Wednesday, January 29, 2014

The 2014 Eastern Conference All-star Game Picks

One idea floating around about the all-star game that has picked up steam is to do away with teams by conferences, pick the 24 "best" players, and let two captains pick sides. During a historically bad year for the east, this has gained traction, and I'd be fine with it since it's always a glorified pick-up game. I say this because I had to seriously consider the candidacy of guys like Afflalo and Lance Stephenson. Meanwhile, guards like Tony Parker and Mike Conley could miss the all-star game. Nevertheless, I had to attempt to scrabble together a competition eastern conference team.

Speaking of Lance Stephenson, a campaign has spread for his inclusion in the all-star game, and I'm not sure what people think they're accomplishing. First of all, insisting that Indiana should have three all-stars is weird and misguided; an elite team will be awarded by winning (by itself it's a reward) and by playoff success. And Indiana's strength is its well-balanced starting five, not having three stars. In fact, you can have a 60+ win team with only one all-star or, theoretically, zero. Think about this: a great team will have a point differential of, say, +7. The top players in the league will be around +4 to +6 overall just by themselves (but not this is if they played an entire game of 100 possessions.) A player that's a +2 is likely a good role player or a fringe all-star.

So let's say our fictional team has one great star who's +6, and he's surrounded by four solid starters who average +1.5. That's 6+4*1.5, equaling +12 overall, if the starters played the entire game. Bench players, especially the guys who ride the end of the pine, are typically "negatives," but if the bench is decent enough, like Indiana's this year, that +12 overall won't be depressed by a huge amount. Last season, Indiana's starters played 64% of their total regular season minutes, so let's pretend this team gets 60% of their minutes from their starters and 40% from a bench that averages a -1 impact (a poor player, but nothing horrific.) That means you have a weighted average for a point differential of +6.8, which would make them one of the best teams in the league. But we gave them only one star. It's evident that there's no perfect relationship between team strength and number of all-stars. As John Hollinger said, the all-star game is not about parliamentary representation; just pick the best players!

(By the way, that works even with other metrics like basketball-reference's Win Shares. The best non all-stars average like 5 to 8 Win Shares a season, and even Indiana's bad bench last season had 14.2 Win Shares. That means you could build a non all-star team with a poor bench that could win, say, 50 games, and if you add one superstar you can close in on 60 games, which is a reasonable facsimile of the LeBron-era Cavaliers.)

And realize we're constructing this edifice for a guard with a usage rate under 20 for an offense on a slightly below average offense. Lance is lauded for his "playmaking," but the results haven't been extraordinary, he turns the ball over too much, and he definitely has useful offensive players around him. Also, note that usage rate includes turnovers, so Lance is an even smaller part of the offense than it looks. As Kevin Pelton noted, the only two guards who made the all-star team with a usage rate below 20 and less than 7 assists a game since the turn of the millennium have been legendary shooter Ray Allen and Andre Iguodala, who's both a better playmaker and defender than Stephenson. If you go back to the 90's you'll see a couple more super-role players like Dumars, but you'll also see some of the worst all-star selections ever like BJ Armstrong. Basically, unless you're a top-tier distributor, defender, or shooter (Majerle and Ainge made the cut), you're unlikely to make the team. (If you go back further you'll find more guards, but as the league got smaller it meant there were more all-star spots for each player.)

If people want Lance to make the team because of his defense, I'm not seeing it anywhere, at least without a great deal of certainty. Indiana's defense is scary, but Hibbert and George are the best defensive players. Stephenson has cut down his mistakes, but I don't think he's good enough as a perimeter defender that he's a lock. Yes, the east is historically weak, but what are we doing with this campaign? All we're essentially doing is making Stephenson more expensive for his team to sign him after his rookie contract expires, and in their small market it might make keeping him impossible. So this strange obsession people have with proclaiming him an all-star will just have the effect of dismantling the dominating starting lineup of one of only two great teams in the eastern conference and the only bump on the road to the finals the Heat will face.

And, finally, here are my picks for the all-star team....


Guard: John Wall

There are a few impassioned debates about who's the best point guard in the east right now, driven by the fanbases of Cleveland and Washington. Wall's case is more traditional than one would expect, given his errant jump shot and his flashy dunks -- he controls Washington's offense with his passing, and their offense falls apart without him. He's fourth right now in assists per game, or fifth if you include Kendall Marshall (i.e. the D'Antoni surprising, resurgent point guard du jour.) He's also fourth in "hockey assists" per game and moves up to third in the league in assist opportunities per game, per SportVu, basically implying that Wall's teammates are deflating his assist totals. Washington, meanwhile, has an offensive rating of 106.9 when Wall is on the court, and that plummets to an unfathomable 93.8 without him. This is the intersection of two strains of data types in this information world, and without juxtaposition they could be meaningless. But together it shows the large amount of control Wall wields -- he's, in fact, leading the league in time of possession per game, again from SportVU. However, since he plays heavy minutes and hasn't missed a game, his raw on/off differences are likely to be dramatic because a large portion of those off minutes consist of garbage minutes without any starters or only one. More nuanced metrics, controlling for teammates and the strength of the opponent, are more modest with their appraisal of Wall, but this is a historically dreadful east and there are few guards even approaching "star" status. For the last stat explaining how Wall sets up his teammates, Bradley Beal's TS% without John on the court is, via NBAWOWY, 43.6.

Guard: Dwyane Wade

I'm not entirely comfortable putting Wade in the starting lineup since he's missed a large number of games, but based on pure quality he's still the best shooting guard in the east (dubious award though) and, hey, Wall-Wade-George-LeBron looks like a really fun quintet. The rumors of Wade's decline have been exaggerated. Wade's motto is: always play hard. This will impress some people, and astonish the rest. But after years of falling down hard (or embellishing contact), and after years of injuries, he's no longer a sure thing on the court. Yet he's still generating a lot of shots on very good efficiency while passing well and finding players at the rim. He's a terror in transition, and he gets a pass for some missed games here because he's made three straight finals. They're keeping him fresh for a reason. I could see an argument for Lowry over Wade in the starting lineup, sure, but I prefer not to have two point guards starting, and Wade has the requisite box score stats and the "impact:" an RAPM version, free of box score stats, has Wade in the top 20 at +3.3.

Forward: Paul George

The reigning Most Improved Player, he's once again a viable candidate. In fact, last season was an odd case: he locked up the award playing for a good team and increasing his minutes by nearly 8 per game. His largest improvement came from his perimeter defense, so in a backdoor way he did deserve the award, but the media typically neglects defense unless it's propped up by gaudy block/rebound stats. He's maintained his status as an elite perimeter defender, but he's surged to stardom thanks to a legitimate high-scoring game. His shooting percentages have risen on virtually every area of the court, depending on how you slice it up, even though he's taking more shots, including difficult fadeaway jumpers. There should be absolutely no qualms about George being an elite player now. He's scoring efficiently on a fairly high volume without much help; it's not like he's being set up by one of the best playmakers in the league. And coupled with his defense, and he's a two-way force. Indiana's offense is overrated because their win total hides the fact their offensive efficiency is below average even with outstanding offensive rebounding, but that's through no fault of George and he's a key part in a defense that's putting everyone else to shame. The Bulls' defense is closer to the 9th rated Clippers than the Pacers.

Forward: LeBron James

I don't need to spend much time explaining why LeBron should be in an all-star game, even when he's "LeCoast." For one fun fact, with a usage of 29.9 via b-ref and a TS% of 65.9, he's literally off the charts in terms of usage and efficiency. (And actually, so is Durant.) Even though he's dialing back the effort on defense and his turnovers are up, he's breaking the laws of basketball -- he's second in the entire league in post-up efficiency at 1.17 points per game from Synergy, and it's not on a trivial amount of plays either; he's posted-up 156 times. He's also tailored himself as an excellent spot-up shooter, as SportVU ranks his effective field-goal percentage third out of 195 players with at least 20 games and 2 catch-and-shoot attempts per game. He's shooting 51% on those three-pointers. Once derided for his lack of finesse skills, his versatile skillset is a nightmare in a variety of situations.

Forward: Roy Hibbert

I should take more credit for defending Mount Verticality in the first half of the 2013 season, explaining that his shooting percentages were likely an aberration due to an injury and that his other stats held steady. Then in the playoffs the Indiana defense lurched into high gear against the Miami Heat, taking them to seven games. They have nearly perfected their own brand of defense and are currently 9.5 points better than the league average on defensive efficiency per 100 possessions. For people unfamiliar with these stats, that is unprecedented in modern basketball. The Ewing-Oakley-Riley Knicks peaked at 8.3 in 1993. The Duncan-Popovich Spurs peaked at 8.8 in 2004. The Garnett-Perkins Celtics peaked at 8.6. Outside of Bill Russell's Celtics, this could be the most impressive defensive season we've ever seen, relative to one's peers. The historically weak east takes a little oomph out of their stats, but this is still mighty impressive, especially since this is an age with an unprecedented amount of information and a wave of people trying to emulate the defensive success of guys like Thibodeau and Popovich. His box score stats may seem modest, but it's the tip of the iceberg -- the defensive stats of the Pacers are gargantuan, and Hibbert is their most important defender, guarding the restricted area. His defensive rebounding looks anemic, but Indiana leads the league in DRB% and their perimeter players get in on the action. He's also has a decent utility on offense for a center, more usable than types who can't create their own shots. If you were to take one center for the game, Hibbert is the choice if you want to win.


Guard: Kyle Lowry

The point guard competition in the east is seriously weak, as one could argue probably two players from the west who won't be picked over any point guard here. With Rose injured again, Irving failing to take a leap forward, and even Rondo limited, Lowry has unexpectedly put his hat in the ring. Offensively, he's taken on a healthy share of his team's offense with his shot creation and assists, and his shooting percentages have been excellent. He's taken care of the ball too with a 3.4 assist/turnover ratio. As a result, the Raptors have an offensive rating higher than Indiana, even with Rudy Gay's epic shooting problems -- if Gay had shot at the league average, he would have had 50 more points over only 18 games. In the blight that is the eastern conference, Toronto has been one of the few bright spots and Lowry has played a central role. Not only is he good offensively, but when properly competing he's one of the best defensive guards too. He's averaging 1.6 steals and 1.2 charges taken a game, which is the highest mark since 2007 from Varejao and Andrew Bogut, via Hoopdata. He's a bulldog defensive point guard who's a plus when he's motivated and not excessively gambling. He's definitely the best guard on the Raptors; let's hope everyone won't be distracted by DeRozan averaging 33% more points ... because he's taking 50% more shots. DeRozan is an "empty calories" scorer who offers nothing else of value and doesn't even score proficiently. Somehow Lowry's on the trading block, according to some people, but their reigning GM of the year will be smart enough to keep him.

Guard: Kyrie Irving

Like Wall, even though his team's offense is disappointing, it's better when he's on the court and better is an understatement: they score at a pathetic 96.8 points per 100 possessions without Irving, compared to 103.4 when he's on the court. Having a real NBA small forward helps, as they've been better with Deng on board. However, I'm still not sold on him as a player with his weak defense, and I think his crossovers and his dribbling skills don't lead a team to results that are as good as one would think. Unfortunately, there is no viable candidate to replace him, and I'd rather bet on a player who's probably just had a weak slump at the start of the season to a role player who doesn't belong in the game or a fluke season. For instance, Afflalo has gotten a lot of attention for the all-star game, but I feel like we're going to look back years from now and view him like we do Mo Williams. Last season in a similar role his shooting efficiency was poor with how large a chunk of the offense he shared. His TS% has risen 6.2 points.

Looking at his playing style, Afflalo takes a heavy amount of baseline jumpers, especially turnaround shots, usually contested. He's shooting over 50% in a couple of these midrange zones, and my instinct screams "fluky shooting percentages." He was a pretty decent midrange shooter last season, however, and the other reason his overall numbers are up is that his three-point shooting has bounced back from 30% last year (he's near 40% for his career) to 42%. Given his high minutes, he's averaging nearly 20 points per game, but beyond that he's not contributing at all. His overrated defense is now fairly porous given all the energy he expends on the other end, and he provides nothing else of significant value including playmaking. As for Lance Stephenson, I've said my share, and another serious candidate thrown around is ... DeMar DeRozan, a one-dimensional scorer whose only fame to claim is shooting a lot. Otherwise you have to consider guys like Deron Williams, who makes Kyrie Irving look like an MVP, and Joe Johnson, who's already been awarded far too many all-star games after that pact with Satan. So Kyrie Irving gets the selection here, but if someone wants to perform a Freaky Friday with him and Goran Dragic I wouldn't mind.

Forward: Carmelo Anthony

If there were more competition, I could have a nuanced discussion on Carmelo, how he's used in an offense, how we shouldn't judge him based on the disaster surrounding, his aloof defense, etc. If you're going to keep him out of the game because the Knicks are losing most games and you think he's a chucker, then you have to consider guys like Luol Deng or Joe Johnson. We all know his faults, but that means he's not an MVP candidate. A TS% of 55.3 isn't special, but it's not bad either; it's above average. He does this without turning the ball over, and his rebounding is underrated: a 21 DRB% is good enough for a center (although starting next to Bargnani helps a player's rebounding numbers, Carmelo has always been great.) I don't want him starting because of his poor defense and how he can stall an offense -- he averages 17.4 frontcourt touches for every assist -- but he's a lock here.

Forward: Chris Bosh

Bosh is a somewhat forgotten piece now, more known for his goofy antics and video-bombing, but he's still an excellent player on both ends of the court. He's only playing 31 minutes a game, but it's through no fault of his own: the Heat have gone the distance in the playoffs three years in a row and they're resting key guys to be fresh for another title run. Per 36 minutes, he's average 19.5 points with superb shooting percentages and a tiny turnover rate for a big man, stretching the floor for LeBron and Wade. His rebounding and post defense against behemoths is disappointing, but his DRB% is good for a power forward, his natural position, and he takes a pounding from enforcing Miami's strategy of smallball and space/pace. Miami's defense has slipped, but it's not Bosh's fault: when he's on the court their defensive efficiency is 103.0, which would be sixth in the league (currently 11th overall) even though he often plays in smallball lineups. In fact, Bosh has the largest raw +/- on the team, and adjusted RAPM method (non-prior informed) estimates his impact at +4.2, one of the best in the league. His scoring stats would be more impressive if he had a larger role on offense, which is a luxury some of the other guys in the east have, but his accurate midrange shooting is a boon for Miami's offense and he's one of the best pick and roll defenders in the league.

Forward: Paul Millsap

Millsap would have likely been ignored for the all-star game if Horford didn't get injured while Atlanta kept winning. He's long been one of the league's most underrated players without a single flashy strength or style. However, he did add one asset rendering him unique -- he's a good three-point shooter now, taking 2.6 per game at 37%. He's also been a larger part of the offense than he's ever, and he's performing well, including a surprising amount of assists. So he's an accurate big man who can stretch the floor, hit a jumper anywhere, he's not a ball-stopper, he can pass, he rebounds well, he's a good defender, he picks up a lot of steals -- he's just a well-rounded power forward. And he's not leaning on anyone here either. The attention is now on him without Horford and without a Joe Johnson-type on the perimeter.

And while I'm talking about the candidacy of a frontcourt player, let me slam have an opportunity to slam Al Jefferson -- I've seen him on people's ballots over guys like Chris Bosh or Millsap, and I have no idea why. They're averaging a similar amount of points, but Al needs over three more field goals per game to reach his totals. Jefferson's percentages are significantly below average and he's a black hole on offense. The Jazz were always better off with Millsap over Jefferson. People complain about people who overuse things like TS%, but Millsap gives you more points in less shot attempts -- why is that bad? And he's doing it more in the flow of an offense, even shooting from further away and stretching the floor. I'll commend Al's improved defense and that of Charlotte's, but guys like Bosh are still better defenders. You can see this with the "eye-test," you can see this with stats -- it should be obvious.

Wild cards:

Forward: Joakim Noah

Once again, the Bulls are without Rose, and once again they're near league average with an excellent defense and an awful offense. There was no Plan B with Rose -- he was going to be their only scorer, besides Boozer and ... Mike Dunleavy. The team has no business hovering around .500, but they get there with a stifling defense and more ball movement than one would think for an inept offense. Noah represents their strengths, and years later I hope people realize he's a multiple-time all-star because of his defense, rebounding, and passing. His offense is better than most centers; he's not a negative here. A team can play through him in the high post, he has a usable jump shot, and he's unselfish. But as a highly active center, he's the key pivot point of the second best defense in the league. The game today involves lots of switching and pick and rolls, and that's Noah's defensive strength. He's allowed 0.65 points per play in isolation via Synergy ranking 26th, which has involved a few perimeter players, and only 0.74 points per play as the pick and roll defender, ranking an impressive ninth.

I understand that people will argue defense doesn't matter for the all-star game, but the way people talk about all-star appearances, and what it means to the fans, the history of the game, and the players, I'm going for best player. If he were in the west, I could see arguments for Duncan or Dragic over him -- for one, Chicago's defense is better when Noah doesn't play, but I don't think that's accurate -- but I'm more certain taking Noah here than any of the other candidates, like the injured Chandler or Al Jefferson.

Forward: Andre Drummond

This is basically me saying "screw it" and picking someone I'd want to see because I don't think there's a 12th candidate who's deserving anyway. David West is a valuable piece for Indiana, but his scoring stats have slipped: he's not getting to the rim like he used to; it's not fluky jump shooting percentages. Afflalo certainly has a case, but I'm not sure I trust his offensive game right now, which might be unfair but he's really not overwhelming. Jeff Teague hasn't been shooting well, and besides his passing, well, it's Jeff Teague. I'd really like to sneak Dragic across the Mississippi, but alas, I gotta choose someone else.

Detroit is a mess defensively, yes, but I don't think it's because Drummond is a horrible defensive player. Even Asik-Howard was a bad combination, and those two are better than Monroe and Drummond. Drummond without one of those guys on the court still doesn't fix the defense, as when he's paired with Monroe or without both opponents are scoring at a high 111 points per 100 possessions rate, but it's worth noting the Josh Smith-Drummond pair allows only 104.2, which is decidedly above average. He's still learning the subtleties of NBA-level defenses, but he's destroying people on the boards, even with a crowded frontcourt. A defensive rebound rate of 27.8 is truly elite, and an offensive rebound rate of 16.5 is Rodman-esque. And you want him off the boards, because he's one of the best at put-back dunks. They don't run plays for him, ever really, because he's still raw and unskilled, but he still manages to score 14 points per 36 minutes on high efficiency. He's not the most deserving pick, but at least he's super young and could turn into a legitimate star, meaning the pick won't look so strange years from now. Let's enjoy this explosive big man while he can still get up.

The Darkhorse All-star Team

As a bonus, here are my crazy, darkhose all-star picks, filled with players who do not conventionally fit the definition of an all-star, but deserve credit for either being high-performing role players or fringe all-star candidates.

Guard: Isaiah Thomas

Before Cousins had a few monstrous games, I would have easily called him Sacramento's best player, and I honestly still think it's closer than people think. Per possession, his stats are on par with all the best point guards in the league, with the exception of Chris Paul.

Guard: Manu Ginobili

When people talk about the weaknesses of advanced +/-, it's because the results often bounce up and down from year to year, and that specific lineup rotation issues, and who's playing with whom, can drive the numbers. However, Manu is 36 years-old and is once again near the top of the league in advanced +/-, and, excluding Duncan and Parker, who don't even start alongside him, it's with an entirely different cast of characters than the players in the mid-00's. Plus/minus is just about outscoring your opponent -- it's all that matters, in the end. And Manu is the master at pushing the score in his team's direction.

Forward: Andre Iguodala

Iguodala was an actual all-star, but like Manu he's missed a few games due to injury and has a smaller role this season (his usage dropped by almost 5 points.) But he's the glue to Golden State's roster right now, keeping a top-notch defensive unit together, distributing well for a team without a real back-up point guard, and he's even hitting his outside shots, doing a laundry list of important tough tasks for a good team. The Pippen comparisons are apt.

Forward: Amir Johnson

With Rudy Gay and Bargnani gone, Toronto was supposed to be bereft of talent, but they actually improved and immediately after Gay's trade Amir had some monster games. Few people know his game, but he's an active defender who rebounds and blocks shots well, while on offense he actually spaces the floor to nearly the three-point line (he's actually been taking one per game this season) -- yet even with all those long two-pointers he's still a very efficient scorer, much like Bosh. The Raptors have been much better when he plays for a long time, and he deserves some credit.

Forward: Anderson Varejao

Like Noah, he's a highly active center on defense who works well as a high-post facilitator, but health problems have limited his career. He's being wasted on a terrible Cleveland squad, but if he's ever traded to a contender more people would understand his value. As a bonus, he sorta looks like Sideshow Bob.


Guard: Wesley Matthews

He's the quintessential 3-D player. Shooting a high percentage behind the arc, even if you close him out he can drive to the basket with reckless abandon. He's also a good, feisty defender, and a key part of a surprising Portland team.

Guard: Lance Stephenson

He's rightfully in the right place here, in the land of role players and sub-stars. In the Pacers disciplined system, he's a tenacious defender and a dynamic offensive guard, rather than a player prone to weird, aggressive mistakes; he's the Stephen Jackson, circa 2003 Spurs, to Indiana.

Forward: Taj Gibson

As one of the game's best defensive players, he'll never receive much attention, especially when coming off the bench, but the value is there. Last year, for example, he allowed only 0.76 points per play on defense, from Synergy. He's strong, quick, blocks shots, rebounds, is fundamentally sound, and has learned from coach Thibs.

Forward: Nick Collison

Battier had a feature about being the "no-stars" all-star, but Collison has held that title for a long time now. Replacing Perkins in the rotation will do wonder for your +/- stats, but he's the role player every elite team would want: he sets great screens, he passes well, he defends at a high level on both pick and rolls and versus post players like Randolph, and he doesn't demand a large part of the offense.

Forward: Andrew Bogut

A former number one pick, Bogut, after years of injury, has turned into an underrated player. At the most important defensive position, he's one of the best in the game, and he's a useful complementary offensive piece due to his passing, finishing ability, and hard screens. Per minute, one could make the argument he's as valuable as all-star centers like Howard and Hibbert. Let's hope he stays healthy (unless he plays against my Blazers in the playoffs, however.)


Guard: JJ Redick

He'd probably make the all-star team in the east, but he certainly doesn't seem like one. However, his valuable shooting and playmaking has helped a Chris Paul-less offense stand the rigors of competition. He's second in the league in distance traveled per 48 minutes, among qualifying players. He plays like a disciple of Reggie Miller and Ray Allen, running around screen after screen for an open shoot, knocking them down with a high degree of precision, and putting an extra gear in an already great offense. I have an appreciation of this type of player and a theory that this is perhaps the best valuable type of offensive support for an elite team relative to its role on the team (i.e. adjusting for volume.) Redick's the piece that makes this Chris Paul team scary.

Forward: Kevin Garnett

Yes, I did say this was a crazy team. He doesn't play enough minutes anymore to deserve an all-star spot, but once again he's quarter-backing the defense and helping the crew in Brooklyn rise to 0.500 basketball while trying to put out wildfires everywhere. Lopez is gone, Deron has been hurt, and the coaching has been ... troubling, but they play better with Garnett. As I said, this is my crazy all-star team.

Tuesday, January 28, 2014

The 2014 Western Conference All-star Game Picks

The ballots are in, and the game will commence with two center-less lineups as Kevin Love edged out Dwight Howard and the east's premier center, Hibbert, doesn't have the flash of Carmelo Anthony. Positions have been an issue with previous all-star games. For years, there was much media attention about the death of the NBA center because the reserve center (and sometimes even the starting center) was usually an awful choice. Duncan insisting he was a power forward exacerbated this issue, and players such as Ilgauskas and Jamaal Magloire were invited to the game.

The NBA's solution, much like their attempt to "fix" the lottery, was an overreaction and just creates a new problem. The real problem with the strict positions of the previous ballot was that many players waffled between PF and C, as the distinction between the two positions had been all but erased on many teams, but for the all-star ballot Duncan was either a PF or a C; there was no in between. The guard-forward-center structure is an archaic one from decades ago, and one possible solution would be to change it the modern standard of point guard-wing-big man. There are still issues with combo guards and picking a point guard in a weak year, but at least it's better to have a subpar year from Kidd in the game than someone like Magloire because at least a point guard can make passes and flashy plays. Instead the NBA oddly decided to group centers in with forwards, and as a result we have two strange looking starting lineups without centers.

While there's nothing abjectly horrifying about this, because at least the most popular players are picked and that's the goal of the NBA ballot, the all-star games represent time capsules to explain the history of the league and it's important they're representative of the current league. Some people like to judge eras by looking at the all-star team to see who was playing, and a George-LeBron-Carmelo frontcourt will look bewildering. To put it simply, I would prefer an NBA starting lineup that makes "basketball-sense." You don't want the first reaction to be, "Huh? LeBron's the center?"

The Correction

You vote for one of each: PG-SG-SF-PF-C

The tweak? A player can be voted for multiple positions based on how he's actually used in a basketball game. LeBron James, for example, would be eligible for SF or PF, Roy Hibbert just C, and Stephen Curry PG and SG. This would eliminate some of the weakest choices in all-star history, as well as any strange lineups where LeBron is the "center."

Western conference picks


Guard: Chris Paul*

Even though he's currently injured, I think that he's the best point guard in the league and the third best player in the game (when healthy.) There will be some noise made about how the Clippers have been playing without him, but it's been against a weak schedule and they just got JJ Redick back. (And Paul does have a +5.1 raw plus/minus so far.)  But Paul's same critics ignore the evidence of his value from previous seasons when he was lost for stretches and his team was significantly worse, including when he switched teams and revived a franchise that was considered a laughingstock. He should receive more credit for this.

Chris Paul is still the best point guard in the game, a master tactician and perfectionist who controls the game -- SportVU credits him with 25.3 points created from his assists, far and away the best in the league, along with his 19.6 points per game on high efficiency (which makes him responsible for roughly 60% of the points created whenever he's in the game.) That he can do this while only averaging 2.5 turnovers per game is remarkable, and it makes him a one-man offense. If he's healthy, he deserves to start.

Guard: Stephen Curry

After arguably the best three-point shooting season ever and a surprising run in the playoffs, he's adorned as an NBA star and probably the least-hated top player in the league. The gap between Carmelo and Curry's shooting efficiency last season was smaller than people would think, due to Curry's mediocre scoring inside, and he causes too many turnovers, but he still creates a lot of scoring opportunities for his teammates due to his ability to shoot off the dribble long distance and even elite defenses don't know how to stop him. He's taken a higher load on offense this season while maintaining his efficiency; he's clearly taken a step forward, despite his lowest three-point percentage ever. His quick shooting ability makes him dangerous, as SportVU has him only taking 2.6 catch-and-shoot three's a game ... relative to his 8.4 total attempts per game behind the arc.

Forward: Kevin Durant

Durant, after a remarkable season of shooting efficiency that was only eclipsed by a legendary one from LeBron, is currently in the pole position for the MVP race, and it's well-earned. He's north of 30 PER, which is a rare threshold; his Win Shares per 48 minutes from basketball-reference of 0.329 would be the second highest mark ever behind Kareem in '72; and a plus/minus model (seeded with a statistical model for a prior) rates him as the top player in the league. To put it simply, his game is unfair to the rest of the league. Given his height and long arms, he effectively has the size of a center, but he's one of the best shooters in the league and has become adept at putting the ball on the floor and passing. He's an impossible cover, leading the league in points per touch, which is impressive for someone with as many assists and rebounds as him, and also leads the SportVU statistic for pull-up three-point percentage at 47% out of 56 qualifying players. What are you supposed to do with a player who's at least 6' 10" in shoes who can hit those shots? If you crowd him, he'll go to the rim where he hits his shots at 72% within five feet and nearly 90% from the line. He's a historic force, and he'll go down as one of the best players of all-time.

Forward: Kevin Love

Due to Minnesota's position out of the playoffs currently and their problems in close games, Kevin Love's credentials are being criticized. He's not a "true star" because he can't carry his teammates to wins; he's putting up empty numbers that don't mean anything; he can never have a central role on a great team; etc. People have labeled him a garbage scorer, somehow, who can't create his own shot ... he's currently fourth in the league in points per game with very good efficiency. How is it possible to score 25 points per game while being a garbage scorer? Via Synergy, 25% of his plays have come from post-ups and a good 0.88 PPP efficiency -- that's a high volume of post-up plays, as last season Duncan was at 28%, Brook Lopez 29%, Dirk 26%, and Cousins at 25%. He's the highest non-point guard in touches per game, via SportVu, and even when ignoring defensive rebounds (frontcourt touches only) he's still ranked high among non-guards, ahead of guys like Aldridge, Wade, and Dirk.

People complaining about the lack of elite post scorers in the modern league can look no further -- here's a guy who clears the 20/10 marks by a mile, posts up pretty frequently, can pass well, hits his free throws, sets up his other teammates, shoots behind the three-point line well, drawing out big men and creating attention. He's not conducive to winning, yet when he's on the court, according to, his team has a point differential of 9.8 and when he's off it's at a pathetic -10.0? Raw plus/minus has its problems, but even after using advanced adjustments he's still near the top. We've been through this song and dance before with another Kevin. In his prime, Garnett missed the playoffs two straight seasons, and people questioned his superstar status. The next season, he was given great teammates, Pierce and Ray Allen, and immediately won a title with one of the most dominant seasons from a team since Jordan retired. We have failed again to learn from history.

Forward: Dwight Howard

There are some interesting arguments for another forward over Howard, but I prefer a more complete lineup with players in their ideal positions. Paul/Curry/Durant/Love/Howard is a lineup I believe will hold up historically and looks more like a Dream Team. Howard's not a popular player anymore, due to personal problems and a disaster in Los Angeles, but he's still an incredibly effective defender and a better offensive player than people care to admit. He cleans up the boards, remains one of the best rim protectors in the game, has regained his quickness on the perimeter, and posting him up is not a good idea -- I've seen players like Pekovic get destroyed by Howard. People obsess over raw point totals, but I'm more interested in how he passes out of the post and sets screens. And his post game is much better than his detractors indicate, as he's learned counter-moves and is nimble when facing up and driving to the basket. With Marc Gasol sidelined for weeks with a knee injury, I believe Howard was the undisputed best center in the west, since Cousins has problems defensively and Duncan can't play the minutes Howard can. Without Howard, Houston's defense falls off a cliff -- and given his high percentage scoring and work on the boards, and he's an easy all-star.


Guard: James Harden

I have some reservations about Harden, though not as severe as many critics, but he does well with box score metrics and impact stats, from RAPM to the fact that he's the lead scorer/creator for one of the best offenses in the league. His concentration on the NBA's hotspots -- the three-point line and the rim -- may seem like a cheap way to score, but it's effective and style points are not given. His defense is awful, one of the worst for full-time starters, but as a guard it's not enough to offset his scoring. He also has the ignominy of covering the third fewest miles per 48 minutes in the NBA and last among guards, via SportVu and for players with at least 20 games and 20 minutes per game. This is a little surprising given Houston's fast pace, but when he gets the ball he's going in a straight line to the rim and he's not known for his activity on defense.

Guard: Damian Lillard

After winning an Rookie of the Year award mainly because he played heavy minutes, Lillard has taken a step forward as a scorer and as a distributor. He's cut his turnovers down, improved his defense, and has been lights-out from the three-point line. If it weren't for Curry last season short-circuiting our understanding of outside shooting we'd be more impressed by Lillard. He's taking 7 a game and hitting 42% of them, many off the dribble. He's also been incredible in clutch situations, making ridiculous game-winning shots and taking over in the fourth quarter. If there are any doubts of Lillard over another point guard in the talented west, there's the fact that he's one of the two leading contributors to what may end up being, by some measures, one of the best offensive teams of all-time. The difference between Portland's offensive efficiency compared to the league average is 7.8 currently (via b-ref.) The only teams since the merger with superior stats are the 2004 Mavericks (Nash and Nowitzki flanked by shooters) at +9.2 and the ground-breaking 2005 Suns with Nash/Marion and a young Stoudemire at +8.4. They're also close to the crowded group 7.7 points better than the league average: two more seasons of Nash, the '02 Mavs and the '10 Suns, the Stockton/Malone '98 Jazz, and the '97 Bulls. And adjusting for strength of schedule, Portland jumps ahead to +8.5.

Forward: Dirk Nowitzki

It's no surprise Dirk is aging so well. His height and shooting won't deteriorate as quickly as the NBA skills others rely on like lateral quickness. His per possession stats, with the exception of his rebounding, aren't too dissimilar from his stats in his MVP-days, and Dallas is riding a very good offense without too much other help to a playoff spot in a tough, tough conference. He's a darling of the advanced stats crowd for a good reason -- he's once again near the top in adjusted +/-, and his sterling shooting efficiency combined with his obscenely low turnover rate provides a tremendous amount of lift to any team's offense. And a more nuanced, closer look at his game uncovers more treasures: his ability to draw out defenders as a big man and space the floor is hugely important for the role players around him, and his offense holds up very well against even the best defenses because there's not much you can do about a one-legged fade-away jumper from a seven-footer. Advanced stat gurus don't love the midrange game, but things change when you consistently make 50% of your shots with few turnovers.

Forward: LaMarcus Aldridge

Box score lovers will complain about Aldridge's shooting efficiency (and sometimes his lack of offensive rebounds), but, as was pointed out with Lillard, he's a vital part of an electric offense. With a usage rate near 30, one of the highest in the league, and a below average TS%, you'll occasionally hear a criticism about Aldridge's "chucking." However, my counter is this: with how good the team is on offense now, and how large a part of the offense Aldridge is, how much better would they be replacing him with an average player? With that kind of simple analysis, you would be forced to posit the Blazers would turn into the best offensive team ever, arguably, and the best team in the league. But perhaps the highest usage player on the league's best offense is, gasp, a very good player. A cursory glance of the box score misses out on a few factors here. One is that Aldridge spaces the floor well, but you can't tell from the box score because he takes long two-pointers. And like Nowitzki, he has a tiny turnover rate, which makes him more efficient than one would think. The team also plays well off him, as any double teams will be thwarted with a kick out to their shooters, and he's a tough cover because of his size and length coupled with his ability knock down fade-away jumpers. Whenever Portland can't generate an open three or attempt at the rim, they can always throw it to him. He's also a very good defensive player, not an "anchor" but a tier below due to his fundamentals, quickness, and length. The MVP talk is probably an overreaction, even though he leads the league in "pure" RAPM; where was this MVP talk the previous two seasons, when he was almost the same player?

Forward: Blake Griffin

For the last required big man in the west, there's one sure pick left: Blake Griffin, who's been earning more respect lately for performing well with Paul out with an injury. Since his stats were virtually the same in his rookie season before the trade for Chris Paul, it was a strange criticism that he would be nothing without Paul creating for him. The talk about his lack of a post game was also hyperbole. Shaq only had a few moves, but variety doesn't matter when can simply dunk over everyone. The legitimate concern was his jump shot and free throw touch. Thankfully, he's cleared the 70% mark from the line, and his jump shot has looked better, smoothing out the hitch in his release and setting a career high in his percentage from the midrange zone. Griffin has actually rated well by advanced plus/minus metrics for a while. Few people praise his ability as a ball-handler and passer, and that versatility typically adds significant value to big men. He's also hedged enough weaknesses defensively that it's no joke he's starting in the all-star game.

Wild cards:

Guard: Mike Conley

I had a difficult time picking between Tony Parker and Conley for one of the last guaranteed spots. Tony Parker is at the helm of one of the league's best offenses, but the counter is that Conley has had very little help in Memphis with Gasol's injury, age bringing down Randolph and others, and little talent besides the core players. Prince shouldn't be a starter anymore, but the Grizzlies have had few other options. Conley has responded with his best season yet, raising his usage level while maintaining his efficiency and curtailing his turnovers by a significant amount. It's remarkable that Memphis had a league average offense, while Parker has the benefit of possessing a large array of toys to work with in the half-court, orchestrated by a Hall of Fame coach. I'd still rather have Tony Parker on offense, of course, for his mastery of pick and rolls and ability to finish inside, but Conley is one of the best defensive guards in the league -- Parker is simply decent at best. Likewise, Conley gets the nod over Dragic for the same reason. Like the Memphis guard, Dragic is doing a lot for a team that shouldn't otherwise be competing, but the Dragon is quietly one of the worst defenders at his position.

Forward: Anthony Davis

Choosing between Anthony Davis, DeMarcus Cousins, and Tim Duncan is torture. They have entirely different situations and playing styles. They own stats in different ways -- Duncan's team is one of the best in the league, Davis leads the league in blocks and flashes an efficient, versatile game, and Cousins crushes people with high volume scoring and rebounding. It's easy to look at Davis' stats and think he's a Defensive Player of the Year contender -- if that's true, then how are the Pelicans last in defensive efficiency? Looking further into this issue, coaching problems on defense are the likely culprit here; Anthony Davis isn't a glaring negative. Since they're cycled through a few awful defenders and useless players with Davis in the frontcourt -- Ryan Anderson is lights-out on offense with Davis, but there are still issues on the other end of the court -- he hasn't found a competent frontcourt partner. Since Davis can stop perimeter players in isolation and recover quick enough to block shots at the rim, and SportVu has him currently 15th out of 80 players in opponent field-goal percentage at the rim when he's near the player, I have to give him some credit here. He's in a terrible environment for defense, and while he's prone to mistakes, I still think he's more of a legitimate all-star overall when taking a step back.

Cousins is lighting up the scoreboard by focusing more on his strengths and attacking the basket, but I still have issues with him as a player. Davis makes mistakes on defense, which is normal for a 20 year-old, but Cousins does too and he can't make up the ground with his quickness and shotblocking. People will point to his steals, but that's indicative of a player who gambles far too often -- he's still limited by foul problems and this is his fourth year in the league. He picks up loads of rebounds, but those are sometimes deceptive because some players, like JJ Hickson, don't box out and would rather crash the offensive glass than get back in transition defense. Per NBAWOWY, the Kings rebound 76.1 percent of the available defensive rebounds when Cousins is on the floor, and 74.5 percent when he's off the floor. That's a significant decline, one that would drop you a few places in the league-wide DRB% rankings. But when Davis is on the floor for New Orleans, their DRB% is 74.2, and off? A pathetic 70.6%. Davis has few rebounders on the bench behind him, but it's the same with the Kings who often used Derrick Williams as a PF.

Duncan makes this issue stickier. His team is much, much better and he is a central reason why. He's slipped, sure, and can't make the same kind of high percentage plays Davis can but his defense is, of course, fundamentally sound and he uses his size as well as anyone. His adjusted plus/minus stats aren't any better than the two young studs, but it's more difficult to provide +/- "lift" to an elite team than a bad one. Plus, Duncan has to contend with an elite bench and a coach who's coasting through the season. I'm almost certain Duncan is the better player right now ... but only per minute/possession. So why take Davis? He can play heavy minutes, has no behavioral problems, and doesn't collect fouls by the bucket. It's a close call, but I'll take Davis because I'm slightly more certain he's the more valuable player to any random team in the league.

*Injury replacement for Chris Paul (or anyone else): Tony Parker

I've outlined the reasons why I'd take him over, say, Dragic above in the Conley section. He's still a very good player and helps run a fine offense. For a fun stat, he's 9th in distanced traveled per 48 minutes from SportVU -- advanced technology telling us what we already know about the speedy, active point guard, but it's nice to have confirmation. He has a slim lead over other candidates like Cousins, but the point guard crop in the west is unusually strong even without Westbrook gone and I'm positive Duncan doesn't even want to participate. And if Popovich coached the all-star game, would we put it past him to bench his own players for the entire game?