NBA.com/Stats Unveils Improved Matchup Data For 2019-20 Season
By Brian Martin
The emergence of player tracking data took defensive analysis to another level as it allowed us to break down individual player vs. player matchups.
Now in addition to the counting stats (blocks, steals, rebounds, defensive field goal percentage), hustle stats (deflections, box outs, loose balls recovered) and on-court/off-court stats (plus/minus, lineup data), we could see how well a player defended their individual matchup.
But how those individual matchups were assigned needed to be improved in order to account for the defensive schemes seen across the league. For the 2019-20 season, the NBA has updated its matchup data to more accurately account for switches and double teams that are so prevalent in today's game.
In 2018-19, a matchup possession was attributed to the defender if they covered the opposing player for the most time during the team’s offensive possession.
Let’s break down this play below as an example.
- With 20 seconds remaining on the shot clock, D’Angelo Russell crosses half court while being defended by Maurice Harkless
- With 15 seconds remaining on the shot clock, Russell gets a screen from Eric Paschall, which Harkless goes under to prompt a defensive switch with Montrezl Harrell
- With 13 seconds remaining on the shot clock, Russell launches a 3-pointer that Harrell contests and the shot is missed
In 2018-19, this possession would have been credited to Harkless since he defended Russell for five seconds in the frontcourt compared to two seconds for Harrell. However, it is Harrell that contests Russell’s shot attempt, while Harkless is inside the key defending Paschall at the time of the shot.
Heading into the 2019-20 season, the matchup data has received an overhaul to resolve situations like this and provide the most accurate information available. Now matchup data is being captured at every moment in the game, so if a switch occurs, two players will be assigned matchup time against that offensive player and the data will show who is actually defending the shots.
In the example above, Harkless would still have five seconds of matchup time against Russell, but Harrell would also have two seconds of matchup time and Russell’s shot attempt would be attributed to Harrell since he was the defender on Russell at the time of the shot.
Key Changes For The 2019-20 Season
- In previous seasons, matchup data was assigned by the number of possessions one player defended the other. Now, the data is shown as minutes guarded since possessions can be split between multiple defenders.
- Shots are attributed to who is matched up to that player at the time of the shot, rather than the player that defended the shot taker for the majority of the possession.
- Double teams are accounted for, so if an offensive player takes a shot while being double-teamed, that shot will be attributed to both defenders.
- Matchups are picked up only in the front court as there is little value to matchups that occur in the backcourt.
Finding The New Data
The new matchup data has replaced the previous data, going back to the 2017-18 season, when matchup data was first introduced. So the locations for matchup data have not changed on NBA.com/Stats, but the information provided has improved.
On any player page, clicked the link at the top left (default is Profile) and select Matchups at the bottom of the drop down menu. Let’s run through an example, using Kawhi Leonard.
The Matchups page displays the individual matchups for Leonard so far this season. Two things to keep in mind when the page initially loads, it defaults to Kawhi being on offense and the stats are being displayed on a per game basis. Both of these can be switched at the top of the page in order to find Kawhi’s defensive matchups and look at the stats in terms of totals rather than averages. At this point in the season, totals and averages against a single player are one in the same, but once the Clippers get deeper into the season and face teams more than once, this could come into play.
Here is a look at which players Kawhi has defended most often through the first two games of the season. The two columns to note here are Matchup Min and Partial Poss as they represent the new way in which matchups are calculated.
- Matchup Min: Rather than showing possessions as in years past, now defense matchups are captured by the total time that a player was matched up with another.
- Partial Poss: The sum total of partial possessions that were spent defending that player. For example: if Kawhi guards LeBron for 10 seconds of a 20-second possession, that is a 0.5 partial possession that would be added to this column for this matchup.
The rest of the columns carry over from previous seasons and essentially provide standard box score stats during that matchup. All of the stats are attributed to the offensive player in this situation with the exception of Team PTS, which tracks how well the offensive team scored while that particular matchup was happening. Sometimes a player can shut down his individual assignment but the opposing team can still find success. In this case, in the 4.1 minutes that Kawhi guarded Robinson, Robinson scored four points, but the Warriors as a team scored 20.
The team pages offer the same matchup data as the individual player pages, but include all matchup data for all players on that particular team. If you want to see how every Clipper defender fared in their matchups against a particular team or player, you can do that here.
Matchup data can also be found at the individual box score level and since it is focused on just a single game – rather than the entire season like the player and team pages – more details can be provided about each matchup.
To find the Box Score Matchup data, open any box score and click the link next to the words “Box Score” on the left just below the line score. The default option is “Traditional” – click that to launch the drop down menu to more advanced box scores, including Matchups at the bottom of the drop down menu.
The page will display every player matchup from the game, with the ability to select specific players and swap them between offense and defense. In the image above, we see the most frequent matchups (in terms of matchup time) from the Clippers-Warriors game on Thursday.
In addition to the standard box score data for each matchup that we saw on the player and team pages, the Box Score Matchups page offers additional info, such as percentage of time for both the offensive and defensive players (%OFF Time and %DEF Time) in the matchup as well as percentage of time both players in the matchup were on the court together (% Time Both On).
For example, Landry Shamet defended Stephen Curry for 5:36 seconds during Thurday’s game. While 5:36 doesn’t seem like a lot of time during a 48-minute game, there are a few things we have to keep in mind:
- Both Curry (30 minutes played) and Shamet (34 minutes played) have to be on the floor at the same time; they shared the court for 59 percent of all time in the game (59% of 48 minutes is 28.3 minutes).
- Matchup time is added only when Curry is on offense and Shamet is on defense so the Warriors have to have the ball, which essentially cuts the above time in half (half of that 28.3 minutes is 14.2 minutes).
- Remember that only time in the front court is counted towards matchups, which further reduces the matchup time.
- Shamet defended Curry for 48.1 percent of his defensive time in the game, while it was 54.3 percent of Curry’s offensive time (due to playing fewer minutes).
- Shamet defended Curry for 25.1 partial possessions, so he was not the sole defender on Curry on every Warriors offensive possession, which again further reduces matchup time.
In that 5:36 of Warriors offensive time with Shamet defending Curry, Curry made four of his nine shot attempts, but came up empty on three attempts from beyond the arc. He also drew a shooting foul and made both of his free throws. Curry scored 10 points during his 5:36 against Shamet, while the Warriors as a team scored 32 points.
This new method of calculating matchups will provide more accurate data to analysts of the game and will help defensive analytics take another step in the right direction.