Finals Preview: Warriors vs Cavaliers — an analytical dive into offense and defense

June 1, 2017 by J.H. Yeh

With the NBA Finals coming up on this Thursday, featuring two of the most dominant teams ever (combined record of 24-1) in playoffs history and 11 players who have appeared in all-star games from both rosters, clashing in at the pinnacle of the NBA postseason for an unprecedented three years in a row, we start to wonder who will take home the Larry O’Brien trophy this time after splitting the previous two meetings.

Aside from the stars decorating the playoffs and the rivalry between the two franchises, I figured it’s time to delve into the analytics once again for a deeper look into both teams — to see which team has the edge to win it all.

The first of the many comparisons between the two elite teams is Dean Oliver’s “Four Factors.” Dean Oliver, a famous statistician and pioneer in basketball analytics, came up with the four most important components to win a basketball game, and he gave the moniker to those four elements  — as simple as it sounds — the “Four Factors.” The Four Factors are, in the order of importance, shooting percentage, turnover rate, rebounding rate, and free throw rate (how often a team gets to the free throw line). I think it is appropriate to start this presentation by taking a look at each team’s Four Factors as both the Golden State Warriors and the Cleveland Cavaliers have almost identical Four Factors percentages in the regular season (playoffs are slightly different due to the small sample size). Note that in the tables below, I further expanded the Four Factors and apply that to both offense and defense — which essentially yields eight factors. I also broke down the rebounding percentages to include offensive rebounding percentage and opponent offensive rebounding percentage. In addition, I replaced the traditional FG% with effective field goal percentage for an understanding of the shooting performance. Not surprisingly, both elite ball clubs share a lot of similarities within the Four Factors — with the Warriors ranked first and the Cavaliers ranked second in eFG% in both the regular season and the playoffs.

26

Capture

Next up is the shooting and point distribution for both teams. The most statistically visible change for the Cavaliers is the 5.9%  increase in points per game ( from 110.3 pts/G  in the regular season to 116.8 pts/G  in the playoffs).

1

20.PNG

As expected from both analytically-savvy teams, the Cavaliers and the Warriors run their offense through perimeter shooting and beneath the basket while deliberately avoiding the mid-range. The Cavaliers, in particular, only score 3.3 points per game from the paint (excluding the restricted area), which accounts for only 3.0% of their total scoring — both feats are ranked 29th in the league, just behind the equally analytically-savvy Houston Rockets. However, that number is upped from 3.3  points in the paint per game to 8.0  points due to the increased intensity in the playoffs as the opponents congest near the basket for better defensive stance and dare the Cavaliers to take more less-efficient range shots from the paint and the mid-range. Similarly, the same phenomenon is also occurring to the Warriors as they are forced to take more of those less-efficient shots.

Despite that, both the Cavaliers and the Warriors’ mid-range points per game and point distribution (against total points scored) still hover around the league average because they have players who excel in shooting mid-range shots such as Kyrie Irving, Shaun Livingston, and Kevin Durant. The presence of these players offer some offensive flexibility in the expense of taking fewer shots (that can generate higher returns) from analytically desirable territories. This forms a stark contrast against analytically-savvy teams such as the Rockets and the Brooklyn Nets, whose mid-range games are non-existent as the Rockets only average 2.5 FGM/G and the Nets averaging 4.2 FGM/G.

Moving away from the restricted area, which is the most efficient zone to shoot from, the second most efficient territory to shoot from is the 3-point territory — specifically, the corner 3s. The Cavaliers excel in that area as they scored league-leading 12.8 points per game from corner 3s in the regular season — and that number is increased to 14.1  in the playoffs thanks to the mid-season acquisition of Kyle Korver. Taking up 12.1% of their scoring in the playoffs (also league-leading), the Cavaliers are taking full advantage of the corner 3s and it has since become a significant chunk of their offense.

Trailing behind the Cavaliers by a whopping 5 points, the Warriors aren’t a strong corner- 3-shooting team for an elite shooting team as they score 9.1 points from the corner 3s  (5th place) in the playoffs but they generate an overall more balanced offense from their shooters as they have scoring coming from all regions, including the paint. With both teams built to shoot from the perimeter, the corner 3s can potentially be an integral factor on determining the outcome of the Finals — depending on how well the teams utilize the corner 3s while preventing the opponent from doing the same.

The Warriors are strong in transition throughout the regular season and the playoffs; they are ranked first with a 18.5% of running transition in a given possession, and that percentage is upped by  1.3% in the playoffs. The Cavaliers, on the other hand, are lagging behind the Warriors by 4.6% in the playoffs and are also downed by 1.1% from their regular season performance. Although the Cavaliers are the slower team (in terms of running transition), they are the more efficient team at generating points in each transition during the playoffs. The Cavaliers have a point per (transition) possession — also known as PPP — of 1.20  points, a 0.18  difference from the Warriors. This may seem trivial; however, on average, an NBA team runs about 15 to 17 transitions and a 0.18 point differential in PPP can mean earning potentially  3 extra points in the entire game. The Warriors’ shortcoming in PPP may be crucial as close games are expected from the series.

3

2

Nonetheless, the Warriors’  1.02 PPP and a decrease in transition points per game should not discount their superior offense because that doesn’t mean they are a weak transition team. Throughout the playoffs, opponents are more committed on setting up transition defense in order to slow the Warriors instead of collecting offensive rebounds for second chance opportunities. This strategy is also a trend in modern basketball as teams are going small (click here for the article). With more emphasis on transition defense to drive down the Warriors’  PPP and their transition efficiency, the Warriors, as a result, see an increase in free throw frequency and turnover frequency in transition. The most notable change during the Warriors’ transition is the visible drop field goal percentage ( by 8.7% – from 56.6% to 47.9%). Contributing factors such as tightened transition defense play a role in the Warriors’ huge drop in transition FG% during the playoffs, but one key factor in explaining the drop is teams are now daring the Warriors to shoot at every chance they manage to find when the Warriors are on a fastbreak. Challenging the Warriors to take more range shots instead of allowing them to shoot a contested layup has been the norm in slowing the Warriors during the transitions — which contributes to their drop in transition FG%.

Drives are not to be confused with transitions, or drives to the basket during transitions. According to the NBA, drives are recorded when a player advances to the paint during a non-transition situation when the opponent’s defense is already established. For another words, drives and transitions should be treated differently.

56

Both the Cavaliers and the Warriors are built to score from range shots and neither team is an overall strong driving-heavy team since their rosters are constructed with shooters. The Cavaliers are mediocre at best at their driving game while the Warriors are ranked last in drives per game and points scored from driving in both the regular season and the playoffs. Despite that, both teams are incredibly efficient at finishing strong at the basket thanks to LeBron James, Kyrie Irving, Kevin Durant, and Stephan Curry, who are excellent drivers and finishers. Furthermore, the high FG% is also achieved with the help from their floor-spacing teammates. Having two very efficient drivers from both teams really brings up their respective teams’ overall driving FG% because no one else on each roster has averaged more than 3  drives per game. James, Irving, Durant and Curry are responsible for most of their teams’ drives and scoring (from drives). Below is a breakdown of the drives. Note that in the playoffs, James and Irving together are responsible for 75.5%  of their team’s total drives and Curry and Durant are accounted  for 52.0%  of their team’s.

30

31

One glaring problem for the Warriors is their big increase in driving TO% as the Warriors’  TO% went from 6.3% in the regular season to 8.9% in the playoffs. While sturdy defense is expected in the playoffs, the Warriors’ spike in TO% may suggest poor decision making in the face of tight defense and failing to protect the ball when driving into the paint due to the opponent double-teaming the driver.

Now that we have transition and driving, here comes another important, yet fundamental, part of the game — passing.

7

8

The Warriors are the obvious winner as they are far superior than the Cavaliers in every passing category because everyone in Warriors jersey can pass. Moreover, the Warriors are far ahead of the Cavaliers in passes made (by about 36.1 passes in the playoffs) and secondary assists (by 3.3 ). Secondary assist is the pass that comes before an assist and is a good metric to gauge good ball movement. Warriors have the highest secondary assists ever recorded, both in the regular season and the playoffs, since the league started keeping track of such stat in the 2013-14 season. The Warriors also lead the playoffs teams in free throw assists. Free throw assist is the pass that leads to at least one made free throw, excluding  and-1. Free throw assist can be viewed as an assist because free throw assist is a good complement for assist as it tells the whole story about a team’s overall offensive execution.

The Cavaliers are significantly lower in passes made, potential assist, and secondary assist because the team features two ball-dominant players in James and Irving, who average second and 4.41 seconds (out of 24 seconds) per touch on the basketball. James and Irving are great passers and much of the Cavaliers’ passes come from them, which reduces the number of passes made per game and secondary assists ( and other passing categories). Interestingly, both teams see a decrease in potential assists but an increase in eFG%  (refer to the Four Factors chart) during the playoffs run. Essentially, a potential assist is a pass that leads to a missed shot attempt — hence the name “potential.” (Click here for article about these kinds of assists)  While fewer passes are made in the playoffs due to greater defensive pressure, one should not see a decrease in potential assists as a sign of bad offense. With higher eFG% in the playoffs and lower potential assist numbers, this shows both teams are very efficient in scoring and are reducing the passes that lead to a missed field goal attempt.

Since we have all the relevant numbers with driving and passing, and their related stats such as FGM from drivings, I plotted number of drives against different potential outcomes to create a visual presentation of how the Warriors and the Cavaliers perform.

6061626364

It’s quick to ascertain that the Warriors tend to pass more during drives, and as a result, have higher number of assists. The Cavaliers, on the other hand, pass less but they are aggressive at driving and are generally effective at finishing at the rim or drawing fouls. In addition, both teams are good at preventing turnovers during the drives.

6567

70.PNG6869

In the playoffs, the Cavaliers are able to keep up with the defensive intensity by notching more FGM while the Warriors falter in that respect. Both teams see a decrease in assists made but the Warriors manage to stay above the average on creating assists from drives. Both teams are also passing less in the face of tougher defense but the Warriors still manage to stay above the average. The Warriors are committing more turnovers in the playoffs than the Cavaliers but both teams are relatively stable in drawing fouls — with the Cavaliers drawing slightly fewer fouls than they did in the regular season.

The next logical step to look for is how teams shoot. Shooting can be classified into catch & shoot and pull-ups. The Cavaliers are very effective in finding the open men for a quick catch & shoot.

2324

The Cavaliers are playing their best offensive game ever as they are first in the league in catch & shoot 3FG% and points generated from catch & shoot 3s. They rank no lower than the sixth place in the rest of the catch & shoot categories in the regular season. In the playoffs, their catch & shoot 3FG% is further improved to an absurd 47.7% — and a big part of this comes from their effective use of the corner 3s and their recent addition of Kyle Korver. (Click here to read about catch & shoot and corner 3s)

Similarly, the Warriors are also extremely prolific with catch & shoot. They produce the highest points in the regular season from catch & shoot and the second most in the playoffs (behind the Boston Celtics). Due to the high volume of scoring, the percentage of their catch & shoot points is lower than the Cavaliers because they have Durant and Curry, who are highly effective in creating their own shots.

That said, in the face of disciplined defense, catch & shoot strategy doesn’t always work out and oftentimes teams need to create shots for themselves. The pull-up shots are less efficient and not analytically desirable in the eyes of the management people but the ability to create shots is the key to gain a rather critical competitive edge.

21.PNG

22

As expected, both teams score more from pull-up shots and their FG% and 3FG% also see a rise in the playoffs. The Warriors experience a 6.5% increase in their overall FG% (the highest increase among 16 playoffs teams). Much of this increase is attributed to Durant and Curry’s capability to effectively knock down contested mid-range shots during isolations. The Warriors’ biggest reason on their high efficiency is Durant, who has an incredible eFG% of 64.5% on pull-ups shots — first in all the players in the playoffs rosters who have attempted at least three pull-up attempts in a game. Another reason for Warriors’ high eFG% is because Durant are taking fewer shots from the perimeter (only 30.4% of his shots come from beyond the 3-point arc this season, as opposed to Curry’s at 54.7%) — a result of diminishing values of adding an additional shooter.

Finally, here comes the defense. As a big fan of defense, I created a regression line using blocks per minute and opponent FG% at the restricted area to see both teams’ actual defense at the rim relative to their blocks.

28

28.PNG

The Warriors average the most blocks throughout the regular season and the playoffs but their regular season blocks over-estimate their actual defense at the restricted area. That said, blocks are not a good indicator to gauge a team’s, or a player’s, actual defense at the rim (click here to read more about it). The Cavaliers experienced ups and downs with their defense in the second half of the regular season. However, both the Warriors and the Cavaliers made significant adjustments in the playoffs and are now second and fourth in opponent shooting at the rim. The Cavaliers improved their rim protection from 0.618  in the regular season to 0.580  in the playoffs and the Warriors went from 0.598 to 0.566.

Most importantly, now the Warriors’ average blocks under-estimate their actual defense at the rim. To sum up, both ball clubs have become significantly better at preventing shots in the restricted area.

Conclusively, the Warriors have a competitive advantage in winning the finals. Though they share similar efficiency with catch & shoot, they are far superior than the Cavaliers in pull-up eFG% (especially the mid-range). Not only they are elite in creating their own offense, they are also very good at finding the right players at the right time as manifested by their strong statistics on every passing category. Moreover, the Warriors have an improved defense at the restricted area from the regular season and last year’s playoffs (they were 0.572 last year) to prevent the James and Irving from attacking the basket. One key concern is their effectiveness in running the transition as they are turnover-prone.

Defensively, the Warriors will need to stretch their defense to the corner to prevent the Cavaliers from running their offense through the corner 3s. Furthermore, the biggest task for them is to find every way possible to force James and Irving (and the rest of the team) to take pull-up shots since 25.9% of the Cavaliers points come from catch & shoot and 24.1% comes from catch & shoot 3s  alone. Another tactic to slow the Cavaliers is to lock-in the perimeter and the restricted area and only allow the Cavaliers to take mid-range shots. Finding different ways to shave-off even a tenth of the Cavaliers points per possession can go a long way in this series.

The Cavaliers, on the other hand, can put up a fight and this series will be a crucible and ultimate testimony for the role payers and the bench. The best way to challenge the Warriors offense is to have wing players to strengthen the perimeter defense by yanking the shooters (particularly Curry) from the 3-point arc and forcing them to drive to the basket to increase the odds of turnovers — just little things to put themselves in a position where they might get lucky. As mentioned before, the Warriors miss more field goals than their per drive stats suggest (refer to the “Drives and FGM – Playoffs” image). Additionally, the Warriors are not an efficient transition team even though they run the most transitions throughout the season, including the playoffs. Cavaliers players should do everything they can to get back on transition defense as soon as they conclude their offensive possession, give up on offensive boards (have one player box out for offensive rebound while the other four players get back on D), and challenge the Warriors to shoot range shots by clogging the paint. Allowing the Warriors to get into rhythm is the last thing the Cavaliers need from the Warriors.

Offensively, the Cavaliers will need to get players other than James and Irving involved by stretching the floor to execute drive-and-kick strategy and exploit the corner 3s even more than ever. Properly manage the offense by driving strategically and taking slightly efficient shots (catch & shoot from the corners) before the Warriors defense recovers can certainly make up the shortcomings on passing.

— J.H. Yeh

(all tables and graphs are created by me, and all sources of stats are courtesy of NBA.com and SportVU player tracking as of May 28, 2017)

%d bloggers like this: