This week's mailbag features your questions on Playoff Rondo, Giannis' 3-point hopes and more.
@kpelton Daryl Morey said on radio recently that it's highly unlikely for a player to make significant improvements to their game after their first three years in the league, noting Clint Capela as an exception to that. Do the numbers back up the claim? #peltonmailbag— Evan Vracar (@Roca_Star) May 4, 2018
Appearing on the Dan Le Batard Show earlier this week, Morey said of Capela, "He really improved in his fourth year in the league, which almost never happens." Morey later added that "97 percent of players are pretty much set where their trajectory is going to be after their first three years."
Because Victor Oladipo made one of the largest leaps in NBA history during his fifth season, Capela didn't get as much attention as a candidate for Most Improved Player, but his development was certainly substantial. He went from a .577 player win percentage (the per-minute component of my wins above replacement player metric) to .670, more than doubling his WARP in the process.
Among fourth-year players, Capela's development was notable but not exactly historic. Looking at players who logged at least 1,000 minutes in both their third and fourth seasons, he ranks 43rd in improvement since 1977-78, though the fact that he improved so much while starting at such a high level is unusual. The other players who made larger jumps in their fourth seasons from a winning percentage of .525 or better are pretty strong company: Giannis Antetokounmpo, Gilbert Arenas, DeMarcus Cousins, Rudy Gobert and Scottie Pippen.
Let's consider how likely this kind of development is to happen by a player's experience level. Here's the percentage of qualifying players (again, minimum of 1,000 minutes in both seasons) who improved their player winning percentage by 90 points or better:
The year here is the player's experience in the first season, so Capela belongs in the Year 3 category. Morey's 97 percent figure is probably a slight exaggeration, since players make Capela-like jumps twice as often as that would imply before we even consider the growth from non-contributor to regular that isn't considered here because the previous minutes total was too low.
It's also worth noting, as we explored earlier this season in a mailbag, that age is the better predictor of development rather than experience. That's obvious among inexperienced players when we run the same chart with age as of the end of the first season rather than years in the league.
Capela was coming off his age-22 season, so his dramatic improvement was about 50 percent relatively more likely when taking age into account rather than experience. Nonetheless, his substantial development makes Capela an exception to the rule.
Given Capela is still just 23 (he'll celebrate his 24th birthday later this month), it's possible he could have more room for growth. In particular, it will be interesting to see if the Rockets continue to extend his minutes during the regular season. After increasing from 23.9 per game during the 2016-17 campaign to 27.5 during the regular season, Capela is playing 31 a night so far in the playoffs. That has helped him boost his scoring average to 15.9 points per game, which along with his 12.8 rebounds, 2.1 blocks and high-percentage finishing makes him a viable All-Star candidate.
The number of quality young centers in the Western Conference will work against Capela's All-Star hopes. Already, Gobert (who was injured much of the first half) and Nikola Jokic missed out on this year's team, even with DeMarcus Cousins injured. So despite playing at an All-Star level, Capela will have a tough time being recognized as such.
"My friends and I all agree that Giannis Antetokounmpo would be virtually unguardable if and when he develops a reliable 3-point shot. However, one friend quipped, 'If he hasn't developed one by now, he never will.' Giannis does have five full seasons under his belt, but he's still only 23 years old. Is there any evidence to suggest that a player loses the ability to develop a shot at some point, either due to age or due to being entrenched in their style of play?"
-- Grant Shao
Speaking of late-career development ... again, defining what qualifies as "a reliable 3-point shot" is key. I settled on attempting at least one 3 per game (minimum 20 games) and making them at league average or better. This is a pretty high bar, however; fewer than 800 players since the introduction of the 3-point line have had at least one such season, and more than a quarter of them did it only once. (Ray Allen has the most qualifying seasons with 18, one ahead of Reggie Miller; Dirk Nowitzki leads active players with 16 qualifying seasons.)
Anyway, by this measure there is certainly no point at which a player's ability to develop a 3-point shot is closed off. About 150 players, or around a fifth of all that ever do it, do so after their fifth season in the league. And the average age at which players qualify for the first time -- excluding players who come over from Europe well into their careers -- is between 25 and 26. There are two players who never qualified until their 16th NBA seasons: Grant Hill in 2010-11 and Pau Gasol in 2016-17, when they were 38 and 36, respectively.
Now, this is not an estimate of the likelihood of Antetokounmpo developing a 3-point shot, which would probably take into account his current 3-point percentage (85 percent of league average this year) and free throw shooting (around league average, which is a decent sign of potential as a shooter). I wouldn't bet on Antetokounmpo qualifying in the next five seasons at even odds. But the possibility can't be ruled out.
#peltonmailbag is there a good way to explain variations in Rondo's effectiveness besides "playoff Rondo" or "National TV Rondo"? Specific lineups make him more effective?— Stephen Herzog (@sherzogNL) May 5, 2018
So first, a confession: Much like Rondo does, I find the concept of "Playoff Rondo" to be somewhat overrated.
Rondo is certainly better in the playoffs, and that's rare in and of itself. On average, weighted by minutes, players see their per-minute win percentage drop about 6 percent in the playoffs because the competition is tougher. By contrast, Rondo's had gone up 2.6 percent before this season, which is pretty good. It ranks in the 89th percentile among players with at least 1,000 minutes played in the playoffs since 1977-78, and 10th when we limit to players with at least 3,000 minutes. That will likely go up a bit this year, since Rondo rates 8.6 percent better than during the regular season.
However, to imply Rondo is a totally different player in the postseason in my opinion both overrates his playoff performance and underrates how good he was in the regular season during his prime.
Anyhow, I think this year there is something of a reasonable explanation in terms of New Orleans' lineup change. Rondo played just 348 minutes during the regular season with the frontcourt of Anthony Davis and Nikola Mirotic, according to NBA Advanced Stats; that trio has already played 178 minutes together in the postseason, nearly three-quarters of Rondo's total. With Mirotic and Davis on the court, Rondo averaged 12.6 points and 12.3 assists per 36 minutes during the regular season, up from 11.3 and 11.3 overall.
That said, Rondo has been even better in the playoffs than he was with the spread lineups during the regular season, averaging 14.0 assists per 36 minutes. There's also no such obvious explanation for his performance in Boston given the Celtics' lineup consistency, or for national TV games where rotations were unlikely to change. The simplest explanation is probably that Rondo truly does focus more in bigger games.
"When accounting for shot quality using Second Spectrum's metrics, which player has scored the most points in a game this season above what would have been expected based on his shot attempts?"
-- Dan Feldman
Using quantified shot quality, which takes into consideration the shooter's location, shot type and nearby defenders, here are the top 10 this season in terms of most points above what we'd expect for an average shooter on the same attempts.
Because shot quality tends to vary less than shot-making, these performances are much more about hot shooting than taking difficult shots, per se; the minimum effective field goal percentage for a game in the top 10 was 83.3 percent by Stephen Curry versus Boston. Only one game here stands out for exceptionally difficult shot attempts: against the Clippers, Kevin Durant's shots had a qSQ of 41.4 percent, putting it in the 98th percentile for difficulty among games with at least 10 shot attempts.
The prompt for Dan's question was LeBron James' performance in the Cleveland Cavaliers' Game 2 win in Toronto, which ranked as just his fifth-best game of the season by this measure, in large part because we're mostly remembering the difficult fadeaways James made in the second half.
In the first half, James's qSQ was a robust 56.0 percent, which he beat by a little less than a point. After halftime, his shot quality dropped to 45.8 percent. Yet James shot an effective 71.1 percent from the field in the second half, scoring 9.6 points more than expected on his 19 shots. So it was certainly among the most impressive halves of shot-making we've seen this season.