January 8, 2009
The Best Point Guard in the East
Part Two

by Kevin Pelton

In Tuesday's part one of my look at the East's top point guards, I focused on how Boston's Rajon Rondo, New Jersey's Devin Harris and Orlando's Jameer Nelson have emerged as All-Star candidates. Today is more about the bottom line, as we take a deeper look at the statistics to try to discern their respective value.

SCORING

Player    TS%         Usg        PAR

Harris   .587   8    .286   3    8.9   2
Nelson   .617   3    .224  13    8.4   3
Rondo    .562  13    .179  39    4.7  15

For each stat, we'll offer the player's performance this season along with their rank among the 59 point guards who have played at least 250 minutes this season.

All three players have fine True Shooting Percentages and have been efficient scorers this season. Rondo is the odd man out, however, in terms of creating shots. He's below-average among point guards in terms of possession usage--though a certain "Big Three" has more than a little to do with that.

There are a number of junk-type metrics that combine the two stats for an overall scoring measure. What I've done here is create a "points above replacement" (PAR) stat that subtracts a replacement-level True Shooting Percentage (80 percent of league average) from each player's own True Shooting Percentage, then multiplies it by usage per 100 team possessions and by two (because True Shooting Percentage is points per shooting possession divided by two). This approximates how many points more than a replacement-level player are created by a player's shooting in about the span of a 48-minute game.

In this case, the specifics aren't important. No matter how you look at the issue, you'll find that Harris and Nelson have been elite scorers this season. The only point guard with a superior PAR is Chris Paul, and by just a hair over Harris. (For the record, LeBron James comes out as the league leader at 11.1 PAR.) Rondo, meanwhile, comes out middle of the pack amongst starting point guards. That doesn't sound like much, but for a player who is not needed to score by his team and is so good elsewhere, that's pretty impressive.

BALLHANDLING

Player    Ast%        TO%        Pass

Harris    8.4  19    12.1  13    4.09  20
Nelson    7.5  28    13.1  18    3.35  28
Rondo    11.2   6    19.8  48    9.26   5

My Pass rating is Ast/Min squared multiplied by Ast/TO rate (multiplied by 50). In most cases, it tracks assist rate pretty well; this time, it was so close it doesn't really add much information. That Rondo is the best passer of the group hardly comes as a surprise, but I wasn't expecting such a large gap between him and Harris. While Harris is averaging 6.6 assists per game to Rondo's 7.6 and Nelson's 5.2, that's inflated a little because he plays the most minutes of the three. Putting them on a level playing field shows Harris to be much closer than Nelson in terms of distributing the basketball than Rondo.

Rondo's high turnover percentage (even before a nine-TO effort Tuesday night in Charlotte) at least partially reflects the fact that he's more of a passer than a shooter. The leaders in turnover percentage amongst point guards tend to be catch-and-shoot guys who don't have the ball in their hands very frequently, while the pure point guards rack up turnovers in the process of creating assists. Amongst guys in the top 10 in assist rate, Rondo is better than average in terms of hanging on to the ball.

REBOUNDING/DEFENSE

Player    Reb%       Stl%       Blk%        PF%

Harris     5.4  38    2.3  18    0.4  25    2.9  17
Nelson     6.2  20    2.0  27    0.0  54    4.6  48
Rondo      9.1   2    3.6   3    0.3  30    3.8  38

As discussed yesterday, the strongest aspects of Rondo's game from a purely statistical perspective are his rebounding and his thievery. When it comes to the combination of those two skills, Rondo, Paul and Jason Kidd are in a league of their own among point guards. Harris and Nelson are much less impressive in these categories. Harris basically comes out average everywhere except rebounding (and then it's his offensive rebounding that is an issue much more than his work on the defensive glass). Nelson can board adequately and gets a few steals, but tends to foul quite a bit, which seems typical for players of his body type. I know it's not exactly in the job description for a point guard, but how has he played more than 900 minutes this season without blocking a single shot?

PLUS-MINUS (from 82games.com)

Player    NetO         NetD       Net+/-

Harris   +12.0   3    + 4.2  32    + 7.8   8
Nelson   +10.0   5    - 1.5  15    +11.5   4
Rondo    + 4.1  16    - 2.8   9    + 6.9   9

There are two surprises in the plus-minus numbers (which still come with an enormous grain of salt at this point of the season). The first is that the Magic has been better defensively this season with Nelson on the floor than with his backup, Anthony Johnson. The other is that Rondo has not been a bigger defensive difference-maker. This may be attributable to the fact that Boston defends well no matter what personnel is on the floor; Ray Allen (-5.4 net defensive plus-minus) has been the only Celtics player with whom the team has defended better than with Rondo.

THE BOTTOM LINE

Player    ORtg         DRtg        Win%       WARP

Harris   111.3   2    105.4   25   .684   3    6.3   2
Nelson   109.5   4    105.5   26   .627   7    4.0   9
Rondo    108.4   6    102.4    3   .688   2    6.2   3

When we put all of this together into my WARP Rating System, the trends we've already seen remain evident. All three of those point guards have been amongst the league's best offensively. Paul is the only point guard who has been more valuable on offense than Harris this season, while Chauncey Billups' sure-handed play allows him to slip in ahead of Nelson. Even without being a big-time scorer, Rondo still ranks sixth (behind Jose Calderon) because of the various ways he contributes to the Celtics offense.

Defensively, Rondo has a big advantage in terms of ranking. However, because there tends to be less separation at the defensive end, especially among point guards, the gap between him and Harris is virtually the same at both ends of the floor. Naturally, defensive statistics are much less reliable than their offensive counterparts, but I don't think calling Harris and Nelson average defenders is unfair to either. It's simply hard to call Harris anything better than average when he's played heavy minutes for the league's worst defensive squad and that team has defended better with him on the bench.

When we move to per-minute winning percentage and Wins Above Replacement Player, it's clear that Nelson has a tough time competing with Harris and Rondo even in what has been a great season. Add in the fact that his hot start will be somewhat more difficult to sustain, given its dependence on hot perimeter shooting, and Nelson finishes third in this group. There's no shame in that.

Choosing between Harris and Rondo is more challenging. Using very different skill sets, they've nonetheless managed to be nearly identical in terms of per-minute performance and value. One could say that both players are perfectly suited for their team's needs, Harris offering the shot creation the Nets so desperately need and Rondo leading his superstar-filled team with his ability to distribute the ball while playing a key role in the league's best defense.

Ultimately, I'd give the slight edge to Harris. The last couple of weeks have shown that teams can gameplan around Rondo's all-too-evident weaknesses (poor outside shooting and an occasional susceptibility to turnovers). Rondo's defense and rebounding contributions are consistent, but at times he can be a liability nonetheless. Harris never reaches that point. In an ideal world, both players would be named as reserves to the Eastern Conference All-Star team, but if the coaches can pick only one, Harris is the choice.

Kevin Pelton is an author of Basketball Prospectus. You can contact Kevin by clicking here or click here to see Kevin's other articles.