2014 NBA Preview: Key Offseason Pickups Should Give The Cavs A Boost

FiveThirtyEight treats the NBA’s return like Cleveland treated LeBron’s homecoming. But while the people of Ohio cheered, we spreadsheeted. To get ready for the season, we took each player’s projected Real Plus-Minus and wins above replacement, calculated a total for each team, and ran 10,000 simulations of the NBA schedule to divine likely records and championship odds.1 We’ve split the teams into the lower and upper tiers in each conference; these are the eight likely playoff teams from the East. (We previewed the West’s lower and upper tiers here and here, and the East’s worst teams here.) Use our stats, x-factors and regressions to prepare in the few remaining hours before Tuesday’s games.

Three years ago, the Charlotte Hornets (né Bobcats) were about to embark on a season that would ultimately set NBA records for futility, including the fewest wins (seven) and the lowest winning percentage (.106) any team ever posted in a single season.2 At the time, it was almost inconceivable to think the same sad-sack franchise could somehow produce a winning record within the span of a few seasons, but that’s exactly what happened last year, as Charlotte finished 43-39 on the strength of good performances by Al Jefferson, Kemba Walker and Josh McRoberts.

Sabermetrically speaking, teams that dramatically improve their record from one year to the next tend to experience a noticeable downturn the following season, an effect known as the “Plexiglas Principle.” An NFL team that improves by five wins from one season to the next, for instance, can expect to step back by about 2.4 wins the following season, as some of the apparent improvement is because of luck and not because of an actual upgrade in skill.

The same effect exists in basketball, and that’s not good news for the Hornets. Remember, Charlotte needed to improve by an average of 17 wins per season to get from where they were back in 2011-123 to where they finished last season. Luckily, though, the Plexiglas effect doesn’t seem to be as strong in the NBA as it is in other sports.4 A team that improves as much as Charlotte did last season (a net of +22 wins) can expect to revert toward its previous win total by 4.6 wins, which would put the Hornets somewhere between 38 and 39 wins in 2014-15.

The same goes for teams that improve greatly over the span of two seasons. Historically, a team that improves by 34 games over a two-year period can expect to decline by about 5.7 wins the following season, which would leave the 2014-15 Hornets with about 37 wins this season.

Now, the consensus among Las Vegas bookmakers is that Charlotte will win somewhere between 43 and 46 games in 2014-15. The Hornets are young — particularly after adding 24-year-old budding star Lance Stephenson — and are generally perceived to be a team on the rise, especially with the return to their original 1990s name starting this season. But Charlotte also overachieved last season, and on top of battling the Plexiglas Principle, its Eastern Conference foes have only improved over the summer. As counterintuitive as it sounds, Charlotte could win slightly fewer games this season and still have made progress from a talent perspective. — Neil Paine

This summer was not a good one for the Indiana Pacers. First, free agent guard Lance Stephenson declined the team’s five-year, $44 million offer, signing instead with the Charlotte Hornets, an up-and-coming conference rival. Then, at the beginning of August, star forward Paul George suffered a catastrophic leg injury while playing in a scrimmage for Team USA. The best-case scenario for George puts his return at the beginning of the 2015-16 season. Between assists (not counting assists to each other) and their own scoring, George and Stephenson had a hand in just over 50 percent of the Pacers’ points last season — 3,989 in all. That volume of offensive production would be difficult enough for a good offense to account for, but George and Stephenson were the two best pieces in an offense that struggled as a whole. Their individual skills were often the only things keeping the Pacers’ offense from complete implosion. To see just how much offense the Pacers are losing, we can look at Seth Partnow’s True Usage statistics. This measure of total offensive involvement incorporates a player’s assists and potential assists5 to elaborate on a player’s traditional Usage Rate. True Usage is the percentage of a team’s offensive possessions that the player was involved in — whether by taking a shot, getting fouled, turning the ball over or creating a scoring opportunity for a teammate. The table below compares last year’s numbers for George and Stephenson with the numbers for Rodney Stuckey and C.J. Miles, whom the Pacers signed this summer to help fill minutes on the wing. By True Shooting Percentage, both Stuckey and Miles were both reasonably productive offensive players last season. But they weren’t George and Stephenson. Miles was efficient, but in a much smaller offensive role. Stuckey played an even smaller role and was far less efficient. The absence of George and Stephenson will put pressure on Roy Hibbert and George Hill to step into bigger roles, and for some new additions, like Stuckey and Miles, to take on different ones. It’s possible that all these obstacles will inspire the Pacers’ coaching staff to be more creative. But even that may not be enough. Indiana’s story will likely be a simple one: Average offensive team struggles without its two best offensive players. — Ian Levy Despite winning just 38 games, the Hawks once again made the playoffs last season — for the seventh straight season — and, once again, exited in the first round (the fourth time that’s happened during the current run). To their credit, the Hawks managed to squeak into the postseason despite the loss of star big man Al Horford to injury for most of the year. This year, our model projects Atlanta to finish with 44 wins, sixth in the East.6 In continuing the team’s trajectory of modest success, head coach Mike Budenholzer will probably emphasize Spurs-y fundamentals: passing and long-range shooting. Last year’s Hawks could really swing the ball around, finding one another in the right spots. The team led the league in assist percentage — or the percentage of made baskets that were assisted — at nearly 67 percent, compared to a league average of 58 percent. That relative assist percentage is good for top 20 all time. Last year’s Hawks could also shoot. The team ranked second, behind the New York Knicks, in made 3-pointers per 100 possessions, at 9.8. While other teams such as the Houston Rockets and the Golden State Warriors were more prolific from behind the arc (that is, they made more 3-pointers, or even more per minute), those teams also played at a faster pace. Ranking in the top two in both assist percentage and 3-point shooting is no accident. Better passing enables a team to find open shooters. And the Hawks’ system is trying to emulate the San Antonio Spurs’ passing. As Zach Lowe wrote in July: The Hawks under Budenholzer are not going to pound the ball with isolations and stagnant pick-and-rolls in the middle of the floor. Budenholzer wants to build a sort of Spurs East, with the ball whipping from side to side in an unguardable blur of passes, handoffs, and picks. So there must be a strong relationship between assist percentage and long-range shooting, right? Nope. There was none that I could find. Data from just last season, or the last 10 seasons, reveals no statistically significant relationship between a team’s predilection for 3-point shooting (after adjusting for its pace) and the share of baskets that were assisted. In other words, there is no tried and true connection between better passing teams making more 3-pointers. For every good-passing, 3-point-draining team like the Hawks, there’s one like the Knicks that shoots just as well (or better) without the ball movement. And, conversely, there are equally good passing teams who make far fewer shots from long range, such as the Chicago Bulls. The Hawks will likely to continue to shoot well — especially with sharpshooter Kyle Korver — and assist on many of those shots. But the diversity of NBA pass-and-shoot styles makes it hard to generalize. — Andrew Flowers Last year, all the Washington Wizards had to do to make the playoffs for the first time in six years was hold strong on defense and improve their offense from putrid to average. If that sounds like faint praise, it isn’t meant to be: The improvement in the Wizards’ offensive production last season was significant, and it sheds light on the team’s 2014-15 outlook. In 2012-13, Washington scored 5.4 fewer points per game than the league median. The only two teams that scored less efficiently in the past five seasons were the historically bad 2011 Charlotte Bobcats and 2014 Philadelphia 76ers. The Wizards’ frontcourt provided little skill to match John Wall’s dynamism off the pick-and-roll, no offensive rebounding ability, and poor spacing. There was a lot to fix. Marcin Gortat helped fix it. Gortat, traded from the Phoenix Suns, had honed his craft running pick-and-rolls with the master, Steve Nash. He ably stepped in as the roll man for Wall. Gortat shot 54 percent on 2-pointers and, according to Synergy Sports, pick-and-rolls were the largest source of his offense. Gortat also helped Wall fix himself. Last season was Wall’s best as a pro. He led the league in assists and was second (behind Chris Paul) in Assist Rate7 at 40.9 percent. According to Synergy, almost half (45 percent) of Wall’s assists came off pick-and-rolls, and Gortat accounted for 152 of those assists. Wall shot better, too. His 2-point shot, bemoaned by many in his first three seasons, climbed to 46 percent, around average for a starting NBA guard. And while Wall will never be a knock-down jump shooter, he shot a serviceable 35 percent from three on 308 attempts.8 All told, Washington improved its 2-point shooting by 6.5 percent last season. This is a large jump, but it only got the Wizards up to just above league average. As currently constructed, the Wizards offense will likely never be a top-10 unit in the league. But the growth of John Wall and the stingy defense should be enough to get Washington back into the Eastern Conference playoffs. — John Ezekowitz The Toronto Raptors made the first significant signing of the summer free-agent season, and, outside of LeBron James, it may have been the most valuable one. The Raptors re-signed point guard Kyle Lowry to a four-year,$48 million contract, an under-the-radar move as other teams vied for James, Carmelo Anthony and Chris Bosh.

After stints in Memphis and Houston, Lowry has blossomed into one of the toughest and best offensive point guards in the league. He ranked third among point guards in wins above replacement and ninth in points created off assists, which includes passes that led to free throws. According to Synergy Sports, Lowry’s pick-and-rolls created 0.89 points per attempt, the eighth-best mark among starting point guards. The only issues with Lowry’s offensive game are his propensity for turnovers — historically he has averaged more than 2.5 per game — and a strange inability to finish drives to the basket. Lowry only made 53 percent of his attempts within 3 feet of the basket last season, well below the league average of closer to 60 percent.

In short, Lowry is a top-10 point guard at a time when the position has seen an incredible influx of talent and has taken on additional importance. That alone makes his below-max salary (in line with Steph Curry’s very favorable deal) a bargain.

Amir Johnson, the Raptors’ power forward, also has a bargain of a contract, but he isn’t as flashy as Lowry. His early years in Detroit were nothing special, but he has matured into an underrated defensive force inside. And that’s most visible when you look at his stats in SportVU’s optical tracking system. When Johnson “challenges” a shot in the paint,9 opponents shoot 47 percent, the fifth-best defensive mark in the NBA for any power forward. And when Johnson is paired with impressive young center Jonas Valanciunas, opponents only take 29 percent of their shots at the rim, deterred, presumably, by Johnson and Valanciunas’s presence. If Johnson and Valanciunas had played every minute at that rate (unlikely, I know), they would have allowed the fewest shots at the rim of any team in the league.

Johnson will likely split time with Patrick Patterson, who provides the floor-stretching shooting that Johnson lacks. Nevertheless, Johnson’s defense, toughness and singing ability are huge assets for the Raptors. Toronto appears to have done what other teams around the league would love to do: secure two quality starters in free agency on cheap contracts. — John Ezekowitz

The first six years of Miami Heat head coach Erik Spoelstra’s career have been nothing if not highly eventful. And yet despite his success, questions still linger as to whether Spoelstra is overrated, underrated or properly rated as a bench boss.

There’s evidence pointing in all directions. In Spoelstra’s first two years at Miami’s helm, when the eventual signing of LeBron James was a mere glimmer in Pat Riley’s eye, the Heat exceeded expectations10 by nearly seven wins per season — the mark of a very good coach. Along the way, he cajoled surprisingly solid seasons out of the likes of Quentin Richardson, Dorell Wright and an aging Jermaine O’Neal, and successfully managed Dwyane Wade’s return from the knee injury that had sapped him of 31 games and nearly all his effectiveness in 2007-08. (Spoelstra also remains the only coach to extract anything approaching usefulness out of Michael Beasley.)

Then again, Spoelstra also oversaw the Big Three’s relatively disappointing regular-season debut, one which was marred by frequent meltdowns in close games and saw Miami undershoot expectations by nearly 11 wins (according to the projection methodology outlined above). In fact, during the James-Wade-Bosh era, Miami underperformed projections by an average of more than four wins a year. Part of this was undoubtedly due to the unavoidable effects of diminishing returns, as some element of the individual benefits James, Wade and Chris Bosh brought to their respective teams before 2010 was lost when the three joined forces with only one ball to share. But you could argue that the sheer amount of talent Spoelstra had to work with should have been able to average more than 58.8 wins per season11 over the four years they were together.

That’s why the post-LeBron period will be yet another referendum on Spoelstra’s coaching chops. He’s steered young, undermanned teams to 43 wins and a first-round playoff departure. He’s shepherded teams with all-time-great talent to 66 wins and an NBA title. The 2014-15 Heat don’t quite fit either of those templates — and that means they’ll provide Spoelstra with another variety of challenges by which to prove himself. — Neil Paine

The Chicago Bulls are projected to win 50 games this season, good for second in the East, with a 19 percent chance of advancing to the NBA Finals, using a purely statistical model.

Compare that to the projection of our colleagues at ESPN The Magazine, who expect the Bulls to finish second in the Eastern conference with 53 wins, or to the Vegas over/under line, which puts the Bulls’ win total at 55.5, the third-highest of any team.

The Bulls’ success hinges on which version of Derrick Rose they get: the elite, pre-knee-injury Rose or the frankly below-average performer in 10 regular season games last year. A purely statistical projection will heavily weight Rose’s recent performance, marking him as slightly above average, but nowhere near elite. The fact that he didn’t play much the past two seasons means he will regress to the mean more sharply.

Our model uses an imperfect metric: Real Plus-Minus (RPM). An advanced basketball metric, RPM is a rate statistic, in that it measures a player’s performance per unit of time on the floor. It does not take into account how much that player plays. In other words, it’s an efficiency statistic. Nonetheless, it’s a useful gauge of how well Rose played.

Before his lost season and the abysmal 10 games last year, Rose was spectacular in the 2010-11 and 2011-12 seasons, when he registered RPM scores of 3.2 and 4.3. But heading into the 2014-15 season, Rose is projected to post an RPM of 1.6, somewhere between the 80th and 85th percentile.12 (The median projected RPM of all 467 players this coming season is -1.3.)

To examine players comparable to Rose going back to 1977, let’s pivot away from RPM and use a similar statistic called SPM (Statistical Plus-Minus)13. By this metric, Rose had extremely successful 2010-11 and 2011-12 seasons, chalking up SPM scores of 4.9 and 4.7, respectively.

Below is a table of the 20 (and yes, there were only 20) players who had two or more seasons with an SPM of 4.5 or higher before the age of 25. (Again, this data only goes back to 1977.)14 The second column shows how those players performed in their age-26 and age-27 seasons, to compare to how old Rose will be this season.

If Rose performs on average with how others on this list performed, his SPM will jump to 6.3. That would make him a little better than Chris Paul and would give him the third-highest SPM in the league (behind LeBron James and Kevin Durant). But that assumes Rose does not face any injury complications, and that any rust completely disappears. Given his injury history, that’s not very likely. Rose playing at the level of Paul is the best-case scenario.

Let’s imagine, instead, that Rose regresses to the worst level of performance of any player on this list: Andrei Kirilenko, who averaged a 2.7 SPM. That’s still an excellent, All-Star level of play. That’s the rating Mike Conley is projected to earn this season, and it’s just above that of Kyrie Irving and John Wall.

But if, as the SPM-only model projects, Rose is merely a 75th-percentile player, with a -0.5 projected SPM, he’d be Jerryd Bayless. Jerryd who? Never mind, Bulls fans. Never mind.

The Bulls go as Derrick Rose goes. And that is a very uncertain path. — Andrew Flowers

The LeBron Show has a new cast: A seemingly perfect sidekick, Kevin Love; some bright young things, Kyrie Irving and (potentially) Dion Waiters; trusty bench specialists, Shawn Marion (defense) and Mike Miller (shooting); and an energetic, post-defending, rebound-gobbling center, Anderson Varejao. And, for comic relief, Varejao’s hair.

But for James to do maximum damage as a penetrator and distributor, he needs space. Spacing comes with good shooters, and these new guys are just as good (and maybe better) than the old ones at shooting threes.

Below are the 3-point shooting attempts and percentages for the top 10 players, by minutes played, on the 2013-14 Miami Heat as well as the projection15 for the same 10 players on this coming year’s Cavs team.

This season’s top 10 Cavs players are projected to shoot slightly more 3-pointers than last year’s Heat (1,681 vs. 1,675) and at a slightly higher rate (36.8 percent vs. 36.1 percent).

James was a good 3-point shooter last year (hitting 38 percent), and is expected to hit 37 percent this season, but on fewer attempts. And these projections assume that last year’s Cavs players, like Irving and Waiters, won’t shoot better by playing with James (as last year’s Heat players probably did). So this Cavs team has gunners — and that excludes the possibility Cleveland signs Ray Allen (38 percent 3-point shooting) and doesn’t count sharpshooter James Jones (52 percent on threes).16

But it’s the distribution of shooting ability that’s interesting here. Last year’s Heat and (most likely) this year’s Cavs each feature two non-shooters in their top eight rotation players. On the Heat last year, Chris Andersen and Dwyane Wade shot threes on only 4.4 percent and 4.2 percent of their attempts, respectively; Thompson and Varejao basically did not shoot threes for the Cavs last season. It will be fascinating to see James swap a team of 3-point shooters and one non-shooting wing (Wade) for a team of gunners where the non-shooter is a big (Varejao, and occasionally Thompson).

Beyond spacing and shooting, other huge questions linger. How will David Blatt distribute ball usage among the star trio of James-Love-Irving? Will Irving, Love and Waiters’s offensive potency help outweigh their mediocre-bordering-on-lackadaisical defense? Those are big unknowns. But it’s a safe bet that the LeBron-led Cavs will be flanked by several sharpshooters, giving the team ample long-range ammunition. — Andrew Flowers

CORRECTION (Oct. 28, 9:32 a.m.): An earlier version of this article originally implied that two teams had scored less efficiently than the Wizards in the 2012-13 season. Only two teams in the last five years have been less efficient than the 2012-13 Wizards.

CORRECTION (Oct. 28, 11:20 a.m.): An earlier version of this story misstated Charlotte’s 2013-14 record; it was 43-39, not 42-40.

CORRECTION (Oct. 28, 12:18 p.m.): An earlier version of this article incorrectly suggested that Marcin Gortat signed with the Washington Wizards before the 2013 season, when in fact he was traded in 2013 from the Phoenix Suns.

## Footnotes

1. The rosters we used came from ESPN.com’s depth charts, and were current as of Monday, Oct. 20.
2. Granted, it was a season shortened to 66 games by a lockout. But still.
3. Prorated to an 82-game schedule.
4. There’s a mild -0.21 correlation between the change in an NBA team’s record from one season to the next and the change the following season; the correlation for the NFL is much stronger at -0.41.
5. As measured by SportVU Player Tracking statistics.
6. This projection uses Real Plus-Minus ratings for each player, which are aggregated to forecast the team’s wins total.
7. The percentage of field goals a player assisted when he was on the floor.
8. Which, given what we know about his efficiency, is probably too many 3-point attempts.
9. A challenge is defined as a defender being within 5 feet of a shot taken from 5 feet or less away from the basket.
10. According to “retro predictions” based on the projection system devised by FiveThirtyEight editor-in-chief Nate Silver in conjunction with my Statistical Plus-Minus (SPM) metric. Applying the projection process to historical teams using SPM data that would have been available at the time (while still using the players’ actual minute totals from the season in question), we can roughly measure how good teams “should have been” for seasons dating back to 1979-80.
11. After pro-rating the lockout-shortened 2011-12 season to 82 games.
12. To project Rose’s 2014-15 season RPM, we took his RPM score for the 10 games he played last year and then applied an aging curve, which regresses his performance down about 15 percent.
13. SPM is created from RPM by running a regression with RPM data as the independent variables and various box score stats (rebounds, assists, etc.) as the dependent variables. RPM goes back only to the early 2000s, so SPM serves as an estimate for years for which RPM doesn’t exist. SPM covers players going back to 1977.
14. Andrei Kirilenko was a Plus-Minus beast!
15. The minute projections are from ESPN’s fantasy basketball tools. The 3-pointers per minute projections are from Basketball-Reference.com’s Simple Projection System.
16. Jones is listed as the third small forward on ESPN’s depth chart, and so not included in the top-10 rotation here. Jones also played limited minutes, and only 20 games total, on the Heat last year.

Neil Paine is a senior sportswriter for FiveThirtyEight.

Ian Levy writes about basketball from southern Vermont. He’s the editor of NylonCalculus.com and a regular contributor at Hardwood Paroxysm.

John Ezekowitz worked as an analytics consultant for the Phoenix Suns from 2011 to 2013 and previously was the co-president of the Harvard College Sports Analysis Collective. He currently works at Sankaty Advisors, the credit affiliate of Bain Capital.

Andrew Flowers writes about economics and sports for FiveThirtyEight.