In what was already the third announced injury replacement for the 2018 NBA All-Star Game, it was announced after the injury to Cleveland Cavaliers big man Kevin Love that his All-Star spot would go to Miami Heat point guard Goran Dragic.
The switching out of a big man for a guard re-balanced the rosters after the previous injury replacement announcement had seen Washington Wizards point guard John Wall replaced by Detroit Pistons centre Andre Drummond. So that part makes sense. And to be sure, Dragic is an excellent point guard. Miami certainly had to scramble to cover for him when he wasn’t around. So that part makes sense too.
What makes less sense is why Goran Dragic got the nod over Charlotte Hornets point guard Kemba Walker.
In his season to date, Dragic is averaging 17.2 points, 4.8 assists, 4.0 rebounds, 0.9 steals and 2.3 turnovers per game, on a 52.7% true shooting percentage.
In his season to date, Walker is averaging 22.9 points, 5.8 assists, 3.3 rebounds, 1.2 steals and 2.2 turnovers per game, on a 56.2% true shooting percentage.
Dragic has a 16.0 PER, 3.0 win shares, a 105 offensive rating, a 109 defensive rating, and a -4 net rating.
Walker has a 21.9 PER, 6.1 win shares, a 114 offensive rating, a 110 defensive rating, and a +4 net rating.
Empirically, by any individual statistical measure be it basic or advanced, Walker has done more on the court that Dragic has. And while it is as ever true that statistical measures only count for a certain amount, the above is very one-sided in Walker's favour.
Fortunately for him, Kemba has now made the All-Star game himself, albeit in unfortunate circumstances. A remarkable fourth injury between the announcement of the teams and the game itself saw New York Knicks big man Kristaps Porzingis tear his anterior cruciate ligament, and Walker got his spot. But this is still not exactly justice. After all, why did Walker ever rank behind Dragic anyway?
The main difference between the two is in the win column. Despite all their injuries, Dragic’s Heat enjoy a 30-26 record and the eighth seed in the Eastern Conference, while Walker’s Hornets languish at 23-33 and tenth place. Walker has in a sense had his All-Star candidacy penalised by the fact that an overexposed combination of the inexperienced Malik Monk, the unsuitable Julyan Stone and the never-developed-and-apparently-also-lost-all-confidence Michael Carter-Williams back him up.
It is of course to Miami’s credit that they have a stronger unit around Dragic. With depth up front and on the wing, with a lot of athletic defensive-minded players, and about 46 players who would happily take a turn bringing the ball up court, the Heat have an identity and a toughness on the court that has seen them be a tough cover all year, even when undermanned talent wise. Walker's Hornets do not have that. Dragic's Al-Star candidacy should not be penalised by this, and his role in making those role players successful in their roles should also not be understated.
What the Heat also have, however, is a Pythagorean expected win/loss record of 26-30. Essentially, they only have the record that they do because of a statistical anomaly.
The Pythagorean expected win/loss record is designed to be a calculation of how good a team truly is as a sum of its total body of work. Win/loss record is of course also a barometer of such, and also the only one that matters in the end. But the Pythagorean approach is designed to account for variance.
In practice, it measures teams out at how good they are overall. In a very loose approximation of things, if a team is first in the NBA in offensive rating, but last in defensive rating, they could well be an average, middle-of-the-pack team, and Pythagorean analysis may well place them as such. But it is theoretically possible for a team to have a .500 record, sneak all its wins by three points, lose in giant 127 point blowouts, and have a winning record despite metrics biased massively towards being an overall poor team.
To a lesser degree, this is what has happened between the Heat and Hornets. Charlotte has a Pythagorean expected win-loss record of 27-29 while Miami's is 26-30. Charlotte has an offensive rating of 107.4, a defensive rating of 108.1, and a net rating of -0.7. The Heat, meanwhile, are at 105.4, 106.3 and -0.9 respectively. That is basically the same calibre of team, with the small edge to Charlotte.
For 44 minutes, these two teams play at pretty much the same level. But down the stretch, one can close out games, and one does not. That is why one team is good to us, and one is not. And therefore perhaps on a largely subconscious level, that is why one player is particularly good to us, and one is not.
To be sure, Walker’s own limitations in the clutch are a part of why the team struggles. Walker is now 0-14 in his career in last-second shot situations to either win or tie the game, and making even a couple of those this season would have made the situation above much closer. This is especially true if he had made one against the self-same Heat team whose clutch success is key to them being as high of a seed as they are back on 21st January. (Or, indeed, in all four games between this season; Miami won all four contests in the season series, but by only a combined 16 points. With better clutch execution against Richardson et al, this problem, and this piece, do not exist.)
It is however significant here to note why Walker missed that shot against the Heat back on the 21st. Miami knew Kemba would get the responsibility to create in the halfcourt, and thus could drape the gangly 6'7 defensive specialist Josh Richardson all over him, knowing that there was scant little chance of any other Hornet player getting charged with the task. The play as designed was extremely predictable. Of course the ball was going to Walker. Who else could do anything off the dribble?
Perhaps this is a failure of coaching; a dearth of options does not necessitate key possessions automatically involving no passes. Perhaps it is partly inevitable; Walker will always be among the smallest players in the NBA, and this will always count against him in single-play situations, when opponents can drape size and double-teams all over him.
Yet perhaps the most obvious culprit is the construction of the Hornets’ roster.
If the Hornets had a more diversified and talented offensive unit, and more depth in the back court, they would be a better team. If they had invested all the money more wisely and more on two-way players, and in more than just one player (or two, when Jeremy Lamb is having a decent night) who can create their own shot, they would not be moribund in this way. They would have a better team, and by default, all the players on that team would then look better.
They would also have an All-Star in Kemba Walker they would no longer perhaps to sell, who would look even better than he does now (especially in clutch situations), and who people would know is better than Goran Dragic.
To be sure, both Walker and Dragic are excellent players. Dragic's constant snaking into the lane, use of angles, pull-ups and reverses make him a constant thorn in the defence, while Walker's more dynamic ability to create space with the dribble makes him a 20 point per game scorer despite his lack of size. Walker, though, is simply playing better. And yet Dragic got the first All-Star nod based largely, it seems, on a subconscious team-related bias.
That bias is also built on a transient foundation. Regression to the mean has been happening for Miami, who have lost five of their last six games, with the five losses coming by a mere 22 points combined. This the same some of tight margin they had built their 29-21 record, fifth seed and Goran Dragic's All-Star candidacy on. The Pythagorean analysis tried to warn us. If Miami had been an eighth seed on a losing streak at the time the Love injury occurred, would he have been the replacement pick?
Late game execution affects record, certainly, and good clutch play is a vitally important thing. But do we overvalue it?