ESPN/J.E.: Rookies provide no impact at all

Home for all your discussion of basketball statistical analysis.
colts18
Posts: 304
Joined: Fri Aug 31, 2012 1:52 am

ESPN/J.E.: Rookies provide no impact at all

Post by colts18 » Sun Aug 24, 2014 5:14 pm

Interesting article on ESPN. According to it, rookies almost never provide a positive impact. Even rookie LeBron was a negative player.

Highlights:

Image

"As shown, NBA rookies -- even high lottery picks -- usually post starkly negative RPM values. In other words, they are hurting their respective teams more than they are helping. "

"According to RPM, no rookie this millennium -- not even LeBron -- has had a major positive impact on his team."

"So if the Cavaliers want to maximize their odds of returning to the Finals while they have LeBron James under contract -- that is, in the next two seasons -- swapping Wiggins for Love is not only a defensible move -- it's a complete no-brainer.

According to the RPM numbers, the trade should be worth an extra 10 or more wins for Cleveland next season. That's enough of a boost to make the Cavs a legitimate title contender again."

http://espn.go.com/nba/story/_/id/11340 ... den-impact

Dr Positivity
Posts: 298
Joined: Thu Sep 20, 2012 6:44 pm

Re: ESPN/J.E.: Rookies provide no impact at all

Post by Dr Positivity » Sun Aug 24, 2014 5:40 pm

It's probably close to true, though I certainly would not use RPM as evidence for it (But the research J.E. did to build how rookies are evaluated in RPM, is more trustworthy I suppose). Guys like rookie Blake, Lillard, Rose had pretty impressive on/off so I'd be hard pressed to buy they in reality had negative impact

J.E.
Posts: 817
Joined: Fri Apr 15, 2011 8:28 am

Re: ESPN/J.E.: Rookies provide no impact at all

Post by J.E. » Sun Aug 24, 2014 6:03 pm

Dr Positivity wrote:It's probably close to true, though I certainly would not use RPM as evidence for it (But the research J.E. did to build how rookies are evaluated in RPM, is more trustworthy I suppose). Guys like rookie Blake, Lillard, Rose had pretty impressive on/off so I'd be hard pressed to buy they in reality had negative impact
Rose had a -8 NET in his rookie season.
Obviously we're not saying that every rookie that has ever played has had a negative impact in his first season, but chances are that it takes them a while before they can provide positive impact

Dr Positivity
Posts: 298
Joined: Thu Sep 20, 2012 6:44 pm

Re: ESPN/J.E.: Rookies provide no impact at all

Post by Dr Positivity » Sun Aug 24, 2014 6:04 pm

Oops, I misread the +8 as offense

Mike G
Posts: 4413
Joined: Fri Apr 15, 2011 12:02 am
Location: Asheville, NC

Re: ESPN/J.E.: Rookies provide no impact at all

Post by Mike G » Sun Aug 24, 2014 11:04 pm

"As shown, NBA rookies -- even high lottery picks -- usually post starkly negative RPM values. In other words, they are hurting their respective teams more than they are helping. "

"According to RPM, no rookie this millennium -- not even LeBron -- has had a major positive impact on his team."
The 2003 Cavs were the worst team in the league, SRS = -9.6
According to WS/48, Carlos Boozer was their only above-avg player. Ilgauskas was just avg, and everyone else was well below avg.

In 2004, they doubled their wins -- from 17 to 35 -- and their SRS improved to -3.1.
After LeBron, Boozer, and Z, their most-used players were Kevin Ollie, Eric Williams, Ira Newble, and Jeff McInnis: all journeymen near the end of usefulness.
How and by whom did they double their win total and cut 70% off their scoring deficit?

If you have a fever of 103, and your doctor brings it down to 100, has he done "more harm than good"?

xkonk
Posts: 294
Joined: Fri Apr 15, 2011 12:37 am

Re: ESPN/J.E.: Rookies provide no impact at all

Post by xkonk » Mon Aug 25, 2014 3:12 am

The guidelines listed in the article certainly seem a little odd. It says that players at 3 RPM or better are All-Star types, but the rankings (http://espn.go.com/nba/statistics/rpm/_/sort/RPM) say that there were nearly 50 All-Star players last year. The cutoff for a 'starter' is set at 1, but only 116 players made that cut - there are only about 4 starter-quality players per team, on average. I'm willing to believe that there aren't enough "quality starters", in some sense, to fill the league, but that doesn't mean that everyone else is hurting the team more than helping.

Rookie LeBron is listed as a 0 RPM player, or a 'minus starter'. Is average really a minus who hurts his team? Last year the most suitable comparison might be DeMar DeRozan, who was slightly negative (-0.17) but played a ton of minutes (LeBron played over 3100 as a rookie). Just by being about average over a ton of minutes, DeRozan was worth almost 4 WARP, which makes him a top-100 contributor. That seems helpful, and would make him a 'starter' by that 116-player count.

All that aside, it certainly seems true that most rookies are not going to be the final piece of a championship team, and they are not going to be as good (in the short term) as a top-10 type player who is only 25. I'm firmly in the pro-Love camp for the trade.

J.E.
Posts: 817
Joined: Fri Apr 15, 2011 8:28 am

Re: ESPN/J.E.: Rookies provide no impact at all

Post by J.E. » Mon Aug 25, 2014 8:45 am

Mike G wrote:
"As shown, NBA rookies -- even high lottery picks -- usually post starkly negative RPM values. In other words, they are hurting their respective teams more than they are helping. "

"According to RPM, no rookie this millennium -- not even LeBron -- has had a major positive impact on his team."
The 2003 Cavs were the worst team in the league, SRS = -9.6
According to WS/48, Carlos Boozer was their only above-avg player. Ilgauskas was just avg, and everyone else was well below avg.

In 2004, they doubled their wins -- from 17 to 35 -- and their SRS improved to -3.1.
After LeBron, Boozer, and Z, their most-used players were Kevin Ollie, Eric Williams, Ira Newble, and Jeff McInnis: all journeymen near the end of usefulness.
How and by whom did they double their win total and cut 70% off their scoring deficit?

If you have a fever of 103, and your doctor brings it down to 100, has he done "more harm than good"?
Please note that is says "major" in the part that you quote. There's obviously a chance that he was better than his RPM (+/- 0) suggests, but if he was it probably wasn't by much
That year, the Cavs scored better when he was on the court (102 off.eff vs 96) but also defended worse (105 vs 101).
In 844 non-LeBron minutes the team got outscored by 4.7 points per 100, hinting at the fact that the team, even without LeBron, had improved from its -9.6 SRS the year before

gzchen
Posts: 7
Joined: Mon Jun 23, 2014 4:29 pm

Re: ESPN/J.E.: Rookies provide no impact at all

Post by gzchen » Mon Aug 25, 2014 3:46 pm

Mike G wrote:
"As shown, NBA rookies -- even high lottery picks -- usually post starkly negative RPM values. In other words, they are hurting their respective teams more than they are helping. "

"According to RPM, no rookie this millennium -- not even LeBron -- has had a major positive impact on his team."
The 2003 Cavs were the worst team in the league, SRS = -9.6
According to WS/48, Carlos Boozer was their only above-avg player. Ilgauskas was just avg, and everyone else was well below avg.

In 2004, they doubled their wins -- from 17 to 35 -- and their SRS improved to -3.1.
After LeBron, Boozer, and Z, their most-used players were Kevin Ollie, Eric Williams, Ira Newble, and Jeff McInnis: all journeymen near the end of usefulness.
How and by whom did they double their win total and cut 70% off their scoring deficit?

If you have a fever of 103, and your doctor brings it down to 100, has he done "more harm than good"?
Theoretically, it's entirely possible that all else being equal, replacing one player with a second player could make the team better without the second player being an above average player, as all this suggests is that the second player is better than the first player. Saying whether rookie LeBron did "more harm than good" depends on whom he's being compared to. If it were to the player whose minutes he replaced, then you're correct, LeBron did not do more harm than good, but that doesn't mean he was anything better than an average bench player (and as JE commented, the rest of the team also improved so we're not really comparing apples to apples). However, in the case of Wiggins, we're more concerned with how he compares to Kevin Love, and according to JE's article, we should expect Wiggins to be much worse than Love next year.

Mike G
Posts: 4413
Joined: Fri Apr 15, 2011 12:02 am
Location: Asheville, NC

Re: ESPN/J.E.: Rookies provide no impact at all

Post by Mike G » Mon Aug 25, 2014 4:27 pm

The best rookies often go to the worst teams. If one is an average player going many minutes, that is a huge improvement.
The confusion of [average = contributes nothing], or [below avg = worse than nothing] is still alarmingly widespread. Even if you know what you mean, in or out of context, the average reader gets a completely wrong impression from such loose comments.

It's not just "theoretically possible" for another avg player to make a team better, it's about 90% likely. Only exceptions being for truly elite teams.
If the Spurs gave many minutes to an avg player, he would be doing more harm than good to their pt-diff. Actually, that would be the coach's doing, but whatever. It's the equal and opposite effect of the same player on an otherwise terrible team.

This year's Cavs are an atypical case: More than one superstar is joining the team. Therefore, avg players get fewer minutes.

Statman
Posts: 548
Joined: Fri Apr 15, 2011 5:29 pm
Location: Arlington, Texas
Contact:

Re: ESPN/J.E.: Rookies provide no impact at all

Post by Statman » Mon Aug 25, 2014 6:23 pm

colts18 wrote:Interesting article on ESPN. According to it, rookies almost never provide a positive impact. Even rookie LeBron was a negative player.

Highlights:

Image

"As shown, NBA rookies -- even high lottery picks -- usually post starkly negative RPM values. In other words, they are hurting their respective teams more than they are helping. "

"According to RPM, no rookie this millennium -- not even LeBron -- has had a major positive impact on his team."

"So if the Cavaliers want to maximize their odds of returning to the Finals while they have LeBron James under contract -- that is, in the next two seasons -- swapping Wiggins for Love is not only a defensible move -- it's a complete no-brainer.

According to the RPM numbers, the trade should be worth an extra 10 or more wins for Cleveland next season. That's enough of a boost to make the Cavs a legitimate title contender again."

http://espn.go.com/nba/story/_/id/11340 ... den-impact
My work suggests about the same - although I would argue LeBron made a definite positive impact as a rookie, whether it was "major" could be debated.

My current projections for the careers of the incoming college to NBA rookies can be found here: http://hoopsnerd.com/?p=626 . If you look at the excel file and sort by HN/48 and then by year - you can see that only 3 players project to be (barely) above average in terms of production per minute (HN/48 above 100) - and one (McDermott) is barely expected to play at all based on historical statistical comps (assumed very poor defender who comes in occasionally to shoot). Embiid is expected to play a little more, but still litte due to foul proficiency comps. So, only Shabazz Napier projects to be barely above an average NBA player in terms of production as a rookie who actually projects real playing time in '14-15.

mtamada
Posts: 161
Joined: Thu Apr 14, 2011 11:35 pm

Re: ESPN/J.E.: Rookies provide no impact at all

Post by mtamada » Tue Aug 26, 2014 1:57 am

Mike G wrote:The best rookies often go to the worst teams. If one is an average player going many minutes, that is a huge improvement.
The confusion of [average = contributes nothing], or [below avg = worse than nothing] is still alarmingly widespread. Even if you know what you mean, in or out of context, the average reader gets a completely wrong impression from such loose comments.

It's not just "theoretically possible" for another avg player to make a team better, it's about 90% likely.

Yes, I think the original article made a poor choice of words when it said this:
According to RPM, no rookie this millennium -- not even LeBron -- has had a major positive impact on his team.
According to their graph, the average #1 pick has had a rookie RPM of about -1.0. A negative player? No, as MikeG says, they were almost certainly making a major positive contribution to their team, relative to whatever stiff they were replacing.

The article committed a common mistake by mistaking a -1.0 rating for a "negative" contribution. It's only negative if we compare it to the average player. Which is far from the only, or even the most useful, comparison to make. Replacement level is usually a better comparison point, albeit somewhat hard to agree upon (although the dialog on Tom Tango's website seemed to have reached a consensus).

In LeBron James' rookie season, the Cavs won 35 games. If we do calculations the way the article did, the team was -6 compared to the average team's level of 41 wins: a negative outcome! I guess LeBron didn't help them!

The Cavs had won 17 games the year before. No "major positive impact" indeed.

SteveI and J.E. do good work and I'm glad they have that column on ESPN. The -1.0 RPM result is interesting and surprising. But they flubbed on the interpretation and wording of their conclusion.

mystic
Posts: 470
Joined: Mon Apr 18, 2011 10:09 am
Contact:

Re: ESPN/J.E.: Rookies provide no impact at all

Post by mystic » Tue Aug 26, 2014 11:08 am

mtamada wrote: Yes, I think the original article made a poor choice of words when it said this:
Why do you think that? They explicitely stated that the RPM number compares the players to an average player. Thus, it should be easy to understand that James in his rookie season did not make a "major" positive impact above an average player. That's what the statement says, if you read the article correctly and not assume that something else was written like Mike G. did.

In regard to James and the 2004 Cavs: James basically replaced Jurmaine Jones and some of Darius Miles' minutes. Both players had below average RPM values with -1 and -0.8 respectively. James replacing them had technically a positive impact for the Cavs itself, but in comparison to an average player James did not have a "major" positive impact. Those two things do not contradict each other. The article didn't state that James didn't help the Cavs per se, but they made a statement about his contribution in comparison to a specific reference line, which in this case is: average. That shouldn't be so hard to understand.
Mike G wrote:The confusion of [average = contributes nothing], or [below avg = worse than nothing] is still alarmingly widespread.
But that is a reading comprehension issue, not an issue based on some sort of confusion by the authors. If someone can't understand what "above average" means or does not understand the term "average" in the first place, what else is there to do? A team consisting only of average players is expected to win 41 games in a given NBA season! How hard can it be to understand that 41 wins does not equate to "nothing"?

Mike G
Posts: 4413
Joined: Fri Apr 15, 2011 12:02 am
Location: Asheville, NC

Re: ESPN/J.E.: Rookies provide no impact at all

Post by Mike G » Tue Aug 26, 2014 12:03 pm

mtamada wrote:Yes, I think the original article made a poor choice of words when it said this:
According to RPM, no rookie this millennium -- not even LeBron -- has had a major positive impact on his team.
A key word being his team. Not some average team or some other team.

LeBron would have loved to have some more average players everywhere he's played. As would any player who has ever played.

One might as well argue that a rookie (or anyone) who can go 40 mpg and hold his own against NBA veterans, giving as good as he gets, is making huge positive impact. If he's surrounded by great, efficient players, that's a great team. If there are a lot of replacement-level guys on the floor, they're a mediocre team.

mystic
Posts: 470
Joined: Mon Apr 18, 2011 10:09 am
Contact:

Re: ESPN/J.E.: Rookies provide no impact at all

Post by mystic » Tue Aug 26, 2014 12:29 pm

Mike G wrote: The key word being his team.
No, it is not. The key words are RPM and major! If you understand that RPM compares a player's impact to the impact of an average player, you understand that the sentence says that no rookie had a "major" impact on his team "in comparison to an average player". It might confuse some that the word "major" is used here without defining the meaning of that word in the specific context, but that is no different than you saying "making a huge impact", whatever the word "huge" means in that context.

So, in comparison to an average player, was James really that much better in his rookie season? Let us use some other metric like PER and WS/48 here: 18.3 PER and 0.078 WS/48 do not scream "much better than average", in fact, PER had him 0.8 standard deviations (minute weighted standard deviation is used here) above average, while WS/48 had him 0.5 standard deviations below average. Now, RPM says he had the impact of an average player per 100 poss. So, where does it leave us? Did James have a "huge" or "major" impact on his team over an average player per 100 poss in his rookie season? Yes or no?

Mike G
Posts: 4413
Joined: Fri Apr 15, 2011 12:02 am
Location: Asheville, NC

Re: ESPN/J.E.: Rookies provide no impact at all

Post by Mike G » Tue Aug 26, 2014 1:10 pm

For good comprehension among one's readership, one should strive to construct sentences that are self-contained, without internal contradictions.

This sentence preceded the one about LeBron:
"...NBA rookies -- even high lottery picks -- usually post starkly negative RPM values. In other words, they are hurting their respective teams more than they are helping. "
We shouldn't have to interpret that to mean, "...they are hurting their respective teams more than a better player would ." That's such an oblique point that it doesn't even make a lot of sense.

If a player is hurting his team more than he's helping, I understand that to mean his net contributions are negative. I don't see why such a player would stay on the floor.
If he's actually displacing a better player who is on the team, that seems like the coach's fault.

Post Reply