TeamSPACE: Excursions of an Award Tour
By Matt D'Anna
On Monday, we took a deep dive into the starting lineups for every NBA Championship team since 1996-1997, by visualizing each player’s Hunting Grounds of shot activity with TeamSPACE charts. Some clear visual patterns emerged regarding the influence of generational shifts, coaching philosophies, and the impact/dominance of a superstar player. Seemingly, we were done examining the spatial characteristics of these lineups.
We’re not.
By slicing and dicing the shot data this way, it warrants further consideration. By comparing the similarities and differences among each season’s best, we may be able to extrapolate trends over time, trends within dynasties, and the qualities of a successful team; the heart, er chart, of a champion I suppose.
Is this even useful? Maybe. Let’s find out. Enjoy the soundtrack (each header) along the way.
Specifically, there are four questions to address:
- Do championship teams occupy similar amounts of total court space?
- Do these teams exhibit similar levels of overlapping space?
- Is there a relationship between overall court space and overlapped court space with shooting percentage and win percentage?
- Do specific teams, coaches, or schemes exhibit similar spatial patterns over time, regardless of field goal and winning percentages?
First, here’s how the chart changes:
Transitioning from 5 players to 1 lineup
Then, there are three new calculations we can infer from this visualization.
- Total space a lineup occupies, represented as a percentage of the possible area (an entire half court)
- Overlapping space of a lineup, represented as a percentage of the total lineup space. This figure represents how often more than one player in a lineup is active from the same space on the court.
- Lineup field goal percentage. For the purposes of this study, we’ll keep it very simple: made shots divided by all shots.
Based on these set of metrics for each team, we can compare over time for similarity and upward/downward trending. Also, it’s important to note that we are working with a relatively small, skewed sample: championship teams are more likely to be similar to each other. Ideally we would compare every team since ’96-’97… But that’s for another day.
Now, some artwork:
Well that was fun…
…But what do all those colors mean?
First, let’s examine each lineup’s total court space over time:
Championship teams occupy between 4.4% (the ’12-’13 Heat) and nearly 12% (the Mavs) of the court. While there are some peaks and valleys every few years (most teams are using 7-9% of the court), there is not a clear upward or downward trend over time. A linear trend line is a virtually flat, ever-so-slightly-decreasing line. The hidden pattern is the subtle growth of the Spurs, but we’ll address that in a moment.
Second, examining the overlapped space over time we notice a different trend:
Ah ha! Much less chaotic than the total court space. There is a slow, steady (albeit not uniform across every team) increase in overlapping court space over time. Taken in conjunction with the previous graph, teams are closing the gap: total space is either remaining the same or shrinking, and overlap is growing. This is potentially indicative of changing offensive systems over time favoring multiple players being active from the same spots. This could be a shift from individual, freelance decision making towards systematic shot opportunities.
Third, let’s compare these scores to field goal and winning percentages:
Huh. Only one metric stands out: a moderate, negative correlation between total court space and field goal percentage. This would indicate that as court space increases, field goal percentage decreases, and vice versa. Stated differently: occupying more of the court results in worse shooting percentages, some of the time. This would point towards a lineup wanting to reduce its footprint of activity.
Side note: compiling shot distances for each lineup reveals something else: a moderate, positive correlation between total court space and average shot distance. This means that as a lineup uses more of the court, they shoot from farther away; an entirely logical and uneventful finding. There are non-meaningful relationships for all variables compared to overlapped court space. As for win percentage, this is a skewed sample (they all have high winning percentages – they won it all!) so nothing noteworthy emerges beyond the shooting/total space relationship.
Fourth, how does each dynasty compare?
When each team is compared this way, we’ve really got a collection of extremes.
- The Bulls use more total space, overlap less, and shoot worse than the other three; we’ll call this the #JordanRules, I suppose.
- The Heat gradually decrease their total space, increase their overlap, improve their shooting percentage, and increase their shot distance. Each time (well, twice) is like clockwork; likely the #LebronEffect.
- The Spurs have increased their total space and decreased their overlap over time, while their shooting has remained virtually the same. Considering each of their charts are very consistent, they appear to just keep refining their precision; #PopSchematics.
- The Lakers are the least interesting here, really. If anything, in the post-Shaq era shot distance increased by nearly one full foot and overlapping space nearly doubled, apparently a function of #MambaRingz
So, having read/experienced all of that — we have learned that similarities across franchises do exist. Six of the seven teams somewhat neatly organize into superstar-driven (Bulls, Lakers, Mavs) and team-driven (Spurs, Pistons, Celtics) shooting patterns. Superstar-driven teams generally occupy more total court space and less overlapping area. They would appear to operate in a more player-centric model. Team-driven champions generally occupy less total space and have more overlap, indicative of a plug-and-play system of scoring. One isn’t necessarily better than the other: superstar-driven teams have eight rings, and team-driven have seven. The outlier is the Heat, which visually and statistically are a hybrid of both models. They appeared to be a study in everything that was right about both models. Heading into next season, obvious questions center around how these models will survive and sustain. A few closing points:
- The superstar model hasn’t secured a title since Kobe in ’09-’10. The Oklahoma City Thunder are an eerily similar representation of that team.
- How Lebron applies what he learned in Miami will be interesting. During his first tour in Cleveland the superstar-model was unsuccessful. An opportunity for a championship may partially hinge on how quickly Love and Kyrie can settle into distinct and productive Hunting Grounds, a la Wade and Bosh.
- The Spurs model has endured multiple tests of time; what does the next iteration look like?
Any one of these would certainly be a plausible, logical and somewhat predictable outcome to next season. Which one prevails is the great unknown at this point.
Data and photo support provided courtesy of NBA.com, Basketball-Reference.com, and Darryl Blackport.