Pelton mail: How early could LeBron have been the No. 1 pick?

This week’s mailbag features your questions on LeBron James, Chris Paul and James Harden, and more.

You can tweet your questions using the hashtag #peltonmailbag or email them to [email protected].


OK, so this question from the proprietor of the indispensable kenpom.com wasn’t posed specifically to me, but as soon as I saw it I wanted to break it down for the mailbag.

Let’s stipulate two important ground rules:

1. The team that drafts James gets him for only the first seven years of his career, after which he changes teams as an unrestricted free agent as he did in real life. (This means that the semi-serious answers of drafting him as a baby aren’t realistic unless doing so establishes your team’s city as James’ hometown and means he returns home after winning two championships elsewhere.)

2. We have perfect hindsight about how James and other players develop. (I’m not sure this is what Ken was actually thinking but it makes the answer more interesting from a statistical standpoint.)

So now we need to estimate how a high school-aged LeBron would have performed in the NBA. To estimate this, I took a look at a best-fit quadratic curve for James’ actual NBA performance by age:

Kevin Pelton

I also compared the win percentages of rookies directly from high school to their minutes per game to get an estimate of how much a less-effective, younger LeBron would have played. That yields the following estimates of his win percentage and wins above replacement player (WARP) season by season:

I wouldn’t actually put too much stock in these but it seems plausible that as a sophomore in high school, James could have played replacement-level basketball as a little-used reserve. Anyway, we can use these estimates to show how many WARP James had in his first seven seasons as compared to the best available player in each draft before 2003. (Remember, we’ve got perfect hindsight here so Pau Gasol is going ahead of Kwame Brown in 2001, for example.)

Best available player, 1998-2002 NBA Drafts
Draft WARP Best WARP
1998 38.6 Paul Pierce 98.2
1999 61.5 Elton Brand 97.8
2000 80.3 Michael Redd 48.2
2001 102.7 Pau Gasol 71.8
2002 127.0 Yao Ming 66.9

You could probably still make a case for taking James over Elton Brand in 1999, given you would have gotten two years of top-five play from James before he left, but he would have basically been just taking up a roster spot the first two or three years.

By the 2000 draft, after James’ freshman year, that’s down to a year or two on the bench and you’re getting three years of elite play from James. In an average draft, that’s probably enough to take him No. 1. As compared to the abomination that was the 2000 draft, that’s a no-brainer. So I would conclude James should have been the No. 1 pick of any of the three drafts before he was actually eligible.


“With two players who have averaged double digit assists in their careers, I was wondering if James Harden and Chris Paul would set the record for the most combined assists between teammates in a season and how they would stack up in comparison to other play making duos.”

— Midilan

Let’s focus on duos since the ABA-NBA merger. Since then, here are the teammates I have making the most combined assists:

Most combined assists, two teammates
Team Season Player 1 Assists Player 2 Assists Total
PHX 1988-89 Kevin Johnson 991 Jeff Hornacek 465 1456
DET 1984-85 Isiah Thomas 1123 Vinnie Johnson 325 1448
UTA 1990-91 John Stockton 1164 Karl Malone 270 1434
UTA 1987-88 John Stockton 1128 Rickey Green 300 1428
LAL 1984-85 Magic Johnson 968 Michael Cooper 429 1397
LAL 1982-83 Magic Johnson 829 Norm Nixon 566 1395
LAL 1981-82 Magic Johnson 743 Norm Nixon 652 1395
LAL 1985-86 Magic Johnson 907 Michael Cooper 466 1373
UTA 1991-92 John Stockton 1126 Karl Malone 241 1367
DET 1978-79 Kevin Porter 1099 M.L. Carr 262 1361
(since 1976 NBA-ABA merger)

With the notable exception of the Johnson-Nixon backcourt, for the most part these “duos” are rather one-sided in nature. In addition to those two seasons, there have been 14 others in which two teammates both accumulated 500 assists, just three of them in the past 20 years: Stephen Curry and Draymond Green the past two seasons and Steve Nash and Boris Diaw for the 2005-06 Phoenix Suns.

So suffice it to say I don’t see much modern precedent for the Paul-Harden combo, which should easily blow past 500 assists apiece. Despite Paul’s missing 21 games due to injury, the two players combined for 1,470 assists last season, which would surpass the modern record. Obviously their assist rates will go down playing together, though I’m not sure quite how much.

My SCHOENE projection system has a really crude adjustment for this, reducing the assist rate for all players on a team when the rate of projected assists to field goals looks too high. So Clint Capela’s assist rate declines relatively as much as Harden’s and Paul’s, in this case, which clearly doesn’t reflect what will really transpire. For what it’s worth, SCHOENE projects Paul (713) and Harden (681) to combine for nearly 1,400 assists, which would put them eighth on this list but also mark the most since the ABA-NBA merger for a team’s second-leading assister.


“A lot of people talk about Isaiah Thomas being on the decline soon because of his age. But his total career minutes are relatively low for a star at his age. IT played three years of college ball and averaged 26 minutes or less three out of his six seasons in the NBA. Shouldn’t his decline (if healthy) happen a little later than others since he hasn’t had as much of a workload throughout his career? IT is the same age as Russell Westbrook, a little older than Harden, but his career minutes are 13,214, while Westbrook is at 22,786 and Harden at 20,688. Shouldn’t we be more worried about decline in those two instead of IT?”

— Brian, Seattle

1 Related

So I haven’t specifically looked at the issue from this standpoint — players who have seen relatively few minutes for their age. In an edition of the Pro Basketball Prospectus annual series a few years ago, I studied it from the opposite side, looking at the Westbrooks and Hardens. There did seem to be a little evidence that those players declined more quickly than similar players who saw more moderate NBA minutes.

That noted, in general when I’ve looked at projecting future development, I haven’t found experience to add anything above and beyond the predictive value of a player’s age. And that makes sense to me because the decline due to age is less about wear and tear specifically and more about the general slowing down all of us experience as we pass our physical prime.

After all, we still see reserves and role players decline due to age at similar rates to starters playing heavier minutes. So I don’t think Thomas’ minutes totals will help him stave off the decline we can expect as he moves into his 30s.


Barring an NBA decision to move the 3-point line back, I don’t think there’s a realistic chance of the trend toward teams taking more 3s reversing itself. And that’s going to continue to put more pressure on centers to defend on the perimeter, both against the increasing number of stretch-5s and in pick-and-roll situations.

This past season in particular, we saw a trend toward teams putting their defensively challenged big men in reserve roles: Enes Kanter was the pioneer of this movement, which reached a new level with the Milwaukee Bucks giving Greg Monroe the most minutes of their centers but refusing to start him.

Bringing such players off the bench makes sense for a few reasons. Their ability to create shots is more valuable when surrounded by reserves than starters, and there are fewer dangerous pick-and-roll threats to deal with playing against second units. So I do think that’s probably the future for Nikola Vucevic and players of his ilk.

Jonas Valanciunas is in a slightly different category because he at least brings capable rim protection. In his case, a bench role might be more about finding a way to feature him offensively than it is hiding him defensively — at least during the regular season.

You might be interested in