The Iverson Effect

[mtamada]

Iverson's normal-season career TS% equals NBA average of his time and is greater than guards' average despite he had taken more shots than almost everyone else

[...]

Iverson is definetely one of the most underrated players of all time.

I guess it depends where you, and we, are rating him.

A TS% that's at the NBA average, while hoisting up huge numbers of shots? That's an all-star, no question. And maybe an all-pro.

But is it an MVP or a Hall of Famer? In my book, someone shooting that many shots needs to be ABOVE the NBA average in TS% to be considered for MVP. Either that, or be a defensive and rebounding wizard a la Bill Russell. Or get a boatload of assists like Isiah Thomas -- but even Isiah was never MVP, nor did he deserve to be.

[Kevin Pelton]

Iverson's normal-season career TS% equals NBA average of his time ...

I've got his career TS% as 1.7% worse than league average weighted by true shot attempts by season. But a lot of that is late-career improvement. Through 2004, he was 3.2% worse than average, and he only was better than league average once from 1998-99 through 2003-04 -- his MVP season, when he was 0.1% better than average.

By comparison, McGrady was better than average all four seasons in Orlando, including 8.7 percent better than average in 2002-03 while using 35.4 percent of the Magic's plays.

[Mike G]

Here are 100 highest Usg% player seasons since 1978 - http://bkref.com/tiny/Lfsk9
Sort by TS% (twice) to see the worst shooters among them.
The list is 'topped' by Jordan's 2002 season with the Wizards (.468)
There are only 4 other instances under .500 -- all of them are Iverson, 2000 thru 2004.
His MVP season, 2001, he hit .518, which is only 13th worst.

The Effect we are looking for in this thread is a player's effect on his Team's shooting (and pt-diff, TO, etc). It matters less -- or not at all -- whether he shot great or crap%, shot frequently or seldomly. The team effect (vs opponents) is what measures his effectiveness.
We're also seeing that some players reduce their team's turnover total, even as they rack up large individual TO totals.

A player's TS% should be compared to that of his teammates and opponents, rather than that season's NBA average. The Sixers and their opponents regularly shot .010 to .025 worse than the league; and that is what's relevant. They were built as a defensive team, not one that just out-shoots the other team.

It's hardly the case that Iverson never passed the ball. Back to that top 100 Usg% list -- he ranks as high as 8th (2005), at 37.6 Ast%
Of 10 players with career Usg% > 30, only LeBron and Wade have higher Ast%. At 28.8, Iverson is well ahead of Jordan (24.9), Kobe (24.3), Carmelo (15.7), Durant (15.3), etc.

[Crow]

Philly's team offensive efficiency during the Iverson era averaged about 19th in the league. That should be kept in mind when evaluating the Iverson effect. The team may have needed Iverson but they basically designed themselves to need him. Few teams were ever designed to need that kind of player as much and for good and bad consequences. Their success was more about great defense for a few years and good defense much of the time beyond that. In the end they won 6 playoff series total in his 11 year tenure. The design had its moments, but they were relatively few. There might have been a better design and a better way for Iverson to play.

[colts18]

to be fair to Iverson, the team wasn't really built well for him. Yeah they were built well when it came to defense and offensive rebounding, but he had no 2nd option to lighten the load and more importantly he didn't have any shooters to space the floor. Imagine if he was on a team like Miami (minus LeBron/Wade). Good 3 point shooters (Allen, Lewis), 3 and D guy (Battier), low usage floor spacing PG (Chalmers), 2nd option fine with being a 2nd option (Bosh). That would be a really good team for Iverson offensively (not defensively though). His usage would be reduced to 30-32% and the shooters would space the floor making it easier for him to drive to the basket. Plus he could make teams pay for doubling him. Iverson came at the worst time for a slashing guard like him. 99-04 was the strongest defensive era, even stronger for perimeter players. No illegal D plus handchecking. Once handchecking was gone, his efficiency went up.

[MW00]

Two points here, one, building a D around him (or catering one to him) was more important than catering a good offense for him, as offensively his greatest virtue was as a usage sponge who could maintain an okay ts% at huge usage rates, whilst on D he really needed a big pg who could cover twos because he wasn't big enough to do so.

The second would be that whilst he didn't necessarily have a great, or perfectly fitting offensive options, throughout his Philly spell he had a few competent 2nd banana types e.g. Derrick Coleman, Jerry Stackhouse, Joe Smith, Tim Thomas, Larry Hughes, Toni Kukoc, Keith Van Horn Chris Webber and other scorers like Corliss Williamson, Clarence Weatherspoon (not all at their peak but still). It didn't work, Iverson needed and succeeded best when surrounded by defender-hustlers who didn't want/need the ball. Only later, in Denver did he work well/succeed with another high usage player.

[Crow]

In his prime 5 season window Iverson played with a 25+% usage 2nd guy only once and averaged only .6 of a fellow starter with usage over 20%. The situations were different (young guys rising. old guys fading) but in each case the 2nd guy's usage was lower with Iverson than either before or after.

During that time period only 40% of team seasons had a 30+ minute guy with usage equal to or above 28%. iverson had 5 of the ten highest (over 32.5% usage). None of the others even made the playoffs in these seasons despite being named Jordan, Bryant, Pierce, McGrady and Stackhouse. It doesn't look like very very usage and pedestrian to weak TS% was the way to win a title. A couple of guys did win a title with usage over 30% but with much better TS% (and a different position in the offense).

If his usage had been lower, maybe even under 28%, there might have been a different enough dynamic. For better or worse I don't know for sure but with Iverson playing the Iverson way, no titles. Iverson only got out of the second round once.

[colts18]

You have to mention handchecking with Iverson. Notice how in 2006 his TS% went up a lot despite him being out of his prime. Imagine if 01 Iverson could drive to the basket without being handchecked. Iverson's top 3 Ftr seasons came from 06-08 when he was 30-32 years old.

I don't think you can win with Iverson at 35 usage%, but you can win with him at 30%. In 2008, he and Carmelo combined for a 58 usage% and .567 TS% for Iverson, .568 for Melo. Both guys were considered inefficient guys yet they had their highest TS% with each other.

[Crow]

Handchecking is a factor, suggesting a tougher environment for him to do what he did and a less wise environment for the team to make him their primary offensive strategy.

The Iverson era was GM'd by Billy King (1 conference title in 12 total years) and coached by Larry Brown (55% regular season NBA win %, 52% playoffs overall, one season above .600 with Iverson). Their one long playoff run beat teams under Isaiah Thomas, Lenny Wilkens and George Karl, none of whom were above .450 in playoffs for their careers. Twice they were taken to the 7th game despite being the top conference seed. Without that one season, Iverson's playoff record would be critiqued harder.

As it was: career playoff eFG% below 44%, TS% below 49%, winshares per 48 just barely above .100 or average. His regular season career number was only .126 which put him in the bottom half on ws/48 for guys with 20,000 minutes since 1980 and a 25+% usage for their careers.

Iverson's career team win-loss record was 466-448 regular season and 30-41 in the playoffs.

Of the 39 players who scored 20,000 or more points in the regular season his winshares per 48 minutes is 6th worst of 39 http://bkref.com/tiny/wq5kN and his overall offensive rating (which one might think would make him look better) was the very worst for the regular season http://bkref.com/tiny/SPDkF and among the six worst in the playoffs among those who scored 2000+ playoff points.

Somewhat surprising, but maybe it shouldn't be, the very worst career offensive rating for qualified players is another small PG with a high to sometimes very high usage- Tony Parker. 5 of the worst 12 on career playoff offensive rating are PGs but only Iverson didn't get a title. He had the worst eFG% and TS% of any qualified player. 2nd highest usage to Jordan. Parker has the highest playoff usage of any point qualified and title winning PG at a bit under 28% (that is the level I'd probably try not to exceed) Still two or his 3 titles came with his 2nd and 4th lowest usage levels regular season and 2nd and 3rd lowest in playoffs. Only a few other qualified PGs averaged over 20% playoff usage (Magic, Isaiah and Dennis Johnson). Most won titles with a different PG usage strategy.

[colts18]

If Iverson had the same free throw rate in 2001 that he had in 2008, he would have had 208 more FTA (170 more points). His TS% would have jumped from .518 to .535

[Mike G]

Again, the Iverson Effect is:
- A player with [among] the worst shooting% in the rotation, taking the most shots, can apparently have the greatest positive effect on the Team shooting%.
- The guy with worst efficiency (per possession used) can have the greatest impact on point differential.
- The guy with the most turnovers can create the best turnover differential.

Nobody really knows that Iverson could have done anything better, by even one iota. It's hard to go up to shoot and then make a great pass. It's even harder to go up to pass and have to then create a shot. Scoring + passing at this productivity is very rare.

[permaximum]

I think previous posts pretty much covered everything about the so-called Iverson effect. However, I have to mention he played through serious injuries in 2003/04. I wouldn't base my arguments around that season. 2001/02 is an another season he struggled a lot because of injuries although he lead the league in steals and scoring.

As a side note, I always thought RAPM would like Iverson a lot more and somewhat prove the bias against Iverson but it seems it's not the case. His defensive x-RAPM values were definetely the reason. According to J.E.'s x-RAPM he was 18th in 2001/02 and it's the best in his career. No surprise 2001/02 was also the only season that Iverson had a positive defensive x-RAPM value (+0.5). On the offense RAPM likes him although his TS% more or less equals the league average, he jacks up shots and turns the ball over more than an average player.

It only proves you can't judge a player's offense by his individual shooting percent, turnover rate or raw efficiency just like you can't judge a player's defense by his individual steal, block, rebound percents. That's why I think RAPM is the future but it's too noisy.

[Mike G]

Iverson's 2003-04 season was chosen because:
a) he missed a lot of games, so there were a lot of with/without AI lineup minutes available.
b) he had a bad year in boxscore stats. If in a bad year he had a profoundly positive impact on all lineups, his effect may have been even more positive in a good year.

Someone may know how to do these analyses quickly, but it takes me some time. It's just kind of fun for a while.
RAPM is nice, but I still am not sure how 'garbage time' is processed. Using only heavily-used lineups, I assume that's not an issue.

[Mike G]

From J.E. in another thread:

Per basketball-reference, here are his +/-, NET and ORtg-DRtg numbers from '01 to '08

║ Year ║ +/- ║ NET ║ ORtg-DRtg
║ ...
║ 2004 ║ -2 ║ 1 ║ -6 ║

There's obviously the chance that he was often backed up by really good players, but I think very few people would argue that

In the OP, in most commonly used lineups, we see a Net gain of 6 to 11 (per 100) when Iverson is on the floor.
How does that jive with a +/- of -2, or with a NET of +1 ?
At 43 mpg, maybe he played lots of minutes with below-avg teammates?