defending someone, but that never accounts for the real reason a basket may have been made, which could've been the lack of team communication, or a missed rotation on the backend of the play.

Synergy does label types of plays, such as spot-up jumpers, cuts to the basket, and even pick and rolls, but it doesn't go so in-depth in that it assesses the entire play that happened. SportVU technology is fairly new and is currently changing the basketball world, but even that can't automatically read a pick and roll play just yet, or any other type of designed play for that matter.

That's why it takes actual film analysis to quantify "credit" and "blame" on a collective defensive effort. It's the old fashioned way, but a more hands on approach is worth it, since no individual player is ever solely responsible for a basket being allowed or prevented.

To find out what's working and not working within the Boston Celtics' defense, I recorded a number of different statistics for every pick and roll play run against them during their 11 games in February.

In the first part of the pick and roll series, let's take a look at some of the new statistics that will be used, and then how the Celtics fared this month using those stats:

Method

Statistics Used

PPP, or Points Per Possession: This stat is very popular today since it signifies how many points a team scores (or allows) per possession. In this case, it only includes the instances where a shot or free throw is attempted, which doesn't really answer the full question necessary to understand how well a team defends a pick and roll.

That's why P4, or "Points Per Pick and Roll Possession" is being introduced. This stat compensates for PPP's weaknesses in that includes whenever an offense resets, or has a turnover, since that should be accounted for as "credit" when a play is made, in addition to shot and free throw attempts.

Reset: One of the issues with Synergy Sports is that it doesn't give any credit to defensive players that force a "reset." If any type of play is run by an offense, but they don't get the result they intended for, they will often pass back out to the perimeter for a "reset of the offense." This is important to use because it gives credit to players who force the offense to do something else.

Another stat that will be used is called "open" or "open percentage." Again, Synergy Sports is a results-orientated system, and it doesn't account for whether or not a player was open when attempting a shot, pass, or anything else. Using this, it gives an idea of whether or not a basket was made even if it was contested. Of course, this type of defense is preferred compared to open shots at the rim. SportVU technology can already account for something like this, as it was used in various studies last year, but it is not publicly available information in this case.

Counting

As was stated above, all of these statistics were found by watching each individual play on film, and then by taking notes on pick and rolls only. Pieces of data that were recorded include:

Side or Middle: Whether or not the pick and roll took place on the side or middle.
Ball Man: The player defending the ball-handler.
Roll Man: The player defending the player who is setting the screen on the "ball man," which is often a big man.
Pick Man: Labeled by a number depending on their position on the court, 4 being a power forward, or 5 being a center. This is especially important when discussing a "switch" on the pick and roll.
Ice, Hedge, or Switch: This is the type of technique used to defend the pick and roll.
Help Defense: This is represented by a "yes" or "no," as it's important to note whether or not other defenders came over to help on a play.
Action: This is a simple description of the play that occurred.
Result: The result of the play, including make, miss, turnover, reset, and whether or not the player was open or covered during the result.
An example of what one sheet of data looks like can be seen by clicking here.

Results

A broad look at the Celtics' pick and roll

The Boston Celtics defended the pick and roll a whopping 441 times over the course of 11 games in February. Out of those 441 chances, opponents scored 353 points, for a P4 of 0.80. This means that a team scored less than one point per pick and roll against the Celtics.

Overall: Boston Celtics February Pick and Roll Defense

Statistic................Number
Total Points ...........353
Shot Possessions....336
Total Possessions....441
Makes...................137
Misses...................171
FG%......................44.5%
PPP.......................1.05
P4.........................0.80
Turnovers...............33
Resets...................72
Open:..Yes/No........165/276

However, when opponents actually attempted a shot or got to the free throw line, they had a normal 1.05 PPP. While the true league average for this specific play is unknown, it does fall in line with what the typical league average is for PPP.

Interestingly, opponents also had a 44.5 field goal percentage (137-for-308) on pick and rolls, which is just underneath the league's average of 45.2.

Out of the 441 pick and rolls run, the Celtics forced 72 resets and 33 turnovers. So, 23.8 percent of the possessions resulted in no shot attempt at all, which is why their P4 is significantly lower than their PPP.

Plays resulted in open attempts 165 times, and they were uncovered 276 times. This means, more often than not, an opponent was covered while attempting their shot.

But looking at these statistics still doesn't provide a truly clear picture of how the Celtics defended pick and rolls. There is a significant difference between "middle" and "side," and the statistics showed that Boston defends one far more effectively than the other. (The differences will be explained more in-depth in the following parts of this series.)

The C's faced 262 total middle pick and rolls in February, and allowed only 197 points, which is a P4 of 0.75. Opponents shot 44.5 percent and had 24 turnovers. In addition, the Celtics were able to force 41 resets. This means that had a PPP of 1.00, which is slightly above average as a defense.

But on side pick and rolls, the Celtics ran into trouble. They faced 179 of them over the course of the month and allowed 156 points, for a P4 of 0.87. However, out of the 139 shot attempts, they allowed a horrific 1.12 PPP. The C's let up points at a high rate, as they weren't able to force teams into many difficult plays, with only 9 turnovers and a 48.4 field goal percentage.

Brief Discussion

Throughout the month, the Celtics have done a reasonably nice job defending middle pick and rolls, with their P4 of 0.75 and PPP of 1.00. They want players attempting mid-range jump shots, so they have the big man defender "sag" near the free throw line, which means he "ices" the pick and roll. The on-ball defender is supposed to angle his body into the screen to force the ball-handler to go a certain direction. (For a little more information on this, read the "Briefing the Pick and Roll" section from here.)



In the three video examples above, the big man defender ices the pick and roll, giving the ball-handler the space to attempt a mid-range jump shot. This is the point of ice, since shots in this range are the lowest percentage in the entire NBA. However, savvy players are sometimes able to beat slow defenders to the rim with the space they are given, which is apparent in the third example with Kelly Olynyk.

When icing the side pick and rolls, the Celtics sometimes allow the ball-handler to get back to the middle of the floor, which opens up space for the pick and pop, or other shooters since help defense is required. The best example of this from the past month happened on three consecutive plays against the Sacramento Kings.



In all three instances, the Kings ran the same exact play: a side pick and roll with Isaiah Thomas as the ball-handler. But there were two different results, the first being an open three-pointer by Rudy Gay. This happened because Kris Humphries and Jerryd Bayless lost containment of Thomas, who snaked towards the middle of the floor. Gerald Wallace was forced to help (to prevent a drive to the rim), but that opened up Gay for the open jumper.

On the next two possessions, the Celtics adjusted by having the big man help across the paint, instead of having Wallace drop down. But this opened up Carl Landry on the pick and pop. Thomas found the open man in both instances for two good looks at the rim. Even though these mid-range jumpers are something the Celtics can live with, too often they allowed open attempts like these.

After all, there's a reason why the Kings ran the same exact play three consecutive times. They knew they could pick their spots because the Celtics have holes in their side pick and roll defense. But the key to good defense is being able to make adjustments on the fly, which is something Boston has gotten better at throughout the season.

Stay Tuned

Keep an eye out for the next few parts of the pick and roll series, where I will breakdown the Celtics' technique of "icing" the pick and roll, and their success when they decide to "hedge" or "switch." After that, I will go player-by-player in reviewing their affect on the team's collective defensive performance when they are put into a pick and roll.



bob



.