BBR Mailbag: Biggest Yearly Declines in 3-Point Attempts
Posted by Neil Paine on July 30, 2010
This is a question from Erik, who writes:
"I noticed the rather severe change in Josh Smith's 3-pt attempts from the last several seasons to the one that just wrapped up. After attempting 87 3's last year, 99 the year before and a cringe inducing 152 in '06-'07, Smith only attempted 7 3's in '09-'10. I was curious if you knew of any other players who had experienced such an attack in common sense? It seems like a huge drop off and a decision made for the betterment of the club. He played more games & minutes, had more shot attempts total, scored more points and did it all while basically eliminating the 3-pt shot from his game. Thoughts?"
From 2006 (his second NBA season) though 2009, Josh Smith attempted a 3-pointer on 12% of his field goal attempts -- despite the fact that he made just 27% of them, a rate 9 percentage points worse than the league average. And then, suddenly, he stopped shooting them: in 2010, threes didn't even make up 1% of his FGA (he went 0 for 7 on the year). In other words, apparently Smith finally got the message that he wasn't good at the 3-ball, and he abandoned it completely. How unprecedented is this? Here are the biggest single-year declines in 3-point tendency (3PA/FGA), relative to the league average, since 1981:
| Season | Previous | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| Rank | Player | Year | Age | Team | 3pTend | 3FG% | 3pTend | 3FG% | Diff |
| 1 | Brian Cook | 2006 | 25 | LAL | -3.8% | 7.0% | 27.5% | 3.6% | -31.3% |
| 2 | Devin Harris | 2006 | 22 | DAL | -15.0% | -12.0% | 16.2% | -2.0% | -31.2% |
| 3 | Keyon Dooling | 2004 | 23 | LAC | -5.6% | -17.3% | 24.1% | 1.0% | -29.7% |
| 4 | Greg Anthony | 2002 | 34 | TOT | 16.2% | -5.9% | 45.8% | 5.5% | -29.7% |
| 5 | Darrell Armstrong | 1998 | 29 | ORL | 1.9% | 2.2% | 31.3% | -5.6% | -29.3% |
| 6 | Ron Harper | 1998 | 34 | CHI | -3.3% | -15.5% | 25.1% | 0.2% | -28.4% |
| 7 | Michael Curry | 1998 | 29 | MIL | -13.8% | 9.9% | 13.7% | -6.1% | -27.5% |
| 8 | John Salmons | 2006 | 26 | PHI | -3.5% | -5.9% | 23.7% | -1.5% | -27.3% |
| 9 | Jamaal Tinsley | 2005 | 26 | IND | 10.3% | 1.6% | 37.2% | 2.5% | -27.0% |
| 10 | Derek Fisher | 2003 | 28 | LAL | 9.2% | 5.1% | 34.3% | 5.9% | -25.1% |
| 11 | Greg Buckner | 2007 | 30 | DAL | 13.2% | -4.8% | 38.0% | -0.4% | -24.8% |
| 12 | Earl Watson | 2008 | 28 | SEA | -4.5% | 0.9% | 19.5% | -3.0% | -24.0% |
| 13 | Tyrone Nesby | 2002 | 26 | WAS | -7.0% | -7.7% | 17.0% | -8.1% | -23.9% |
| 14 | Dale Ellis | 1987 | 26 | SEA | 10.5% | 5.7% | 33.0% | 8.2% | -22.6% |
| 15 | Muggsy Bogues | 1998 | 33 | TOT | -11.0% | -9.6% | 11.3% | 5.7% | -22.3% |
| 16 | Nick Anderson | 1998 | 30 | ORL | 12.4% | 1.4% | 34.7% | -0.7% | -22.3% |
| 17 | Kyle Korver | 2007 | 25 | PHI | 15.1% | 7.2% | 37.3% | 6.2% | -22.2% |
| 18 | Craig Hodges | 1991 | 30 | CHI | 25.2% | 6.3% | 47.1% | 15.0% | -21.9% |
| 19 | Yi Jianlian | 2010 | 22 | NJN | -14.9% | 1.1% | 5.6% | -2.4% | -20.5% |
| 20 | Quentin Richardson | 2006 | 25 | NYK | 20.7% | -1.8% | 40.8% | 0.3% | -20.0% |
| … | |||||||||
| 233 | Josh Smith | 2010 | 24 | ATL | -21.5% | -35.5% | -12.2% | -6.8% | -9.4% |
Viewed this way, Smith's transformation was not very rare in NBA history. However, Smith wasn't a mere role player like most of the names on that list -- he was big name, playing more than 2,000 minutes in both 2009 and 2010, and taking more than 800 shots each season. As a percentage of his shots, the difference wasn't huge, but Smith's change is far more unprecedented when you look at the decline in his raw 3-point attempts.
Among players who played at least 2000 MP in back-to-back seasons, Smith was actually the very first player in NBA history to go from taking 80+ 3-pointers to taking under 10 the following year (the previous record for fewest 3PA the year after an 80-attempt season was Derrick McKey in 1990, who went from taking 89 to taking 23). This was a very good thing for both Smith and the Hawks -- by cutting back on his threes, he turned into a more efficient player (109.4 ORtg vs. 103.5 in 2009), a more willing passer (he passed on 57% of his touches instead of 43%), a better offensive rebounder (9.0 ORb% vs. 6.5 in 2009), and the Atlanta offense improved to become the NBA's 2nd-best during the regular season.
The moral of the story: It's amazing what cutting out a gratuitous, inefficient aspect of your game can do.
July 30th, 2010 at 12:31 pm
My guess is you could show the inverse using Antoine Walker.
July 30th, 2010 at 1:05 pm
I don't understand the numbers in the table. How can 3pTend (3PA/FGA) be negative? How can 3FG% be negative? Do these numbers refer to the differences between seasons? If so, what is the difference column for?
July 30th, 2010 at 1:13 pm
It's all relative to the league average. For example, if you have a 3pTend of -15%, it means your 3PA/FGA was 15 percentage points lower than the league's 3PA/FGA, etc.
July 30th, 2010 at 3:34 pm
No mention of Shawn Marion? How about a complete career reversal. From 2002 to 2008:
364
265
341
290
252
198
192
2008-10:
53
53
19
I didn't bother doing the math percentage, but I'm going to imagine it's one of the most drastic changes in approach to a game we've ever seen.
July 30th, 2010 at 4:34 pm
@Anon X 2:
The scope of the article was largest yearly decreases. While Marion's 3PA's were certainly a huge reversal in the long run, his year-by-year declines weren't as tremendous. His shot selection changed more gradually -- 72.4% decrease for Marion from '07 to '08, compared to a 91.9% decrease for Smith from '09 to '10.
Also of interest, Marion's changing numbers are likely due to new teams to which he'd been traded. His big shift in 3PA's occurred in '08-'09, after his first offseason with a new team (MIA). It probably owed as much to a new system he was in as it did to any personal decision.
Smith, on the other hand, had no observable catalyst for his change. He had the same basic core around him (Bibby, Johnson, Williams, Horford) and the same head coach (Woodson). It appears that he simply decided over the offseason that he sucked at shooting the long ball and gave it up cold turkey.
July 30th, 2010 at 4:42 pm
It also makes me laugh when I see Marion on NBA 2k10 with a "D" for outside shooting, because I distinctly remember him running around with a 3 pointer icon next to his name in the NBA Live games around the middle of the decade.
It's hard to account for "could but doesn't" vs "can't" on a video game.
July 30th, 2010 at 4:47 pm
Smith changed because he realized he stunk at shooting 3s.
Marion changed because....? While not a great 3 point shooter, he wasn't bad.
You'd think in Dallas he'd shoot em. I understand that Josh Smith's % decline is greater, especially year by year, but he topped out at 152 3s. Marion shot 364! From 4.5 3s a game to 1/3! During that stretch he averaged over 3.5 per game compared to like 1.5 for Smith.
My point is that Smith was from an ineffective 3 point shooter who at times shot them to a non-existent 2 point shooter while Marion went from a high volume effective 3 point shooter to a non-existent 3 point shooter.
It's a matter of scale. One seems more drastic, even if more gradual.
July 30th, 2010 at 4:49 pm
2 point shooter should have been 3 up there. darn typos.
July 30th, 2010 at 5:06 pm
@JB, it is because NBA Live sucks. 2k beats it in every aspect. If you haven't seen the light and switched to 2k you may never recover from weakgameitis.
July 30th, 2010 at 5:12 pm
Josh Smith finally listened to his coach and stopped shooting them. He said he really never thought about it before. But wanted to do whats best for the team. Out of those 7 he shot last year im sure 3-4 were buzzer beaters at the end of the qtr or shot clock. Hes really became a beast because of it.
July 30th, 2010 at 5:36 pm
Marion has shot under 20% from 3 since leaving PHX; sounds like he's still shooting too many to me. He had two outlier seasons where he shot about .390 from three (before Nash btw), but otherwise he was never better than avg from 3, even in PHX.
July 30th, 2010 at 6:43 pm
Wait, wait... did I read that right? Atlanta was 2nd in the league in offense last year? I knew Jamal Crawford was da bomb!
July 31st, 2010 at 6:32 am
I wish Lebron were on this list...
It always pains me to see him average between 4 to 5 3PA per game and making just a third of them year in year out.
I mean, last year he shot .56 from 2 points. Why the hell did he keep launching so many 3s !!
July 31st, 2010 at 2:36 pm
Neil, where did this come from?
(Also, can you implement comment previews?)
July 31st, 2010 at 8:38 pm
That's from Bob Chaikin's work on "touches"; I talk about it in the Blog-tionary:
http://www.basketball-reference.com/blog/?p=1120
It estimates the # of times a player touched the ball in an attacking position on the floor, and from there you can calculate the % of times each touch ended in a completed pass, a shot, a turnover, or a foul drawn. Smith's pass/shoot/tov/fd in the '09 regular season was 43/37/13/7 on 0.97 touches per minute, but in 2010 it was 57/28/10/6 on 1.28 t/mp.
Also, we'll see what we can do about comment previews.
August 2nd, 2010 at 3:37 pm
Smith's percentages could also be affected by criticism on and off the court. This, however helped the Hawks and made Josh Smith a more strong conscious player on-court.
August 5th, 2010 at 2:47 pm
It should be pointed out that Josh Smith changed positions. He entered the league as a SF, last year he was listed as a PF. All these numbers and crazy stats... um... he's a power forward now. Might have something (A LOT) to do with it.