July 26, 2014

Statistical observations following another non-playoffs Cavs season

NBA: Philadelphia 76ers at Cleveland Cavaliers

Remember how much fun the playoffs were? From 2005-2010, spring time basketball in Cleveland was fun and exciting. If you recall, there were chants of “Eric Snow for MVP” one year – perhaps my favorite sports memory ever – and that’s how in love we once were with playoff basketball.

If the past few days of amazing back-and-forth games showed me anything, it’s that as a fan, I yearn to be back in that position again soon. Very soon.

As a stats-minded individual, I also know that the Cavaliers haven’t had it that bad … yet. Missing the playoffs four straight seasons isn’t that rare in the NBA. Several teams are on longer droughts currently. Heck, imagine how difficult life would be out in the Western Conference.

Continuing with that stats-centric hat, here are six observations I’ve been mulling over from the conclusion of this funky #seasonofhuh. Before we get started, go check out the pieces from my WFNY colleagues Ben Cox and Andrew Schnitkey so far. Both were outstanding.

Maybe the expectations for this Cavaliers team were too damn high in the first place?

Last season, the Cavs were 24-58. They were the third-worst team in the NBA, fortunately again jumping up in the lottery to control the top pick in the draft.

At the end of the year, I wrote an article called “Regression to the mean and the Cavaliers.” In it, I shared the fate of the 20 previous NBA teams to go exactly 24-58. On average, they improved by 7.5 wins the next season. But that jump was heavily skewed by the 2007-08 Boston Celtics, who improved to a 66-win season with the addition of the Big Three.

The 2013-14 Cavs improved by an above-average nine wins. Yes, of course, they had a major offseason spending spree and expected to improve significantly. But the majority of those new players, including the rookies, failed to make an impact – or actually made a negative net impact.

Jarrett Jack played 2,252 mostly underwhelming minutes, third-most on the team, at a cost of $6.3 million. His presence perhaps led to increased confusion with the roles for young assets Dion Waiters and Kyrie Irving. Mid-season acquisitions Luol Deng and Spencer Hawes combined for 2,157 more minutes too. They were mostly helpful. Much more on them later.

The Cavs had an average age of 24.9 this season, seventh-youngest in the NBA. No team younger than the Cavs had a better record. Last year, the team’s average age was 24.1, second-youngest in the league. Largely, this was the same team and the same main contributors. As a sum, they got better.

Before the season, we can’t forget that some were skeptical about the Cavs’ supposed no-doubt improvement. Sports Illustrated’s Ben Golliver predicted the Cavs as the biggest potential “disappointment” in comparison to their 40-win Vegas odds. Grantland’s Zach Lowe infamously wrote “no one has any clue with these guys.”

If fans want to be mad about this ongoing rebuilding process, they should be mostly mad about the missteps leading to last year’s 24-win performance. Then, they should be mad about the failed free agent signings of summer 2013. The young contributors are mostly getting better and a nine-win improvement is nothing to be that angry about.

Yes, the defense did improve pretty significantly this season.

Where did the returning contributors improve? One easy place to start is defense, where Byron Scott’s teams ranked 27th (106.9), 26th (106.0) and 29th (109.1) in defensive efficiency over the past three seasons. They were continuously abysmal on that end of the court.

Mike Brown entered 2013-14 with a stellar defensive reputation. During his last 11 full seasons in the NBA, no Brown-influenced team ranked worse than 14th in defensive efficiency. He also hadn’t missed the playoffs in any of those seasons.

The Cavs finished this season with a 104.8 defensive efficiency, tied for 17th in the league. That is a large upgrade. It cannot be ignored. WFNY’s Scott Sargent wrote about some of the other defensive numbers last week. Undoubtedly, there was a clear improvement.

One thing to note within the context of defensive numbers: It might seem that the Cavs regressed toward the end of the season upon the addition of Deng (on-court DRtg of 106.4). But that’s not exactly true compared to the league average.

Drastically, NBA offensive efficiency shot up in the second half of the season. Before Deng’s debut, the league average was 102.6 and after, it was 105.1. That’s how the team was able to go 17-16 down the stretch, despite the apparent sudden drop in defensive efficiency.

I also think that we all might be over-emphasizing the need for a rim protector, at least somewhat. The Cavs ranked 24th this season in opponent field goal percentage in the restricted area. Yes, that’s fairly on par with the 23rd, 26th and 28th rankings in the Byron Scott era. That’s not ideal.

Tristan Thompson doesn’t have the size to contest well inside (but that doesn’t mean he’s a worthless basketball player). Spencer Hawes isn’t that defensively skilled, nor has he ever been. Tyler Zeller is OK, but is best served as a backup. The Cavs do indeed lack legitimate interior defensive talent.

However, shooting efficiency is only one part of the equation. Teams attempted the third-fewest percentage of shots in the restricted area against the Cavs this season. That might be surprising to the average fan, one who continues to complain about the lack of a “true center” on this roster.

To me, that surprising stat screams that opponents aren’t exploiting the team’s interior defensive issues, but are merely capitalizing occasionally. It means that there are other Cleveland defensive weaknesses that are much more glaring. More on that topic in a bit.

[Related: Playing the Odds: Everything we know about the Cavaliers and analytics]

Assists and interior scoring ramped up dramatically with the addition of Luol Deng.

On the offensive end, this season was largely disappointing. Last year witnessed a 38-game stretch where the team had the NBA’s seventh-best offensive efficiency, as led by departed bench heroes Luke Walton, Shaun Livingston and Wayne Ellington. No such magic occurred this year with the Cavaliers scoring the basketball.

Where to begin? Oh, the Cavs shot only 50.9% in the restricted area in the season’s first 35 games. That’s not good … at all. In fact, it was the worst such mark since 1997-98, according to NBA.com/stats. No team since then had completed a regular season shooting worse than 53% in the inner circle. The Cavs seemed set to make the record books.

At least, that was the pace until Deng’s debut. Deng arrived in early January with a decent reputation not as an outside shooter, but as a slasher, passer and defender. He was shooting 66% in the restricted area with the Bulls this season and 62% in the previous three seasons.

And although Deng’s play with Cleveland was a bit worse than many hoped, there’s no doubt that the Cavaliers offense changed for the better with him (and Spencer Hawes) in tow. Two areas where the stats particularly stand out: Restricted area shooting percentage and assisted field goal rates.

cavs shooting before-after deng

For the season’s final 47 games, the Cavs were nearly average at 59.7% shooting in the restricted area. That’s a far, far cry from their historically inept start to the year. Andrew Bynum – champion of Opening Night hope – was no longer around. Tristan Thompson mostly returned to normal. The guards all jumped from terrible to average.

And look at the assist rate improvement. Before Deng’s debut, the Cavs ranked 28th in the NBA with assists on 45.8% of their two-point field goal makes. After Deng’s debut, they ranked ninth at 55.2%.

That’s not to say assisted field goal rates necessarily mean that much. This stat actually has a very low correlation with offensive efficiency; the Bulls ranked first and the Suns ranked last in the stat. But the increased assists are likely conflated with the interior scoring improvement and fit the team’s second-half narrative of floor spacing, better ball movement, more trust and easier shots.

I’d argue that Luol Deng’s footprint was all over these offensive improvements and he didn’t bomb his free agent value that much, as some have theorized. With Deng (and Hawes), the Cavs had an offensive identity and rhythm down the stretch. It was good.

The Cavs still need a lot of work on the three-point line, offensively and defensively.

The NBA game is changing. During the days of the Bad Boy Pistons in the late ‘90s, teams were attempting three-pointers on only about 7-8% of field goal attempts. By now, you know that the three-point shot is increasing in frequency and importance.

In 2013-14, NBA teams attempted threes on a record 25.9% of field goal attempts. That number keeps jumping up each season. Teams are becoming more and more strategic on how to attempt (and defend) the three-point ball. The Cavs, meanwhile, seem stuck in a style of game from a few years back, at least.

 

cavs three differential

The three red lines in the chart – which was inspired by Ben Golliver – represent Cleveland over the past three years. Unfortunately, yes, they are in order. In 2011-12, the Cavs had a differential of -0.12 three-pointers per game. The next season that mark fell to -0.77. This year, it dropped way down to -2.21.

That means that in 2013-14, the Cleveland Cavaliers were out-scored by 6.63 points per game from three-point field goals alone. They were out-scored by only 3.29 points per game, meaning they made up significant ground on two-pointers and free throws. But this shows how teams might’ve been abusing the Cavs on the wings, not in the paint.

As I wrote in early February in an article on defensive analytics, the Cavs allowed more and more three-point attempts as the season rolled on. They finished the year allowing 30.4% of shots to be threes, highest in the NBA. That’s a strategy failure. Teams like Portland also lacked one-on-one defensive talent, but purposefully schemed to take away three-point attempts. The Cavs seemed to over-blitz defensively, leading to constant open shots behind the arch.

On the offensive end, as I wrote in The Diff in early February using the ShotScore statistic, the Cavs seemed to take among the worst shots in the NBA. They finished the year taking 46.7% of shots as non-restricted area two-pointers, the fourth-highest such rate in the league. Those shots carry with them the lowest possible expected value.

The Cavs haven’t substantially invested in three-point shooters over the past several seasons. Luol Deng was never more than a mediocre shooter. The team’s three-point attempt rates were actually on par with the NBA average with either Spencer Hawes or CJ Miles on the court. It’s amazing how much better the team played with real, live shooters.

The roster desperately needs more three-point prolific wings in order to catch up with the rest of the league. And defensively, Mike Brown (or whoever the coach might be) needs to wake up and realize a better system for how to discourage three-point attempts s from happening so frequently.

Advanced stat loves Luol Deng, Anderson Varejao, Matthew Dellavedova and C.J. Miles.

In case you’re a hard-core basketball fan living under a rock – which likely isn’t possible – by now you’ve heard of Real Plus-Minus, ESPN’s new foray into the world of basketball analytics. As an overview, read the mostly vague introduction and play with ESPN’s sortable stats. Then, check out the phenomenal follow-up explainers from Kevin Ferrigan and Alex Suchman.

Real Plus-Minus is an adjustment to APM, RAPM and xRAPM, three statistics that have been living in the NBA’s underground analytics lairs for years. It attempts to regularize, normalize and fix-ize all of the many caveats that arise from actual raw plus-minus data in the game of basketball.

In a nutshell, the basics of Real Plus-Minus: It adjusts for teammates and opponents, it includes priors (stats from previous seasons), it still relies somewhat on box score statistics like rebounds and points, it factors in player height, 0.00 is set as the league average and there is some magical proprietary fairy dust sprinkled in too. To calculate WAR, everyone’s new favorite baseball statistic, you just need to know that the replacement level is set as -2.35.

What does RPM say about the Cavs? Well, take a look at the chart below. The DIFF O and DIFF D columns represent the net differential between the player’s ORPM or DRPM compared to the averages at their respective position.

cavs and real plus-minus

The only four players with above-average marks in both marks per their respective position: Deng, Varejao, Dellavedova and Miles. Let’s take a quick in-depth look at each of these four players.

For Deng, it’s important to note that this includes his season stats with Chicago as well. Again, I think it’s vital that we not outweigh three (quietly productive) months in Cleveland over 9.5 (very good) years in Chicago. His all-around game will still get paid significantly this offseason. By now, I’m assuming that won’t be by the Cavaliers.

Varejao still had defensive value, perhaps aided by his past years of success, but it’s shocking to see the offense. He had the third-best ORPM among centers (as of stats pulled on Saturday). His mid-range shooting, which I once criticized, was a huge asset for the Cavs. It’ll be fascinating to see if he has now played his final game in the Wine & Gold.

GotBuckets’ Kevin Hetrick is lobbying for Dellavedova as the Rookie of the Year based on RAPM stats. Delly was a constant catalyst for improved on-court play for the Cavs, leading the putrid 2013-14 rookie class in WAR. He’s a near-lock to return as he has a unguaranteed $816K salary next season. There are many worse backup point guards in the NBA.

And speaking of unguaranteed deals, the Cavs had another bargain year from Miles, who likely will seek greener pastures in free agency too. He was an underappreciated hero for the team this season, providing quality minutes (and shooting) on the wing. This likely was the best free agent signing of Chris Grant’s tenure. That’s pretty sad.

I’m not exactly certain what to think about Kyrie Irving just yet … and that’s OK.

Kyrie Irving is the 22-year-old reigning All-Star MVP. He has accomplished seasons that very few guards his age have ever done before in NBA history. And he’s also an enigma.

Is he regressing? Is he ever going to be a “superstar,” for whatever that tag is worth these days? Is he necessarily better than some of his peers, like Damian Lillard and John Wall, who are shining bright in the playoffs right now? Is he actually worth a max contract in the NBA system?

Mostly yes. Probably not. I don’t know. And most likely yes. Those are my four answers. Yes, I think Irving is overrated – I said that before the season too when #NBARank pegged him as the league’s eighth-best player. That’s just insane.

Looking at the numbers, there’s no doubt Irving’s productivity has changed drastically since his electric rookie season. In fact, one can pinpoint a specific, albeit mostly arbitrary, turning point: the 2013 All-Star Game. Take a look at the before and after stats.

cavs kyrie irving before-after 2013 asg

The shooting efficiency drop is the first notable line, especially from three-pointers. He got off to such a hot start for his career, but how much better could he get than right around 40%? Not many dominant scorers excel much beyond that. Regression was likely, at some point, but he really, really struggled through November of this past season. He shot 37.4% from distance from December-onward.

The usage statistic is also interesting. It dropped precipitously from January-onward this season, dipping all the way to 26.5% in his final 40 games. At the same time, his assist rates improved and turnover rates dropped. He was working off the ball more, ceding point guard duties to Jarrett Jack.

So what does this all mean going forward? ESPN Insider’s Amin Elhassan did a good job of breaking down all of the possible scenarios. Cavs: The Blog’s Nate Smith is one of many who remains in favor of trading Irving for assets.

I’m of the mind that the Cavs should keep hold to this transcendent talent. He still could improve, undoubtedly. The narrative that “he’s not a winner” doesn’t sit well with me. If the Cavs had enough complementary talent, that wouldn’t be an issue. The team controls the situation, undoubtedly, so it’s practically unlikely he’ll be playing for any other team for many, many years.

The Cavs are then somewhat stuck in a Catch-22 situation. Management wants to improve immediately in order to build a better environment around Irving and contend for next year’s playoffs, actually. But the guaranteed long-term core of the roster (Thompson, Waiters, Bennett) remains very young and obviously filled with currently untapped potential. The team wants to win, but likely needs in-their-prime veterans to win. That means perhaps gambling away the other young talents. And, of course, we still have no idea who might end up making such decisions.

This ragtag roster improved from 24 wins to 33 wins in 2013-14, perhaps despite Irving’s stagnation and all of the other miscues. Now, anything and everything appears to be on the table for how this organization will look in just six short months when the 2014-15 season begins. It’s sort of a shame, but it also might not be a bad thing at all. I have no idea what to expect next.

David Richard-USA TODAY Sports

  • boomhauertjs

    I’m officially on board the Doug McDermott bandwagon for 3-pt shooting help.

  • The_Real_Shamrock

    Start Dellavadova, Miles, Deng and Varejao next season because advanced statistics is where it’s at baby!

  • Steve

    Remember that our expectations at the beginning of the year did not include Bynum playing from Opening Night, and Deng and Hawes to be added. That nine win improvement is nice, but expectations should have been for more, especially as the East got even weaker this season.

    And everyone knows I’m one of the stat-guys, but we need to mix in some scoutiness here, which you only briefly touched upon. The Cavs were able to stop teams in the paint because they were absolutely selling out to defend it, which is what caused all those threes to be attempted. I don’t think the Cavs defense would have been any better if they stayed out on shooters. They would have been rolled over in the paint.

    The collapse-and-rotate Brown defensive scheme may need some adjustment, but it’s definitely going to need smarter, better defenders. And that rim-protector is certainly one of those better defenders. It will allow guys to stay out on shooters.

    Regarding Portland, they went from Hickson to Lopez, whose 1.7 blocks/night finished 8th in the league, in the middle. Between Lopez and Aldridge, they had plenty of length to contest guys driving in the paint. That wasn’t just scheme.

  • Steve

    Or, if you’re going to keep Irving as the focal point of the roster, put him on the court with guys who can hit jumpers and/or show some serious effort on the defensive end.

  • Steve

    That could help the league average 3PM/night, but it would certainly exacerbate the bigger problem, the league highest 3PM/night allowed.

  • The_Real_Shamrock

    As far as Portland goes love what they did with the big men. Aldridge is the offensive big man so they brought in a defensive/rebounder/shot blocker in Lopez to pair with him. Lets face it the paint has been a problem for the Cavaliers for a very long time. I really think this teams a low post presence more then anything. The trouble for me is fit. With Tristan Thompson on the roster I feel they need an offensive center I would have that Hawes was the perfect fit because he could space the floor for the guards while allowing Thompson to rebound. But I didn’t see much of that happen. Also it’s been reported Hawes will be looking westward to play next season.

  • The_Real_Shamrock

    I was using my own reverse psychology no way do I want those four starting. Besides I think Deng won’t resign and I think the time has come to bid Varejao adieu. Or if they do bring him back change the personnel so that he can be saved physically and return to the bench as a super sub or sixth man. Bring him and Waiters off the bench to lead the second unit.

    I am keeping Kyrie and building around him starting hopefully with a big man down low.

  • mgbode

    just trade those 3 for Kevin Love!!! (oh sorry, NBA trade silly season doesn’t start until after the NFL draft)

  • mgbode

    and, I don’t think it would help all that much. he gets wide-open looks in college that will not be there. I think he’s going to struggle mightily on his transition (both getting his shot & obviously defense).

    if he was good enough to play defense and sit in the corner, maybe. but, he’s not.

  • mgbode

    Random points/thoughts while reading.

    *Would love to see average age as a component of minutes played. Easy to have youngsters at the end of the bench, harder to have them on the floor more often.

    *I had nearly forgotten about the ill-fated Earl Clark signing. Sigh.

    9 win improvement is nothing to be angry about

    *not angry, but disappointed. some of those wins were expected in Irving being healthy more and getting better (I’d argue the most important component of the expectations). well, he was healthy more, but he had a step-back year. I am positive about him long-term, but it was a rough year.

    the team was able to go 17-16 down the stretch

    *important stat. and, why it would be nice to get Deng & Hawes to stay (both contributed towards it as you note, but it doesn’t seem likely (especially for Luol). then add Andy and CJ possibly leaving too and it’s easy to get depressed. those are 4 of the biggest positive play guys from this year.

    *obviously agree on the need for 3pt shooters.

  • WFNYJacob

    The average age statistic, which is from Basketball-Reference, does account for minutes played.

  • BenRM

    I can’t remember exactly, but I think I “predicted” wins in the upper 30′s to very low 40′s…something like between 37-41 wins.

    I probably drank too much Kool-Aid and didn’t realize exactly HOW terrible the 2013 Draft Class was….I mean I knew it was bad…but holy cow.

    I also really expected Kyrie and Dion to be a “thing.” I also expected that “thing” to not be a complete dumpster fire.

  • The_Real_Shamrock
  • The_Real_Shamrock
  • Dave

    Delly isn’t quite as good as Irving stats-wise – Irving passes better (A/To) and is a somewhat better defender (really, look at Steals, Fouls, and Blocks per 48!).

    But there’s a very good argument that you’d rather have Delly in there running the point than Dion Waiters: Delly shoots more accurately (TS%, eFG%) and passes more accurately (A/To) than Waiters.

    And there’s consensus among the stats-heads that Anderson Varejao is in the running for most underrated player in the last decade. The reason he’s so underrated is that scoring isn’t the focus of his game, rebounding and passing and defending is.

  • The_Real_Shamrock

    Varejao also never plays a full season too.

  • mgbode

    fantastic. nevermind, the stat is great : )

  • TV63

    This was put together very well. Thank you. You have support from http://www.nba.com/powerrankings/ as early as week 19 on your evidence that Deng had given the Cavs an offensive boost. It was sorely needed.

    “Cleveland
    (24-40) Week 19
    Pace: 95.5 (20), OffRtg: 99.9 (24), DefRtg: 105.0 (20), NetRtg:
    -5.1 (25)
    The Cavs’ playoff chances are likely gone after back-to-back losses to
    Charlotte and New York, because they now play six straight games against teams with top-10 records. Luol Deng has given them an offensive boost, but their defense has been much worse in the 29 games since the trade (allowing 108.0 points per 100 possessions) than it was in the 35 games before he arrived (102.5). This week: @ PHX, @ GSW, @ LAC “

    There is a discrepancy however on the notion that it appeared Cav’s DRtg regressed at the end of the season with Deng. Stats quoted from nba.com during week 19 show the pattern was much earlier than that. Drtg stats show a decline from his arrival from 1/10/2014 well past his first 29 games. The stat of 108.0 per 100 possessions had not changed a whole lot by season end. If
    you took away Deng’s absence of 3/20,3/22, 4/11,4/12,4/16 the figure would be
    106.6 per 100 possessions. (still a far cry from 102.5 prior to his arrival)

    You did state the Real +/- chart included Deng’s stats from Chicago. Those numbers are huge compared to the +/- values he actually had with the Cavs. His Best 5 man line-up was +13.9 under the Bulls, Best 2 man with Boozer was +6.3. This is a stark contrast to Deng’s Best 5 man Deng, Jack, TT, Zeller ,
    Kyrie +7, Best 2 man with Andy @ +1.4.

    Though you pointed out Kyrie’s shooting % drop; so did Deng’s remarkably so. FG%41%, 3P%31.5%, FT% 77%, TS% 50% EFG%45%. It is curious as to why specifically Kyrie, Deng and Jack shooting% dropped down like this while Hawes and the bench players of Dion, Delly, Tyler, Andy improved quite a bit. I can’t help to wonder if the inconsistent starting rotations with injuries just wreaked havoc more than we could know was a factor. Andy/Miles started until 2/17, injuries of Andy, Cj , Waiters soon started to mount, Jack starts 2/19, Hawes starts 2/25. If you do some digging on +/- player line-ups; you can see there were line-ups that were just toxic. The worst 5 man was Hawes, Gee, Jack, TT, Dion. You can assort many different line-ups and you will come away realizing Hawes/Tristan combos in a 5 man, 3 man or just the 2 of them are horrid. How much did this impact the team? Data taken from http://www.basketball-reference.com/players/d/

  • Dave

    He doesn’t, but when he’s healthy he’s so effective that the advanced stats have him contributing more to the Cavs’ wins this year than any other player. Mostly it’s his incredible ability to follow and snag rebounds.

    A player who is also often heavily underrated for similar reasons is Ben Wallace back when he was with Detroit.