The All-Star game is almost upon us. This means it is time for the media to tell us who they think are the ‘best’ players in the game. And it also means it is time for the people who look at stats to say “sorry, that player is not among the ‘best’”.
So let’s get the season started with a comment on Derrick Rose. Fans and media have fallen in love with Rose this year. The Bulls are winning and Rose – who leads the Bulls in scoring – is clearly their ‘best’ player. Andres Alvarez, though, disagrees. What follows is a column he originally posted at Nerd Numbers, which makes two interesting arguments. Rose may not be the ‘best’ but he may be another Kobe.
Derrick Rose may be the new Kobe. By that I mean he’s a good player that people have definitely gone overboard on. Derrick Rose has been called a lot of things lately and one may question if they are true. The Automated Wins Produced Numbers can squash a few of these rumors. Derrick Rose is not currently in the top ten Wins Produced Players for the season. Derrick Rose is currently listed as the 7th best point guard in the league behind Chris Paul, Steve Nash, Rajon Rondo, Deron Williams, Jason Kidd and Russel Westbrook. It’s somewhat hard to even consider thinking of him as an MVP candidate with that in mind.
DJ at the Wages of Wins Journal sent me a question based on the recent podcast Arturo and I put on; Does Derrick Rose deserve to start in the All-Star game? An interesting thing about how Derrick Rose is perceived is that he has “ramped it up” since the start of the season and “become a leader” in light of the injuries to Boozer and Noah. With that I decided to take a quick look over the season piece by piece to get a slightly better understanding of Derrick Rose using the brand new Game Splits tool.
Table 1: Top Bulls throughout the season in 2 week increments.
At the start of the season Rose wasn’t very good. He improved slightly for about two weeks before going back to merely being an above average point guard. For the first month and half of the season this team belonged to Noah. When Noah went down with injury it was problematic, but indeed Rose did step up his play and has been playing at star level since the start of December. However, it turns out that Boozer returning to the Bulls in top form was a bigger boost to the Bulls. In fact, it has only been in the last month with both Boozer and Noah experiencing injury that Rose is “leading” his team.
Let’s look at Rose in another light going in two week increments. Rose has been listed as an MVP and All-Star candidate. To be an honest option for All-Star I’d expect Rose to be a top 5 guard in the East at minimum. To be an MVP I’d expect Rose to at least be top 10 in the league. How has that looked?
| Date Range | Player | G | GS | MP | WP48 | Wins | MVP | East Guard | Team |
| 10/26 to 11/9 | Derrick Rose | 6 | 6 | 226 | 0.04 | 0.19 | N/A | N/A | 5 |
| 11/10 to 11/24 | Derrick Rose | 7 | 7 | 280 | 0.194 | 1.13 | N/A | 5 | 2 |
| 11/25 to 12/9 | Derrick Rose | 6 | 6 | 230 | 0.129 | 0.62 | N/A | N/A | 4 |
| 12/10 to 12/24 | Derrick Rose | 7 | 7 | 246 | 0.244 | 1.76 | 9 | 1 | 2 |
| 12/25 to 1/8 | Derrick Rose | 8 | 8 | 299 | 0.215 | 1.34 | N/A | 3 | 2 |
| 1/9 to 1/23 | Derrick Rose | 9 | 9 | 355 | 0.211 | 1.56 | N/A | 2 | 1 |
| Total | 43 | 43 | 1636 | 0.1936 | 6.6 | N/A | 4 | 1 |
Table 2: Derrick Rose’s rank as MVP, East Guard and Team Leader.
In December Derrick Rose did indeed crack the ranks of the top players in the NBA as well as play better than all the other guards in the East. Still, I think many would argue one week does not an MVP make. He was still second fiddle to Boozer though. Rose has only enjoyed one span of time as the “best player” on his team and it required the top two players to fall to injury. Rose as MVP and the best player on the Bulls looks a little weak to me.
One thing that doesn’t look so bad to me though is the All-Star possibility for Rose. In the last month and a half he has indeed been playing as a top guard in the East. The only real issue is that the three guards in front of him currently; Wade, Rondo and Fields are much better guards. So while I do consider Rose an elite guard out East I still can’t consider him worthy of starting the All-Star game.
Let’s make sure to end this on a positive note though. Derrick Rose has been playing like a star guard as of late. He is very easily worth his contract and if he keeps up his play will definitely be worth the extension the Bulls are likely to give him. To say Derrick Rose is not the best player in the league or the best guard out East or the best player on his team is not an insult. I am merely pointing out what others have said; that people are overreacting to Derrick Rose’s play right now. Derrick Rose is very talented and young and is a huge part of his team’s success. However, it would be foolish to give him all of the credit or consider him better than he really is.
-Dre
P.S. To make sure Arturo is happy, It should also be noted that the deserving starting point guard out East is playing better than Rose with fewer minutes and just as many team injury concerns on his team. Also Rose isn’t clutch. . .
Wages of Wins Journal 
Chadreius
January 27, 2011
Statistics are useless. Watch a game. Then you understand why he should undoubtedly be in the All-Star game in a starting position.
Tommy_Grand
January 27, 2011
Responses are useless.
Chicago Tim
January 27, 2011
“… if he keeps up his play will definitely be worth the extension the Bulls are likely to give him.”
Thanks, that is what I was most worried about. Just to be sure, are you saying that if Rose gets a max contract he’ll be worth it?
By the way, if you add up wins produced for the season Rose is the MVP for the Bulls. Also, it may be a fluke, but apparently Rose also leads the league in adjusted plus/minus.
http://basketballvalue.com/topplayers.php?year=2010-2011&mode=summary&sortnumber=94&sortorder=DESC
If nothing else, the way that the Bulls play when Rose is on the floor, and the fact that he has been on the floor a great deal, may account for the voting. In contrast, Noah and Boozer have missed significant time with injuries, and so, indeed, has Rondo.
Nerdnumbers
January 27, 2011
Chad,
Dang! Turns out Rose had the best game of all Eastern guards on 12/21/2010. Wade is second on 11/29/2010. Guess we use one game to pick our All-Stars and one week to pick our MVPs. I’ll update my notes. I agree stats are useless, that’s why I cover up the score and game clock when I watch the game.
Tim,
Going over to inconsistent stats to make yourself feel better over Rose? And yes at current level he’s worth any contract the Bulls give him. I agree Rose has been the *single* most valuable player to the Bulls. A lot of that has to do with injuries limiting Boozer and Noah and Brewer getting limited minutes. The point is him as unquestionable leader of the team and MVP candidate is not true. He definitely belongs in the All-Star game. I think he doesn’t deserve starter but I won’t lose sleep over it (4th place getting 2nd place isn’t the end of the world). Now Melo on the other hand. . .
dberri
January 27, 2011
Dre (nerdnumbers),
You are watching games? When did this start?
Actually I have heard that you can do this, but it just seems wrong. Apparently the players just are playing with this round object (called a “bas-ket-ball”) and they hardly even mention any numbers. So I am not sure what this “watching” is all about. Seems like a waste of time.
Tommy_Grand
January 27, 2011
Agree. Rose deserves to be an all star, but not MVP yet.
ilikeflowers
January 27, 2011
The Bulls are in great shape right now. Rose is 23 and almost 0.200. He seems a lock for 0.250 in his prime and could surpass 0.300 at some point. Assuming that Noah increases a little to cover for Boozer’s expected thirtysomething decline this team (if healthy of course) is a serious contender for the title next year and with a little luck might do it this year.
Chicago Tim
January 27, 2011
The talk about Rose for MVP has had five stages this year. First Rose said it was his goal during the preseason and the media scoffed. Then Boozer went down but the Bulls still won and the media took a second look. Then Noah went down and the Bulls still won and the media took a third look. Then Boozer went out for a week with Noah still out, the Bulls won most of their games against weak competition, and the media took a fourth look.
Now comes the backlash. With people like Barkley saying Rose is MVP on national TV, and Rick Reilly writing fluff pieces for ESPN promoting Rose as MVP, some of the stat-conscious writers started taking this campaign seriously, and lately even the mainstream media has had articles saying that although Rose is a serious candidate, he just doesn’t have the stats to back it up.
I know the WoW blog network was way ahead of the game, but you guys are no longer voices in the wilderness. I really don’t think Rose will be MVP unless he improves tremendously between now and the end of the year and the Bulls finish one or two in the East.
Apparently it is too hard for the media to pick which star is responsible for the success of teams like Boston, Miami, the Lakers, and San Antonio. And players for average to mediocre teams (like Kevin Love) have no chance if another candidate from a winning team is available.
I think the MVP race this year comes down to Dwight Howard and Chris Paul, and will probably be decided by the success of their teams, as well as their respective scoring averages. If they both finish third in their respective conferences, Paul may get more credit because of low expectations for his team. If, on the other hand, Orlando finished better in the East than New Orleans does in the West, Howard may get more credit because he has a higher scoring average this year.
ilikeflowers
January 27, 2011
Chad is right, btw. Even when I ignore stats like games won or points scored I too can tell simply by watching his body move that Rose should be MVP. Not only should Rose be NBA MVP he should also be the next People Magazine Sexiest Man Alive. Derrick Rose is dreamy.
Lior
January 27, 2011
Second Chaderiu:
It’s a point I made before on this forum: the All-Star game is not about the best players (the most productive ones). It’s about the most entertaining players (the ones that would be exciting to watch in a game with no defense and emphasis on acrobatic scoring). So I definitely think that the main criteria for ASG selection should be looking at the player and deciding “how excited would I be to watch him play in the ASG?”. I think you’ll agree with me that statistics are pretty useless for answering this particular question.
It’s telling that while the title of the post mentions “start in the all-star game”, the actual post tries to answer the question “is Rose an MVP candidate?”. I think we need to agree to have different criteria for “All-Star” [exciting player to watch] and “MVP” [player who makes the team win].
ilikeflowers
January 27, 2011
Lior, it’s not really about the most entertaining players. For the starters, it’s about the most popular players among those who care enough to vote (I guess you could argue that just gazing upon Yao Ming is entertaining though), for the rest it’s about who the coaches think are most deserving according to whatever criteria they wish to use. Regardless, for the starters your point holds about stats (you’re giving Chad too much credit though in assuming that this was his point) but for the rest it might not hold since it’s quite possible that most coaches are trying to select the ‘actual’ ‘best’ players.
Italian Stallion
January 27, 2011
Most fans, former players, and members of the basketball media are impressed by players that can do things on the court that most other players cannot. Rose does a lot of those things. He’s freaky talent. There’s little doubt about that. Just watch the games. ;-)
However, there is a difference between freaky athletic ability and basketball talent and production on the court. That’s the part that some less sophisticated fans miss. At this stage he is not translating his gifts into production as effectively as he might some day and IMO is not an MVP candidate. However, all I can say is I wish the Knicks had him.
brgulker
January 27, 2011
The Rose/Kobe analogy is perfect. Both are incredibly valuable players to their teams, and their durability — relative to productive but often injured big men on their respective teams — is definitely worth noting.
Daniel Poneman
January 27, 2011
It is clear from this article that you have watched a maximum of 2 or 3 Bulls games this year. Statistics are good for some things, but if you actually WATCH the games (novel concept, I know) it is abundantly clear that Derrick Rose IS the Bulls.
Thomas
January 27, 2011
Rose may well turn out to be the next Kobe. The guy who has a fantastic career, wins a lot of games, wins championships and works hard every summer to improve himself and his team. All the while advanced stats will come up with insane conclusions saying that he’s not as “efficient” or “productive” as a score of lesser players who won’t go on to accomplish a fraction of what he does.
Makes you wonder why anyone even pays attention to these stats. Oh wait, nobody does besides the masturbatory “advanced stat community” that is not in fact growing in size or influence.
entityabyss
January 27, 2011
Thomas, how did you even find this blog? That’s a serious question. I wonder how people who disagree with stat heads find stat sites. It’s interesting to me.
Also, how much of ths articles on this blog have you read? Have you considered that maybe all those players that went from utah to chicago could have made this team improve so much?
Do you know how wins produced works? Is there arguments you have against it?
Once again, this is serious, because I’d like to hear your arguments against it and see how I can explain it. Here’s a smiley to let you know it’s all good. :-)
Jed
January 27, 2011
When opponents make a run, Rose steps up and takes over the game. He scores at the rim, dishes or gets fouled attacking the basket. He’s done it time and time again this season and while the numbers may not be up to par overall, I think it’s important to note the timing of his points, assists, rebounds, steals, and blocks throughout the game.
alvy
January 27, 2011
dre,
nice article, but to suggest rose is not the leader of the chicago bulls on the premises that he is not the most productive player on the team… i’m not sure i can agree with that. even as a third year guard, it seems that he has certainly gained control of the ball club.
the most egalitarian team i can think of would be the boston celtics.
Jared
January 27, 2011
Guys, I’m a math geek, a sports fan, and a real believer in advancing statistical analysis but I think it’s more likely our analysis with respect to guys like Derrick Rose and Carmelo Anthony is imperfect than it is that these guys aren’t great basketball players.
Not to open a philosophical discussion that deserves its own forum but if the analysis tends to go against something we know empirically (in this case, that these gents are great ball players), then we need to go back and refine the analysis.
Like I said: great believer in APBRmetrics and the growth in all of sports analytics but I can’t help but think it’s a tremendous waste of time to throw so much effort into disproving things we know to be true empirically.
Brian
January 27, 2011
I don’t think Jared knows what “true” or empirically” mean.
ilikeflowers
January 27, 2011
Brian,
Add to that a complete ignorance of the specific stat being used (wp48) which says that Rose is already a great player (around @0.200) and based on his age is likely to become one of the top three PG’s in the league in his prime (0.250 – 0.300). Apparently, none of this is great enough.
I’m guessing that this post got a link on ESPN and we’re hearing from some pissed off ‘experts’.
Valas
January 27, 2011
I must say that although I find Prof. Berri´s insights and statistical method (despite the absence of a substantial defensive statistical correction) highly accurate and valuable, I must point out that in short sample instances (such as the playoffs), statistical predictions based on his model have been severely inaccurate. This, to me, suggests that psychological and “leadership” factors can have a very heavy influence in limited number series. The so-called “clutchness” of a player is silly, until you look at actual playoff wins. History is revisionist– yet some players manage to revise it very consistently. Perhaps poise and attitude have a much larger effect in short, high-reward situations than in a long season.
jigga
January 27, 2011
Landry Fields better than Rose. Oh God.
Replace Rose with Fields on the Bulls and see how far they get. It’s called common sense, statheads.
Brian
January 27, 2011
Valas,
Or chance.
daddy
January 27, 2011
you shouldnt be writing about basketball. This is by far one of the worst written articles ive ever read and you obviously dont know what your talking about. landry fields better than drose? ARE YOU INSANE?
Sam Davis
January 27, 2011
@entityabyss – Although you weren’t asking that question to me, I found my way here for the first time 5 minutes ago from a link on RealGM and I’ll never be back. The circle jerk of nerdom is a bit too much.
I also doubt that anyone who matters will find this place either and if they do, they won’t be taking any of it seriously.
ilikeflowers
January 27, 2011
You’ll be missed, Sam.
Justin
January 27, 2011
“Wade, Rondo and Fields are much better guards” … really? You really feel that fields is a better guard. Jesus christ man. Stats are nice and needed to be taken into account but can not be used without also watching the games.(which I’m guessing you don’t do) Fields is a 5th option on his team who doesn’t have to put up with scoring a lot, facing double teams, facilitate the offense, or serve as the only ball handler to his team(no one else outside of watson and noah(sadly) can put the ball on the deck and create). Also, I think Rondo is a top point guard. He is a great floor general and knows how to facilitate but people need to realize not all of his passes are behind the back-no look passes creating an easy opportunity for someone. The guys is surrounded by great shooters and no amount of age(to the big 3) will
Rose is the best player on the Chicago Bulls and you shouldn’t need any stat to know that.(just read my first paragraph to understand the stuff he deals with). Let’s see they play without Boozer = winning record … they play without Noah = winning record. The constant being? You guessed it derrick rose.
tgt
January 27, 2011
@Sam & others from RealGM,
I guess the NBA teams that have offered Professor Berri jobs are not important. Sheesh. Unsupported vitriol and poor arguments? Did this get linked from a creationism site?
koolaid
January 27, 2011
stupidest hoops article i’ve read this year. landry fields better than rose?! stated as an objective fact no less. you just set the advanced stats movement back a decade. don’t drink so much koolaid son.
ilikeflowers
January 27, 2011
koolaid – “you just set the advanced stats movement back a decade”
Justin – “Let’s see they play without Boozer = winning record they play without Noah = winning record. The constant being? You guessed it derrick rose”
daddy – “ARE YOU INSANE?”
jigga – “It’s called common sense, statheads.”
Jared – “we know empirically…in this case, that these gents[Melo and Rose] are great ball players”
Thomas – “the masturbatory ‘advanced stat community’ that is not in fact growing in size or influence.”
Daniel Poneman – “WATCH the games”
Chad – “Statistics are useless. Watch a game.”
Keep em coming guys. Jared and Thomas are winning so far but there’s still time! Be all that you can be!
saadlink
January 27, 2011
entityabyss: Truehoop (on espn.com) posts links to stats blogs and pages a LOT. So its easy for a person who doesn’t care much for stats but sees something like “derrick rose is no landry fields” and says “wtf? I need to check this out.”
Speaking of which: generally I agree stats can be useful. But when I see a statistic that says “Landry Fields is better than Derrick Rose,” I find it hard to think anything but “okay these statistical methods need to be reconfigured.” If I calculate that my lead brick should be flying towards the sky, I don’t doubt the phenomenon that my lead brick fell, I take a second look at my calculations. We’re basically taking the whole “wins added” stat as gospel truth, and again, while I think it’s useful, it isn’t perfect. This is not a random criticism of Berri; unless someone can come up with some one number that takes into account every single variable that goes into one’s success on the basketball court and how they affect the game, the previous sentence is a simple truth. And how true it is becomes patently obvious when we see things like “Landry Fields is a better guard than Derrick Rose.” Unless, of course, you want to say that if we take each off of their respective team, the Knicks will suffer more than the Bulls. Personally, I am very willing to take a bet that the opposite is true.
john
January 27, 2011
relatively new to the advanced stats approach in basketball, so i have one question. Where does Michael Jordan sit on advanced stats measures historically (non Wizards seasons, adjusted for injury year, etc.)? Because if it’s at the top then that kind of correlation would seem to imply that there’s merit in the advanced stats approach.
ilikeflowers
January 27, 2011
saadlink,
The issue is indeed in the presentation here. While Landry Fields has been more productive on a per minute basis, it is still a small sample size so this may not continue to be the case going forward. The real question for me is how well does rookie production from the first half of the season predict their prime years productivity? And just for a spot check who are the players who have posted a 0.250+ wp48 in the first half of their rookie season. This would be telling. If the list is entirely made up of superstars then Landry Fields will almost certainly end up as a superstar, if no then the jury is indeed still out on him.
ilikeflowers
January 27, 2011
prof,
I seem to recall a post on how well rookie production from the first half of the season predicts a player’s prime years. Am I mistaken? Anybody know?
Mike
January 27, 2011
What amazes me most, is that the very people who hate advanced stats stuff harp on about “fundamentals” and “doing the things that don’t show up in the boxscore” and yet, when a metric like WP48 comes along that accounts for EVERYTHING, rebounds, efficiency, defense, people harp back to scoring.
The Bulls, clearly, are a defensive team, and yet everyone talks about Rose, an offensive star. The disconnect is staggering. I mean, no one ever comes in and says “watch a game and you will see that the rotations against the high pick and roll using a small forward to cover the lane and a floating two guard is why they are so effective”. NEVER.
Instead, you hear this obsession with offence, to the point that I wonder if people who claim to watch games watch anything other than the ball. Do they watch defensive rotations? Do they watch if a team shows on pick and rolls? Do they count the number of ugly, inefficient, early in the shot clock shots a player takes?
As a Suns fan, I remember watching Earl Clarke to see why he was playing poorly, and it came down to being late on every defensive rotation, being lazy in transition and not being in the right spot on offence without the ball. None of that is what happens with the ball in his hands.
I’m all for watching games, I love it myself, but I find myself increasingly at odds with the people who use that argument, because all they seem to “watch” is the the ball, and whoever’s hands it is in.
Nerdnumbers
January 27, 2011
Hey guys,
I know I didn’t go into too much depth (in part because this blog and mine have talked a lot about Fields) but I did link the tools I used for this and you can use them tool (queue infomercial music)
http://nerdnumbers.com/automated-wins-produced – Let’s you see regular season totals in Wins Produced and WP48 for last decade including day by day updates of this year!
http://nerdnumbers.com/game-splits – Lets you view game by game analysis of every game this year!
ilikeflowers
January 27, 2011
I’m telling you what’s real. Ulcers plus Basketball means Leader, all day long, and it’s never gonna change. And that’s factorial.
Take that Jared!
Man of Steele
January 27, 2011
john, MJ was excellent. He was the best guard of his generation, although Magic Johnson had a slightly higher WP48 at his prime. Although excellent big men tend to distinguish themselves further from average, Jordan is as many standard deviations above the mean for his position as anyone since 1977. In short, advanced stats say MJ was the best SG in the era for which we have the modern box score, and was about as valuable as any player during that period, with Magic Johnson being the most comparable player.
Jonathan
January 27, 2011
Do you really believe that Kris Humphries is one of the top 15 players in the NBA so far this season? Any system that tells me Kevin Love is the best player in the NBA and that Kris Humphries is in the top 15 might have some problems.
Philip
January 27, 2011
Why is it that the more well-known the player, the more “positive” that people that people who rely on statistics “don’t watch the games” when their evaluations don’t match? It seems like the more exposure a player gets, the more sure that people are that no one who disagrees with them has ever seen that player in action.
Let me put it this way: would we get the same emotional vitriol for a “Ronnie Brewer is better than Corey Brewer” post?
Jonathan
January 27, 2011
No. Because Ronnie Brewer is better than Corey Brewer. Watching the games tells you that.
mrdope
January 27, 2011
This article is stupid.
DRose has been a fantastic player this season. His efficiency has been hit by injuries to core star players, this also effects his assist numbers. Still the Bulls are winning and its because of him, rest assured the Bulls don’t have the 3rd best record in the east right now because of Kieth Bogans, Kyle Korver or Ronnie Brewer. Rose does what he needs to do to pull out the win under circumstances. Stick Rondo or Fields on this injured Bulls team and see how they go.
@ Philip: because Ronnie and Corey Brewer would be reasonably well matched for comparison, DRose and Landry Fields?!?!, GTFO!
Alien Human Hybrid
January 27, 2011
Sad comments. It is disturbing that people judge empirically derived, scientific analysis based upon whether or not it confirms preexisting beliefs.
How far have we come from Ptolemy, Copernicus, Kepler, Galileo and the Inquisition?
Seriously. The world is flat, right? Isn’t that what your eyes tell you?
mrdope
January 27, 2011
@ Alien Human Hybrid
You are a complete moron.
This is not pre-existing beliefs comparable to the the cultural advancements and progress in modern science humans have experienced over the centuries.
It’s Basketball, you nitwit. If you watch it regularly, know the game, perhaps have played the game competitively. It is obviously that the above article is complete drivel.
You don’t even have to know basketball to know your comment is pompous rubbish of the lowest kind and nothing more.
Jeremy
January 27, 2011
Wow, this is possibly the worst basketball related article I have read in a long time. Stats don’t show everything that happens in a game. I would love to hear your starting five for the East according to WP48. Please tell me you have the awesome backcourt of Zach Randolph and Kris Humphries. I can’t wait for Landry Fields to cross over Chris Paul as he makes his mind-blowing “efficient” shot. Even mentioning that Fields could be in the discussion of top guards in the East is beyond ridiculous. WP48 is skewed towards those that do a few things very efficiently like rebound or have high TS%. My understanding that there is no weight to how these measures impact a game. For example, Kevin Love’s rebounds don’t actually translate to wins for the Timberwolves. Someone has to get the ball in hoop off those rebounds. Fields may make the most of his time on the court but his skill sets are limited to needs of his team. He’s not handling the ball, or seeing double teams. Look beyond the spreadsheet and realize that WP48 is skewed.
It’s one thing to argue that Rose doesn’t deserve to be MVP. The truth is no one deserves to be MVP through half the season. That’s why they play 82 games. The starting guard in the NBA all star game is chosen by the fans. The starting five has never been about who deserves it. It’s about the fans. That’s the only way the NBA can get anyone to watch the best marketed pickup game ever played. No one wants to see the most efficient guys in the NBA duel it out for the “All WP48 team.” The All star game is never about who deserves it. An NBA player who cares about winning doesn’t care about their performance in the All star game.
Finally, enjoy your glass tower of skewed stats and BS. I’ll continue to enjoy Rose for his skills, competitiveness, and drive to win. BTW you might be the smartest NBA fan to write something this stupid. You can take that as both a compliment and insult.
Tom King
January 27, 2011
Statistical formulas will never be able to perfectly judge the value of a basketball player. The variables are limitless and will never all be taken into account. Anyone who says Landry Fields is better than Derrick Rose is unqualified to discuss the game of basketball PERIOD. Bring up whatever stats you’d like, they don’t make any difference. Ask Mike D’Antoni who he’d rather have on his team. Ask any GM, scout, talent evaluator, who is the better player. It’s not a debate.
Drew
January 27, 2011
Landry Fields being called better than Derrick Rose is an example of what ails wins produced: it overvalues efficient scoring. That sounds strange to say, but hear me out.
Yes, ideally you want high-efficiency scoring all the time. But the reality is, most high efficiency players are high-efficiency because they play a limited offensive role– which role is the most they can handle. This is why teams don’t just give the ball to guys like Landry Fields and let him do his thing. He can’t handle that, his scoring would be abysmal. But if you give him a limited offensive role, let him “play within himself,” he can be an effective component to team success.
But that does not mean that the Landry Fields of the world are better than the Derrick Roses. Because Derrick Rose is a guy you can just give the ball to and let him do his thing. Wins Produced values shooting efficiency a lot, but it is not like Derrick Rose gets his offense at the expense of more efficient teammates– his less efficient shots are what make the more efficient shots of teammates possible.
Rose draws a double off a screen, opening up Boozer for a clean jumper on a pick and pop. Rose blows by his defender and draws help defense around the hoop, letting him kick it out. Or maybe he doesn’t kick it out, and gets off a tough shot around the hoop (say, one of his floaters) that he misses just a bit more often than you’d want. The fact that he could convert this at a reasonable rate means that there is a threat which the other team must account for, and this is what makes the kickout option possible. If Rose never took those tough floaters, the defense wouldn’t have to respect them, and then there would be less there for the rest of the team.
Those are just examples to illustrate the point I’m trying to make. In a sense, teams do just dump the ball off to their most efficient scorers, and those are guys like Derrick (or Kobe). They may appear less efficient than a Landry Fields-type, but that is because their degree of difficulty is so much higher– and it is their ability to convert even the tough shots that make defenses respect them, and thus opens up the opportunity for other players of lesser abilities (but who can be effective in the right role) to get into their offensive game.
Really, this is just the mirror image of PER. PER rewards players just for being volume shooters, since it does not take much in the way of efficiency to gain positive PER per shot. Wins produced punishes a player for being a volume shooter, because these players tend to convert less efficiently. Often, this is well-deserved: there are definitely chuckers out there who hurt their teams by just wanting to “get their shots off.” But sometimes volume shooters are actually helping their team: their high-difficulty shots make the more efficient shots of teammates possible.
How can you tell which is which? Statistically, I’m not sure (but I bet there is fame and fortune to the person who finds the answer). Maybe something like adjusted +/- would help here. But you can tell pretty well by, as they say, watching the games. Of course, that involves subjectivity and it is time-intensive. But if enough informed observers agree than you can feel pretty confident in their opinions. And I think Rose has reached that point, or is getting there anyway— like Kobe before him.
arturogalletti
January 27, 2011
Ah, I love this kind of post. Hello world and all that.
If you’re interested in the science behind the article (and I’m assuming you might be since you’re reading this far down) go here:
http://arturogalletti.wordpress.com/2010/10/05/the-basics/
One general note, did you watch Knicks beat the Heat tonight? Who do you think the Garden crowd thought was the most important Knick in that comeback win? (I’ll give you one hint: he wears number 6 and Spike Lee and Charles Barkley love watching him play ball)
Some answers.
Ilikeflowers. I put up the list of rookies with >.250 WP48 here:
http://arturogalletti.wordpress.com/2011/01/17/fanservice-followup-notes-on-melo-rookies/
Griffin and Fields project between Bird,Marques Johnson,CP3,Larry Smith,Hakeem and Sabonis. Pretty awesome crowd.
As for the best of all time according to Wins Produced (since 1978):
http://arturogalletti.wordpress.com/2010/12/27/the-mvp-race-so-far/
Barkley,Stockton,Magic,Malone, Jordan are the top five with Garnett and Kidd knocking on the door and Larry Legend,Hakeem and Shaq rounding out the top 10. That’s total wins. That meet everyone’s standards?
Italian Stallion
January 27, 2011
Drew,
IMHO, you are pointing out some of the issues with giving TOO MUCH weight to efficiency and not enough to the actual scoring volume/ability. But I think as many people do you may be overestimating the value in the other direction.
IMO many players have at least some flexibility in their usage without much impact to efficiency.
If each of 4 low usage high efficiency players up their usage 1-2 shots per game, they can make up for a lot of the scoring of the high usage low efficiency player. In addition, if that high usage player is really creating a lot of value drawing double teams, he will accrue some value via his assists.
Personally, I am in the middle on this one.
I think most people vastly overrate wildly athletic high usage scoring, but I think these high usage scorers are often adding value that is not captured via efficiency alone.
Alien Human Hybrid
January 27, 2011
@ Mr. Dope-
Something tells me millions of people watched the sun rise and set for millennia, all the while believing the sun went around the Earth.
Just like everyone “knows” “x” about basketball, mainly because they’ve watched a lot. Or played a lot. Or even just a little.
Parenthetically, I’ve noticed what appears to be an inverse relationship between frequency of insult and frequency of insight.
Italian Stallion
January 27, 2011
I recently read some stuff about basketball written by Haralabob. While he didn’t give away much in terms of how he value players or create lines, he did say some things that were interesting to me as a gambler.
I get into endless debates on these issues. Sometimes it gets quite heated.
Half the time I’m arguing with Knicks fans that think the Knicks HAVE to get MELO because he’s such a great player etc… and half the time I’m arguing Melo is better than being give credit for in the advanced stats community.
It’s really the same debate as this one about Rose.
I think perhaps I am wasting a lot of time. I have yet to convince anyone on either side of anything. So what’s the point?
After about 40 years as a fan and 3-4 years of more serious study of the game, this year I began testing my beliefs against the line makers in Vegas.
I’m with Haralabob on this. Trying to beat Vegas is pretty much the ultimate test of your model and thinking.
If you can beat Vegas, you definitely know a lot and if you can’t you don’t.
Mike
January 27, 2011
> Please tell me you have the awesome backcourt of Zach Randolph and Kris Humphries.
Um, Backcourt?
> and it is their ability to convert even the tough shots that make defenses respect them, and thus opens up the opportunity for other players of lesser abilities (but who can be effective in the right role) to get into their offensive game.
Really? So teams practice, they have plays they design, and they pay coaches all this money and it all comes down to give it to your best offensive player?
The mistake people make is to assume that inefficiency is caused by trying to do more in a GOOD way. A contested 18 footer with 18 seconds on the shot clock is a bad shot, no matter who takes it. Sure, Kobe might hit 38% of those, and some scrub 32%, but so what? Don’t take that shot. As in… NEVER.
Jordan shot over 50% for his career. He took efficient shots. Ditto Kareem and any number of greats, INCLUDING Lebron.
My rebuttal would be to compare the famous “Battier is MVP” game based upon what Battier knows: http://www.nytimes.com/2009/02/15/magazine/15Battier-t.html vs what “Common wisdom” from Hollinger says: http://sports.espn.go.com/nba/dailydime?page=dime-090114
Lets look at this telling part(s). Hollinger claims:
“The difference-maker, in the end, was Bryant’s presence”
But was it? The game hinged on a long Kobe three. From the NY Times (last page: http://www.nytimes.com/2009/02/15/magazine/15Battier-t.html?pagewanted=8&_r=1)
Lets quote that again:
So that was an AWFUL shot. But, 13.7% of the time he gets it in.
Common wisdom is that Kobe is just “clutch” and, as Hollinger writes, “… took over the game in the fourth”. Statistical analysis, however, says he took a bad shot that he will, on rare occasion, make.
I give Batter the last word on that long three:
Italian Stallion
January 27, 2011
Arturo,
Fields picked a great game to have his lifetime best effort.
At MSG, against the HATED Heat, on TNT.
I’ll bet you there are conversations going on in Denver right now about who they might want in a Melo trade. I’m just hoping Walsh and D’Antoni understand how valuable this kid could become if he expands his arsenal on offense.
Franklin
January 27, 2011
I feel like this stat short changes rose because of the unique way he plays. He is able to get to the rim at will, but until recently he avoided or at least appeared to avoid contact on the drives and consequently wasn’t getting foul calls.
The help defender tries to disrupt or block the shot and when Rose misses the acrobatic layup. Noah/Deng/Boozer are able to get the easy rebound and score because the help sacrificed rebounding position.
Its really as good as an assist, but it counts as missed shot. I know this same thing happens to most players from time to time, but I think you see it happen with Rose fairly regularly because he drives so much and until recently (with in the last month) hasn’t been getting calls.
arturogalletti
January 27, 2011
IS,
Much like Charles Barkley, I love watching that kid play ball. Fantastic vision and play without the ball. He was seemingly involved in every key play down the stretch for that game (doing something positive for his team). Totally owned the moment.
I really think he and the other young Knicks will be fantastic in the playoffs and a rough,rough out.
marparker
January 27, 2011
This could be a great discussion. WoW challenges alot of pre conceived notions. Its not perfect but I’m going to throw out there that I think it is right on over 80 percent of player comparisons. I think certain players can end up with really easy roles where their weaknesses are completely covered by the teammates. I also think players can end up on teams where they have alot of weaknesses but their teammates are even weaker. These are the guys who change teams and their wp48 plummets or have drastically different wp48 from season to season.
All in all from what I’ve seen guards are pretty much what wp48 says they are. SF, PF, and C are much more dependent players. When these guys play with above average guards there FG% skew higher. When they play with below average guards fg% drops but net posessions increase. So, we have to be keenly aware of the situation big men are coming from.
That being said Landry Fields is a guard on an NBA team. He may not shoot as much as your average shooting guard but he has plenty of opportunities to handle the ball and doesn’t seem to screw up. Maybe there is an advantage to maximizing shot opportunities vs. more difficult shots in a higher volume. The Bulls have the 21st best offense in the league and the 3rd best defense and the 6th best record. Maybe just maybe they are winning despite their offense not because of it. On the other hand Fields is the most efficient guard on a top 10 offense. Maybe just maybe he is actually more important to team success.
Of course not many around here are willing to support their argument with coherent independent thoughts. So we are stuck with the assumption that the more athletic player is automatically the better player.
Philip
January 27, 2011
Mike,
Fantastic post. You hit on a great post – that most fans only watch the ball.
Drew,
You talked about some of the situations where Rose is better – mostly isos or running the pick-and-roll. I don’t disagree with your assessment.
But who would you rather on your team when someone else drives? Fields is a better decision maker and a better scorer and scorer in these situations. You can tell this “by watching the games”. Or you can take a look at the box score and see that Fields is a better shooter and a turns the ball over less.
Who would you rather have when the shot goes up? Fields is a much better rebounder any way you slice it. You can watch him rebound. Or you can take a look at the box score and see that he’s a drastically better rebounder than Rose.
Maybe the issue is that people overvalue unassisted scoring. Think you can’t be good offensively without someone creating shots? Look at the Jazz. They’ve had exactly two players that excel at “creating their own shot”, and they’ve been good offensively for years. Part of that is due to Boozer and Williams. But it’s also due to efficient scoring from players like Korver and Brewer who.
Italian Stallion,
A great point that Kevin Pelton hit on regarding Melo was that although Melo has only average efficiency, he actually has a low turnover rate when you factor in his usage. So his points per possession (ppp) is actually more impressive than his pedestrian TS% belies.
When players increase their usage, efficiency generally remains stable. But their turnovers almost invariably go up. Perhaps Melo helps his teams not by increasing their efficiency, but by helping them keep their turnovers down.
Ben
January 28, 2011
so youre saying it would have been better for the bulls if derrick was hurt now instead of Noah, or if derrick missed the first 2 months instead of boozer? derrick’s impact cannot be fully evaluated by stats. the fact that you dont acknowledge that is why so many people think you dont watch basketball
EntityAbyss
January 28, 2011
It’s a lot of people here with some valid points. I will say this, however.
It’s not always the case that what you see is automatically the truth. Yes, I believe Lebron James is better than Luol Deng (which most stats support), but I don’t think it’s smart to just dismiss a stat because it tells you something that you don’t agree with. What you believe is the case may not be the case. Looking at wins produced, I didn’t just come to this blog saying it’s right because it correlates with wins or it’s wrong because it says that carmelo anthony is overrated.
Of course, when I came on to this blog, I had doubts, but as a fan of basketball, check out the information on the blog, rather than dismiss it because it goes against what you believe to be true.
In the case of Allen Iverson, most people “knew just by watching the games” that he was better than Andre Miller. When they were traded for each other, the sixers got waaaay better and the nuggets got worse. Most people then “knew” that Allen Iverson was better than Chauncey Billups, but once again, after the trade was made, Detroit got worse.
Most people “knew” that Carlos Boozer was productive because of Deron Williams and that D Will would make Al Jefferson just as good, or Al Jefferson would get those numbers in that “system”. As you can see, the jazz are struggling. Most People “knew” that the Spurs were getting too old to contend… so many examples.
What’s my point? Do what I had a hard time doing, but did because I wanted to learn more about basketball. Instead of dismissing the statistic because the results don’t go with what you believe, check it out. Check out the blog. There’s a faq page. There’s articles for many questions, such as usage, shot creation, the effect of changing teams. There’s a network of blogs connected to this with many articles.
At the very least, ask questions. Ask why the wins produced metric says that Landry Fields is better than derrick rose. Ask why it doesn’t account for Derrick Rose’s usage. Don’t just dismiss it. Please someone, ask a question or search the blog.
Josh Party
January 28, 2011
Obviously a internet troll, stop feeding him.
EvanZ
January 28, 2011
There are plenty of disagreements on Rose. I have him rated fourth (behind Paul, Nash, and Rondo) by ezPM, and Adjusted +/- has him ranked first, slightly ahead of Paul. Truth be told, the errors in all of these models (both statistical and semantical) are large enough, that I would have a hard time defending any one particular point guard as “the best”.
I should say, any one player, not named Chris Paul.
Man of Steele
January 28, 2011
marparker, you’re right. People tend to use the draft version of the eyeball test: for the draft, teams watch tape to figure which players have the best physical skills, which is somewhat appropriate when you’re talking about 19-22 year-olds. However, this “metric” is a measure of potential (at the most), not of productivity. Using the same measure to compare NBA players against one another is problematic. I think IS said “people overvalue wildly athletic high-volume scoring.”
Ben, “watching the games actually cannot support your argument. We haven’t seen Noah and Boozer together for more than a week and a half, so we can’t know if the Bulls would be better off with the two of them healthy and Rose injured. It’s okay to speculate, but we don’t actually have the data to demonstrate the accuracy of that particular speculation.
Will
January 28, 2011
(English’s not my first language and I am not a stat person, bear with me if you find wrong grammar and wrong takes on WP48)
Not convinced. I read the FAQs and am not convinced. Call me a moron that believes the sun goes around the Earth, but I am not convinced. In this case, the stats probably did not tell the entire story. WP48 highly values efficiency, thus the more efficient player looks better under WP48. But can we say that the more efficient player is always better? I don’t think so. I think what WP48 tells you that which player is more efficient more than which player is better skilled. Granted, most great players are efficient, but not all efficient players are great players. If based on efficiency alone, you concluded that player is great (all star), then it is illogical, as it is outside of the deterministic value of that stat.
Problem is, each team has a different situation. Fields is way more efficient because and only because he is in a different situation which allows him to be so. I believe basketball is not played in a vacuum, and if you take into account the load on Derrick Rose, you know why he has to take inefficient shots.
Some will (and did in the above) argue that inefficient shots shall never be taken. But if Rose pass up the inefficient shots, it is either his teamates throw up a worse shot or no shots at all! Yes, if he pass up that shot, then probably his efficiency would be higher but his team will undoubtedly suffer. You can definitely blame him for inability to create a better look, but his inefficiency is not mostly his fault. Part of the onus is on his teammates and the coach. They share blame.
And look at Fields. He is the 4th or 5th option on that team. Looks are created for him. 71% shots of him is assisted. Of course he is efficient. His efficiency is not entirely due to his merits, as in Rose’s case. Granted, Fields’ efficiency is partly because he plays within himself, but a larger part is because is role is on the receiving end, not the creator. He also is a good rebounder, but that is because he play a different position and his team is weaker at rebounding.
As for defence, I don’t understand how that affects WP48. But what I do know is Rose is a good defender this season, shown by other advanced stats, I believe the discrepency is not due to his defence.
Basketball is not played in a vacuum, teams are different. You cannot just put stats into formula and think that is be-all, end-all. On the other hand, you also cannot just watch the game and think your eyes tell you everything you need to know. It is probably somewhere in the middle.
As in the case of Kevin Durant, I think this is where flaw of advanced stats is shown. Instead of trying to say Fields is all star, maybe you should revise the system to make presented like what he really is : a role player, an extremely good one at that, even maybe a starter on championship type player.
You will say I am being decieved by my eyes, but I do think the truth probably lies somewhere in between of fancy stats and naked eyes. Both need to adjust. Once again, to use the Sun-Earth example, Corpernicus was right about Suns-Earth thing, but his calculations were way off. Everyone thought Newton is dead on, but then Einstein came, and then quantum mechanics made Einstein look bad. Every good theory has flaws.
(BTW, as a scientist, I am not convinced of treating economics as science. Economics are more metaphysics than science, is often inaccurate. Economists sometimes cannot even explain the economy itself, I doubt the accuracy of economics when put on to a volatile subject like sports.)
Keep up the work, I enjoy reading this blog.
Italian Stallion
January 28, 2011
I’d like to elaborate on something I said in a previous blog entry.
I think the ability to score has varying value depending on the make up of the team.
IMO Rose’s ability to score is very valuable to the Bulls because they don’t have a lot of scorers (especially when Boozer is out). However, if you put Rose on a team with some efficient/high usage scorers and he was putting up the same exact numbers, his scoring wouldn’t be nearly as valuable. In fact, I could imagine a situation where the best thing Rose could do was not try to score as much because he was hurting the team.
It’s a lot tougher for me to imagine situations where getting fewer rebounds, blocks, assists, reducing turnovers etc.. could be good. But the value of shooting/scoring seems to be more variable.
After 3-4 years of giving this whole scoring/efficiency a lot of thought and reading the studies of people more gifted in mathematics, I have become somewhat convinced that the reason it’s so tough to come to a consensus on the value of scoring is because the value is a variable depending on the circumstances.
Michael
January 28, 2011
So to clarify, the second best point guard in the east according to wins produced is selected to start in the all star game? Heaven forbid!
Christian Sorensen
January 28, 2011
Great post, and even though I winced through it as a hardcore Bulls fan, I love Landry Fields and its great to see him be the “unheralded” guy that is statistically beating up our Poohdini.
Three things: even though your blog was swamped with “YOUR A MORON” commenters, maybe they learned something – advanced statistics in basketball are at a different place than baseball, but its good to see the different models for evaluating games – what would happen if you put Landry Fields on the Bulls in place of Keith Bogans? If that increases DRose’s efficiency by creating more assists/opening the lane for more efficient shots, then how does the debate change? Adjusting for “tempo” doesn’t do enough to adjust for the fact that team philosophies are placing these players into specific structured shots/opportunities.
That seems to be the biggest hurdle for advanced stats in basketball versus baseball – there’s only one way to score in baseball, but in basketball you don’t have such static scoring opportunities, you not only have to account for the ability of the player and the training ability of the coach, but there are many more “style” or “strategy” options open to basketball coaches that may skew the player’s statistics. As someone who had to sit through Vinny Del Negro, poorly designed offenses can’t possibly positively contribute to a player’s efficiency measures, but they undoubtedly have negative effects.
Italian Stallion
January 28, 2011
Here’s an interesting question:
Assume an incredibly simple model with no turnovers and free throws to complicate matters too much. We have two players.
One takes 22 shots and scores 23.5 points
One takes 9 shots and scores 11 points.
If we remove the player that takes 22 shots, could the rest of the players (including an average substitute) score an extra 12.5 points on 13 shots?
Naturally “it depends on who the other players are”, but I’d be willing to bet in many cases they would get very close and in some they would make it all up or even exceed it.
EvanZ
January 28, 2011
“but I’d be willing to bet in many cases they would get very close and in some they would make it all up or even exceed it.”
Of course, it will. But when even 1 point on average can mean an extra 2.5 wins over the course of a season, it’s extremely important to get the details right. In this case, we need to know, for example, how many possessions were used. We need to know how many free throw attempts the player got, how many turnovers the player had, and how many his teammates have on average. A point here or there may not seem like much in this one example, but it makes all the difference over the course of a season.
As always, we can look to Einstein, who said, “Make everything as simple as possible, but not simpler.”
Wilhelm
January 28, 2011
This guy couldn’t tell if it was daytime unless he looked at his watch — and if it turns out his watch wasn’t working and the sky was black, he’d still refer to his watch to determine if it was day or night.
It really doesn’t take much to look at these numbers and compare them. There’s no real critical thinking here, just retrieval of data and surface level interpretation. Let’s not pretend this is especially intelligent or insightful. That he can draw conclusions that are so absurd (Bogans contributing more wins than Rose in the first two weeks of the year? Hilarious.) and not even challenge them or question their validity is just intellectually lazy. A disappointing piece to say the least.
Philip
January 28, 2011
EntityAbyss,
Fantastic post.
Will,
What do you mean when you talk about skills? The ability to use a dribble in a spectacular fasion? Or make difficult shots? Or throw fancy passes? I look as those as aesthetic skills. From a purely aesthetic, I enjoy watching them. I’m guessing you do too. They can help a player be better – for instance if that spectacular dribbling and difficult shot making leads to scoring more efficiently, or if those fancy passes lead to assists.
We talk about what Rose and others do as skills. But do we credit Fields with skills related to what he doesn’t do? He doesn’t take difficult shots doesn’t make risky passes, and in general doesn’t rely on highlight-reel type displays of athleticism or “skill”.
Ultimately, it’s the results of what a player does that matter, not how visually pleasing their “skills” are. To draw a paralel with baseball, it took decades for the ability to draw walks to be recognized as a skill; even today, it’s almost certainly undervalued by most fans. The same is true today for basketball for almost everything other than scoring and what shows up on Sportscenter.
(And your English is fine; I had no problem understanding you).
Man of Steele,
I think that Chicago is better off with Boozer/Noah injured than with Rose, but that’s because they’ve got decent backup bigs in Thomas and Gibson. Rose is backed up by CJ Miles. Yeeesh.
frankdatank
January 28, 2011
@ Drew–It’s called an R^2
While DRose is obviously a much better basketball player than Fields, etc, “efficient” players maybe have a better grasp of their limitations. If coaches are the real drivers of volume shooting (as some have suggested is the case w/ DRose), it does not change the fact that DRose has many 7-25 shooting nights.
Scott A
January 28, 2011
Look, Rose may be a bit overvalued at the moment. Rick Reilly columns are like a signaling mechanism for national overreaction about an athlete (see: Cutler, Jay). But to seriously claim that Landry Fields is objectively better then Derrick Rose is insane.
I love advanced statistics, they’re incredibly useful, but things like this make me wonder if they undervalue the main thing sportswriters overvalue: the ability to generate offense, even at suboptimal efficiencies. Landry Fields is efficient because he doesn’t have to take many risks; he hits open shots, takes advantage of defenses playing towards Amare and Felton, and rebounds aggressively. I don’t think risk-taking is well incorporated into advanced stats (I don’t think usage adequately covers it); major shot creators absorb a lot of risk, particularly taking shots in terrible game situations, and keep their efficiency at solid levels. Anyways, don’t know where I’m going with this; just something to think about.
entityabyss
January 28, 2011
Great, I see some arguments. Now we’re getting somewhere. I found out about this blog last year and have heard just about every argument, and it has been my goal to see every argument and not dismiss it.
Before I start, everyone should have knowledge of efficiency differential (aka point differential) which is essentially points scored – points allowed per 100 possessions. It has a correlation of about 95% with actual wins. For instance, the 2006 miami heat had a point differential of a team expected to win 50.1 games and they won 50 games that year. The main things that affect point differential at the team level are scoring efficiency and possessions factors. Your teams scores efficiently (however that be), doesn’t turn over the ball, rebounds well, and forces misses and turnovers, you’ll win more basketball games. Simple stuff.
The main argument I’ve been seeing is that wins produced values efficiency and therefore players with a bigger load suffer. This is a valid argument (one I like to hear). The thing about this is, how do you know for sure that if a player had a lower load, he’d shoot more efficiently. This is common perception, but may not be true. Lebron james, kevin durant, michael jordan, magic, and bird were all efficient with big loads. This year CP3, steve nash, and d will are way more efficient than d rose while carrying big loads. The question becomes, how much does the increased load effect derrick rose. For that, read this https://dberri.wordpress.com/2006/06/11/the-law-of-diminishing-returns-in-the-nba/
That’s a great post that goes over that. Any other arguments you have, or any I missed, just say them.
Once again, if you feel that derrick rose is not so efficient because he has a greater load, check out this post https://dberri.wordpress.com/2006/06/11/the-law-of-diminishing-returns-in-the-nba/
Italian Stallion
January 28, 2011
Evan,
The two players are Fields and Rose.
I understand what you are saying. I rounded things off to make the mathematics simple and still be able to make my point about scoring, TS%, and how low efficiency scoring has less value than perceptions.
I don’t think breaking out the FT attempts would change the point I am making much because I added them into the shot attempts (not perfectly, but in a rounded off way). If we added in turnovers, I think if anything it would improve the way Fields looks.
The thing most fans miss is that all else being equal, possessions don’t much, if at all, when you change players. So when you add or subtract a high usage scorer you don’t change the results nearly as much as people think. When you are talking about a low usage very efficient guy and a high usage lower efficiency guy IMO there’s not much difference in their scoring value at all. The thing that makes one better than the other will often be rebounds, turnovers, blocks, steals, defense etc… and not the scoring difference that everyone focuses their attention on.
EvanZ
January 28, 2011
“So when you add or subtract a high usage scorer you don’t change the results nearly as much as people think. ”
Landry Fields is highly efficient, but has extremely low usage (13%). You put Fields on that Bulls club, instead of Rose (31% usage), and now you’ve got to make up 18% points. Boozer is already at 28%. Now, if you were a coach, wouldn’t you go to Boozer as much as you possibly could? Of course. And he’s already at a very high usage. He probably could only take another 1-2 % points, which would amount to a couple more shots. That still leaves 15% points between Deng, Fields, Noah, and Gibson. I’d love to see Fields and Noah split those extra 15 points and stay at their same high efficiency, but I have a feeling Deng and Gibson would inevitably get some of those shots, too. And both of those guys are under 0.500 TS.
The point is that for your argument to have merit, you have to assume that there is *no* change in efficiency while nearly doubling a guy’s usage (in this case Fields). If that were really the case, don’t you think D’Antoni would have realized it long ago? I mean, really, why is it that Fields’ usage is 13% and not 30%? Just random?
(Why doesn’t James Jones take 25% of his team’s shots, not 15%? He’s so efficient, Miami would never lose, right?)
Italian Stallion
January 28, 2011
Scott,
>Landry Fields is efficient because he doesn’t have to take many risks<
That's partially true, but it's not as easy to find a low usage efficient wing player player like Fields that also rebounds like him as people think. It's clear he has a lot of value whether you use box scores or adjusted +/-.
It's also a lot easier to replace a high usage low efficiency player than people think even if you have Fields on the court because the extra shots don't all have to go to Fields. They can be divided among 5 players with minimal impact on the efficiency of each.
What you can't have is 5 guys like Fields on the court at the same time and expect to score enough. But you can't win with 5 Cs or 5 PGs either and everyone seems to grasp that even though as individuals they could all be great players.
People seem to understand diminishing returns when it comes to rebounds, but scoring has the greatest level of diminishing returns of pretty much everything we discuss.
I'm not extreme in my views on this issue.
In fact, I think scoring has more or less value than this model suggests at the extremes of usage, but it's obvious that scoring is massively overvalued by the fans and the media and perhaps even overvalued by many in the advanced stats community.
That's the way I bet my money with good success.
Italian Stallion
January 28, 2011
>>The point is that for your argument to have merit, you have to assume that there is *no* change in efficiency while nearly doubling a guy’s usage (in this case Fields). <<
I never said we could double Fields' usage. I said the difference between Fields' usage and Rose's usage could be divided by 5 with less impact than many would think .
One of the main points I have been making is that the value of scoring is a variable that depends on the makeup of the rest of the team.
The Bulls are not an offensive powerhouse. So substituting Fields for Rose would be problematical, but IMO even there it wouldn't be a total disaster the way a casual glance would suggest. In other situations, substituting Fields for Rose might actually make the team better.
There is more to this game than positional adjustments.
Sometimes the skill sets of players that play similar or even the same position are dramatically different. Then it's up to the coaches and GMs to mix and match the players to get maximize the value out of the whole, but the value of the whole is not perfectly correlated to the value of the individuals.
IMO Fields has a lot of value. It's just a different type of value than Rose and it doesn't belong on the Bulls. Rose does.
entityabyss
January 28, 2011
Evans, you should’ve checked the link I sent above. Maybe you did. Maybe not.
Anyways, scoring efficiency seems to be about shot selection, and ability, not amount. WP does account for that. Once again, I’ll use allen iverson.
When allen iverson was traded to the nuggets for andre miller, people said although allen iverson was inefficient, there was no way miller could take those shots and that someone was needed to take those shots.
What happened was the entire team slightly increased their shot amount without a noticeable decline in scoring efficiency. How does this apply to derrick rose and landry fields? Well, if landry fields replaced rose, he wouldn’t take all those extra shots. The team would. Although landry fields is efficient, he’s not a high volume scorer, so his WP is not highly dependant on his scoring, but also his rebounding and lack of turnovers.
Everyone will increase their shots, but not too much and probably won’t see a decline in scoring efficiency. Like in the link I sent (once again https://dberri.wordpress.com/2006/06/11/the-law-of-diminishing-returns-in-the-nba/ ) scoring efficiency doesn’t decline when you take more shots. It’s more dependent on shot selection and ability. Landry fields could probably take one more shot a game and not see a change in his efficiency.
Does that mean that all scoring is irrevelant because it can easily be replaced? No. The average ts% in the league is 54.5%. A ts% below 50% would be hurting your team. When a player shoots, he’s taking a shot away from another player, but if his ts% is above 50%, he’s helping. Unfortunately though, if it’s below 54.5%, his scoring is below average. So a kevin durant (ts% around 60%) is taking shots away from his teammates, but it’s a good thing. Derrick rose’s scoring helps too. Just not that much. Landry fields’ scoring helps the knicks, but he doesn’t score much, so not that much either.
I hope that explains things well.
Adam C. Morrison
January 28, 2011
The main reason I cannot take any of the trolls recently is because they are not specific in their criticisms. The FAQ is quite thorough in deconstructing everything you are saying. For example:
“Some people believe that the NBA box score doesn’t do a very good job of capturing a player’s contribution to team success. In Stumbling on Wins, we express disagreement with this sentiment. Here is what we specifically said in Appendix A:
Statistics from football tend to be inconsistent. This suggests a player’s numbers are influenced by his teammates. Although it’s suspected this is true in the NBA, the consistency of performance across time suggests that teammates don’t have much impact on an individual player’s productivity. Consequently, it seems safe to assume that the statistics tracked for an individual player represent that player’s contribution to team success. ”
It’s just a defensive reaction from people who are not being honest with themselves. imo
MKSE
January 28, 2011
I came here from ESPN, and I frequent RealGM and a few other Bulls blogs…just to let you know, you guys are getting hammered everywhere. Not the “Rose isn’t the MVP and may not be starter” thing, because that’s debatable, but saying “Fields > Rose” really set people off…
Chicago Tim
January 28, 2011
Will — I am not an economist or scientist, but I do know that one reason economists turn to sports is that it is simpler than the broader economy. And one reason Berri focuses on basketball is that it is more predictable than other sports, due to the short supply of athletic tall people. Baseball can be predictable during the season but becomes a total crap shoot in the playoffs. Football is pretty much a crapshoot every year. By comparison, basketball is not very volatile — absent some variances due to injury, we can pretty much predict which teams will do well, which will not, and which will lag in the middle.
Italian Stallion
January 28, 2011
>but saying “Fields > Rose” really set people off…<
IMHO, Fields is not better than Rose, but the the critical point is that they have two entirely different skill sets.
Rose's skill set is the one that tends to be massively overrated by fans, former players, the sports media etc…
Fields' skill set is the one that is almost universally underrated by the same groups. However, Fields has been so impressive at what does that even the mainstream can appreciate it.
On a net basis, the gap is not nearly as large as the mainstream thinks even if Rose is better.
Sam
January 28, 2011
I like the Rose/Kobe comparison. Both fantastic athletes with very high skill levels, but both obsessed with putting up shots. Russ Westbrook has the same problem.
They still end up as well above-average players but they waste away too many possessions taking bad shots instead of executing offensively and helping their teammates get better ones.
Bulls Fan in Krakow
January 29, 2011
There are several concepts being argued over in this post, and it’s easy for an objective person to see both sides of the matter: that statistics are useful when used in context, and can often uncover very interesting dynamics, while one must also watch the game and analyze the volume of the player’s work on the court, and not just in a spreadsheet, in order to have the full picture.
I’ve read many defenders of this blog and this post state that Bulls fans and Derrick Rose defenders need to challenge their beliefs because these statistics paint a different picture. However, as a fan of the Bulls, I think that the defenders of statistics must do this also.
For example, as was stated in a comment above, if these statistics state that Keith Bogans has added more wins to the Chicago Bulls at any time this season, then it is simply intellectually lazy not to challenge the statistic and look into why it might have arrived at that particular result. Joakim Noah, Carlos Boozer, and Luol Deng are the only Bulls that have been more valuable than Derrick Rose for more than one quarter this season. Even Ronnie Brewer, as much as I like him, hasn’t been more responsible than Rose for wins produced (in the non-statistical sense) at any time.
As for the nuances of these particular statistics, and how they might portray different players of different skill sets and usages, e.g. Derrick Rose being high usage and Landry Fields being low usage, well, that’s precisely what we use statistics for – to look at different situations while trying to ‘make everything else equal.’
So when the author writes this statement:
“The only real issue is that the three guards in front of him currently; Wade, Rondo and Fields are much better guards.”
We have no choice but to take this at face value. This author believes that Landry Fields is a much better guard than Derrick Rose. There are no qualifiers listed here. There is no mention of usage, efficiency, shot creation, type of play, etc.
And all of the folks commenting here have the right to challenge this on its face, without having to try to explain away any qualifying statements, because there were none given.
In fact, one way in which this statement hasn’t yet been challenged is grammatically. It’s wrong. It’s bad grammar. It’s a poor use of a semi-colon, or a misunderstanding of how to use non-defining clauses. And if the person writing this really is a ‘Professor’, I would think that he or she would at least have the self-respect, or the word-processing ability, to make sure that such inflammatory statements are grammatically correct.
But beyond grammar, it’s simply not true. Yes, Landry Fields is playing great; he’s a highly-efficient scorer and competent rebounder, while at the same time being the 4th or 5th scoring option on a fair-to-poor rebounding ballclub. Maybe he WILL BE a star in the NBA. But Derrick Rose already is.
And if it can’t be judged as a fact, it can certainly be judged per the collective opinions of the 30 General Managers of the NBA. I submit to all here, that ANY GM of an NBA franchise would immediately trade Landry Fields for Derrick Rose straight up – or would be fired by the owner and ridden out of town on a rail by the fans for NOT doing so.
If John Paxson and Gar Foreman call Donnie Walsh and Mike D’Antoni today, and say “We’d like to trade Derrick Rose for Landry Fields,” well, suffice it to say that Walsh and D’Antoni would likely be SO stunned that they’d forget their poker faces and reply “Okay, who else do you want?”
As for those here who argue that Rose is only an offensive player, and lacks any type of defensive prowess, here’s a link to the type of analysis that I personally prefer: one that considers statistical rankings, yet also ‘breaks down the tape’:
http://offthedribble.blogs.nytimes.com/2011/01/19/rose-dwarfs-other-improvements-with-defense/
According to that writer, Rose has been better defensively this season than Chris Paul, Rajon Rondo, and Russell Westbrook, by no small margin.
In fact, take a look at the Bulls game last night. Rose was not efficient at all; he only shot 6 for 21. Yet he had 12 assists, 6 rebounds, went 8-8 from the line, and gutted out 38 minutes against the Orlando Magic a few days after being diagnosed with two stomach ulcers. Orlando had very little contribution from anyone other than Dwight Howard, and it’s very difficult to imagine the Bulls putting the Magic down like that with the services of D Rose.
Even if he’s not the League MVP, last night was an MVP-caliber performance by Derrick Rose. But Landry Fields is a “much better” guard?
Hogwash, sir. Total rubbish.
Bulls Fan in Krakow
January 29, 2011
One correction: “Orlando had very little contribution from anyone other than Dwight Howard, and it’s very difficult to imagine the Bulls putting the Magic down like that withOUT the services of D Rose.”
Druthers
January 29, 2011
So your system ranks Jason Kidd as more valuable than Derrick Rose this year?
Time to fix your system.
EvanZ
January 29, 2011
@Druthers,
Yeah, it’s interesting how many comments focus on how important “efficiency” is, yet Jason Kidd is having the least efficient season of his career (and that’s saying something, considering how inefficient he was early on). He’s shooting 0.468 TS on almost 9 attempts per 36. Luckily for Kidd, WP bails him out with its (true) emphasis on rebounding. Same reason why Landry Fields is so high. I venture to guess that if Fields averaged even 1 less rebound per game, his WP48 would drop 50 points or so.
Italian Stallion
January 29, 2011
Bulls Fan in Krakow,
>In fact, one way in which this statement hasn’t yet been challenged is grammatically. It’s wrong. It’s bad grammar. It’s a poor use of a semi-colon, or a misunderstanding of how to use non-defining clauses. And if the person writing this really is a ‘Professor’, I would think that he or she would at least have the self-respect, or the word-processing ability, to make sure that such inflammatory statements are grammatically correct.<
You weren't doing badly until this.
People try to communicate as best as they can. Some are more gifted than others. Some take more care to spell check and proof read everything several times because they have nothing else better to do in their life.
Some of us (like me) are 99% interested in the quality of the content and 1% interested in the spelling and grammar.
Pointing something like this out politely is usually appreciated, but being a dickhead about it is not.
Italian Stallion
January 29, 2011
Evanz,
Rebounding, especially at the guard position, can have a ton of value. When Kidd gets a long rebound off a jumper he is immediately in a position to start a fast break that often leads to a high efficiency score by a teammate.
Fans are clearly more impressed by the heroic offensive athletics of Derrick Rose and his 6-20 shooting nights that give him a TS% of .449 than they are with steady Landry Fields going 7-9 for 15 points and a TS% of .833 like they did last night.
What people have to comprehend is that all these things have value, not just volume scoring and athletic ability.
EvanZ
January 29, 2011
IS,
You nicely sidestepped Kidd’s poor shooting, eh? He’s shooting 0.468 on the *season*, barely above the number you posted for Rose on a “bad night” (I assume you thought it was a bad, you didn’t actually say).
It’s interesting, though, that Berri says this in his FAQ:
Huh. More responsive to shooting? Really? Realllllly? Kidd shooting 0.468 TS, yet he’s ahead of Rose with a 0.536 TS. Kidd averages 0.5 more rebounds/36 than Rose, yet is some 70 %-points behind in TS and has a higher WP48? 1/2 rebound?
Come on. You can’t seriously think that Kidd is playing better this season. Can you? Well, I guess you can.
And I haven’t even mentioned the elephant in the room: diminishing returns. Take that into account, and Rose shoots past Kidd (literally) without breaking a sweat.
EvanZ
January 29, 2011
IS, just for fun you should take a look at Kidd’s shooting performances when Dirk was out of the lineup. Here’s what it looked like:
12-28: 3/11
12-30: 5/15
1/01: 5/12
1/02: 3/13
1/04: 3/12
1/06: 0/6
1/08: 2/8
1/12: 4/9
1/14: 3/8
Not exactly stepping it up. Good thing Dirk came back.
Druthers
January 29, 2011
<<>>
Right, because Derrick Rose’s typical shooting night is 6-for-20. When he has two stomach ulcers. Nice job picking one example and presenting it as the norm.
He’s averaging 26 points on 20 shots in the last month, mostly because he’s started to figure out how to get to the free throw line (where’s shooting a much-improved 82 percent this year).
But sure, Jason Kidd grabs a few more rebounds when he’s on the court than Derrick Rose, which TOTALLY makes up for the massive gap between their scoring and scoring efficiency.
Druthers
January 29, 2011
Sorry, the quote I mentioned got clipped above, which was:
“Fans are clearly more impressed by the heroic offensive athletics of Derrick Rose and his 6-20 shooting nights.”
Cool Hand Luke
January 29, 2011
Better to say true shooting %s and points druthers, points per shot doesn’t tell you how many possessions a player uses up. He’s putting up 24.5 ppg in ~22.85 possessions or 23.2 points per 36 in ~21.64 possessions (not including TOs here). You’re talking about the last month, and I’m sure the numbers are higher, but points per shot is less useful than telling how many possessions the person is using.
Bulls fan in Krakow, these rankings are examined, that’s what people like Italian Stallion do. If Rose is wrongly evaluated by a significant amount than there is a significant mistake in the way WP is calculated. Criticizing this post based on how it rates a single person rather than how it rates in general seems flawed. You may only think to look at what’s going on because of a particular seemingly strange result but then, to show it’s faults you have to find a problem in the calculation. The calculations can be shown and examined. The rationale for Rose’s score has been given making your point seem kind of empty to me. The arguments you give seem to be a) Rose is better than Brewer/Bogans duh, b) Berri’s grammar is bad, c) No GM agrees with this, d) his defense is underrated (I agree with this), e) he had a good night. Only one of these is an actual critique of the methods by which rose has been evaluated. Berri is not just making the declaration that Wade, Rondo and Fields are better, there’s a mathematical argument behind it.
I think there are faults in WP, an example being how change in expectancy is considered. For example, a possession is worth ~1, so scoring 2 raises that by 1 and you’re credited with that 1. In this instance we considered the changed expectation. But in the case of defensive rebounds say a team has a 70% chance of getting a dreb (which gets a possession value of 1). Surely getting that rebound increases the odds by 30%, so .3×1 = .3 which is what I think defensive rebounds should probably be credited as. The remaining .7 should go to the defense that created that 70% likelihood to get the rebound. Giving the entire value of 1 to the rebounder seems absurd to me. On that note, I consider drose’s defense slightly better than the average for the bulls team this year (he’s improved a ton) and that makes him slightly underrated by WP in that way. What I consider other flaws in WP overrate him though. I think Rose is actually less valuable than WP gives him credit for, but that’s a different story. The point is, if you’re going to attack how he’s rated you need to attack how that number is calculated (if the result is clearly wrong then either the logic is invalid one or more of the premises must be wrong as well). This can be tried with efficiency arguments, defense arguments, rebounding arguments, etc… I think I’m repeating myself. Better stop.
Oh, and Landry Fields’ 7 of 9 was obviously better than rose’s 6 of 21. Though rose’s 12 to 2 assist to turnover ratio was very good.
Cool Hand Luke
January 29, 2011
Oh, let’s do a quick comparison of rose and kidd too. Given the time they’ve had and comparing to average pg expectations (the expectations might be a little off, haven’t changed my spreadsheets in the past month or so, an example is 6.7 assists per 40 is the average for PGs) we can see how many more/less in each stat they’ve had then league average:
Kidd: 264 assists more than expected, 29 D rebounds, 16 fewer O rebounds, 31 more steals, 23 fewer turnovers, 2 fewer blocks, 59 fewer fouls, 7 fewer charges taken.
Rose: 239 more assists, 5 fewer D rebs, 4 fewer o rebs, 5 fewer steals, 67 more turnovers, 14 more blocks, 64 fewer fouls, 4 fewer charges taken.
The quick comparison says the major differences are kidd’s way higher steal count and waaaay fewer turnovers. For Rose to be better his scoring has to be way better. It is way better. Don’t think it’s enough to make up for the turnovers though, but that’s a question of relative weights (and how you calculate scoring value). Just wanted to give some values though. Oh, defense could make up the difference too.
EvanZ
January 29, 2011
Cool Hand Luke,
When you calculate “expected turnovers”, does that take into account usage? Rose has a usage of 31.6% compared to Kidd’s 15.3%. Just to make the point crystal clear, Rose’s TOV% is 13.2%. Kidd’s TOV% is 21.1%.
For reference, Chris Paul’s TOV% this season is 14.7%.
Waaaay fewer turnovers? Hardly.
Cool Hand Luke
January 29, 2011
If those numbers looked stupid it’s cause they are, they were calculated for the wrong positional expectations (kidd was calculated as an sg, and I think rose was too, though I’m not sure).
If I had calculated it right thought evan, it wouldn’t have been %wise, and that may be a better way of doing, but then agqin, you’d expect, if Rose is using that many extra possessions he’d do things that’d make up for the extra turnovers, right? Because even if he’s turning it over less per possession he uses, he’s still turning it over. And that mitigating factor would have to be scoring I think.
EvanZ
January 29, 2011
The flip side of that argument is that by taking those shots, Rose guarantees possessions end without TO. If Rose uses fewer possessions, his teammates will take those shots. Assuming their TOV% don’t go down with more shots (more likely to be the converse, I would argue), he’s actually helping his team reduce overall TOV. This is the same effect that Kevin Pelton suggested for Melo’s improvement in overall team efficiency. On the other hand, by handing off shots to other players, Kidd effectively increases the number of team turnovers, unless all his teammates have lower TOV%, of course. Can’t imagine that’s true.
Cool Hand Luke
January 29, 2011
Three of the next five guys in usage (Boozer, Deng, Noah) have TOV%s of 13.4, 10.8, 14.7. They also have higher TS%s. Rose is taking away turnovers from them and taking away shots. And assists, and anything else higher usg% will boost. Gibson has a lower TOV% but horrible shooting, Watson is higher on both. But the main other three guys than rose on the team by usage/minutes have generally lower TOV%s and higher efficiency. Rose has a 2.34:1 ast/to ratio. Kidd has a 3.7:1 ast/to ratio. What looks probable to me is that rose uses a lot more possessions shooting than kidd and doesn’t turn it over on those. So where he needs to be more valuable than kidd is scoring (and he is). So isn’t it an argument of how much more his scoring is worth?
Admittedly, a lot of kidd possessions look like shooting turnovers…
EvanZ
January 29, 2011
He’s taking away shots, only if you assume those players can take shots whenever they feel like it without any loss of efficiency.
Let me ask you this simple question…Boozer has a 58.3 TS%. Why doesn’t he take all the shots? Clearly, in a world where the assumption is that every shot is an independent event and treated the same, that would be the best strategy. So, why doesn’t that happen? And why doesn’t James Jones take all of Miami’s shots?
Nerdnumbers
January 29, 2011
Bulls Fan in Krakow,
I will first say thanks for pointing out my grammar mistakes. All of my writing and editing since college has tended to be in engineering. IS is right in that we do what we can. I tend to spend about an hour on a post and sometimes don’t proof it as much as I need. I appreciate any freelance (emphasis on free) editing I can get, at least in regards to syntax. (For the record I am not a professor. Dr. Berri was kind enough to site me at the top of the article. Many people seemed to have missed this)
Now in regards to my face value statement about Fields and Rose. I will take your comment IF you tell me that you leave a similar comment to every major sports blog that makes similar statements. I have been a Nuggets fan and seen tons of these for years (“Melo is top talent in the league”, “Melo is a top five Forward”, “Without doubt Melo IS an All-Star”) I gave a few hyperlinks to the tools I used and provided a decent amount of stats before stating this. If your argument is that I should only make objective statements and list all data I used, then that’s fine. However, if you don’t hold the rest of the sports blogging to this standard well then I can’t really give you much credit for telling me to.
Thanks everyone else for the awesome comments. Sweet to see it hit 100!
Bulls Fan in Krakow
January 29, 2011
Italian Stallion,
Now I understand that this post was not actually by Professor Berri, but by ‘Andres Alvarez’, whoever he is, and Andres may or may not be a native English speaker.
But I stand by my statements. If someone writing in such a forum wants to be taken seriously, they should take pains to ensure that the most contentious statements of their writing be written with correct syntax and grammar; otherwise, they will lose readers because their message will be considered illegitimate. If Mr Andres Alvarez would like to contact me, I would be happy to proofread and correct his writing in a timely manner – for a fee, of course.
Cool Hand Luke,
“Criticizing this post based on how it rates a single person rather than how it rates in general seems flawed….[Berri] is not just making the declaration that Wade, Rondo and Fields are better, there’s a mathematical argument behind it.”
I’m not criticizing the post because of the rating of a single person, nor am I criticizing the statistical system used. I’m criticizing the particular statement that Landry Fields is a “much better” guard than Derrick Rose. It’s wrong. And the author didn’t just make the declaration that those players are “better”, he declared that they are “MUCH better”.
As we all know, there are positives and negatives of any statistical system. Each particular NBA metric has its ‘pluses’ and ‘minuses’ (pun intended), times when it should be used, and times when it should be disregarded or dismissed.
However, making a qualitative statement, such as Landry Fields being a “much better” guard than Derrick Rose, based on any quantitative statistical method or system is making a gargantuan leap of faith, regardless of the industry – telecommunications, consumer packaged goods, internet traffic, or professional basketball, for example.
As such, I understand EXACTLY why this particular statistical system values Landry Fields over Derrick Rose. However, I cannot agree with using this system to make an overarching qualitative assertion that Landry Fields is “much better” than Derrick Rose (and I could even take issue with Wade and Rondo as well.)
My example of NBA GMs was also qualitative, and not statistically nor quantitatively based. But humor me, please. I interpret the statement that Wade, Rondo, and Fields are “much better” than Rose, for the purpose of choosing them to the All-Star game, to imply that they are better NBA basketball players for their respective teams, night in and night out, and not simply more valuable within the confines of a particular statistical system.
So does anyone here really believe that any NBA GM would value Landry Fields in the real world, and not pertaining to a particular statistical metric, enough NOT to trade him for Derrick Rose??? Unless it was Memphis or Minnesota dumping salaries, or afraid of him leaving as a free agent, I’d say that idea is beyond imagination.
Again, I fully understand the value of Landry Fields on the New York Knicks, and in a spreadsheet. And Landry Fields’ spreadsheet value, in this system, is greater than Derrick Rose’s. But to suggest that this spreadsheet value can be projected to the real world of an 82-game season and playoffs, from everything we’ve seen in Derrick Rose’s career to this point, from leading the Bulls in seven games against the Celtics in his rookie season to his heroic performance with ulcers last night against the Magic, is faulty, illogical, simply hyperbole in my opinion. It is not taking into account many factors in the real world that are not present in the spreadsheet.
If the author had made the statement that those players are “much more valuable within this statistical system”, I wouldn’t even bother writing here. But “much better”, period? Hogwash. Total rubbish.
Italian Stallion
January 29, 2011
Evanz,
I’m not here to defend Kidd’s shooting or his ranking relative to Rose.
Kidd’s value has never come from his scoring efficiency or volume. It has always come from his rebounding, assists, steals, limited turnovers and PFs etc…
Kidd is not even having a great year on the boards this year.
The reason he is ranked well relative to Rose on this metric is that Rose’s own high usage scoring does not generate enough value because of his low efficiency to offset all the other things that Kidd still does a little better than Rose.
We are all free to debate the values of those other attributes, but it doesn’t change the facts.
The question is not who is the more efficient scorer.
The question is whether high usage low efficiency scoring generates enough value to offset a bunch of other important skills that definitely have value.
Cool Hand Luke
January 29, 2011
Evanz, no. He’s taking shots away whether or not others would hit them. But I think I get your point. I could ask a similar question though, if rose’s usage at that efficiency is valuable than him taking every shot at that efficiency would surely be even more valuable? It’s the same kind of taking to extreme and it’s not why either of us are arguing. Rose takes ~30% of possessions and shoots at approximately average efficiency. Someone has to take these shots but him taking it this way doesn’t seem worthy of reward. A shot or two away from him to Boozer I’d guess would easily increase team efficiency. While my argument seems to propose that people could keep up an efficiency forever (it doesn’t actually, it just assumes a specific replacement level, WP pick 50 ts%, I’m not sure where I’d set it), it seems like your argument presupposes that replacement shots would necessarily be worse and thus the scoring that happens should be rewarded.
Bulls Fan, I think it’s a given on this site that everything people in the wages of wins network says is based on WP unless specifically noted otherwise and that people don’t have to explicitly state it every time, similar to how when people hold opinions they don’t say “I think blah blah blah” every time, rather they just say “blah blah blah.” Seems semantic. Anyhow, if you think it gets this wrong it’d be nice to state why rather than the who would GMs choose question. This stat obviously thinks those GMs are choosing wrong, so merely stating what their choice would be isn’t going to convince someone who generally agrees with the stat. If he’s got value off the spreadsheet, then what is it missing? If the stat mis-values things, then what?
Italian Stallion
January 29, 2011
Evanz,
I agree that Kidd did not step it up a lot when Dirk was out. That’s what I would have expected because he is no longer a scorer.
Dallas was hurt by Dirk’s absence for a whole variety of reasons, but it was compounded by Butler’s absence. Normally, I doubt Dallas would miss Butler all that much even though he’s the #2 scorer among the starters. But the loss of the two primary scorers made the distribution of those shots way more difficult.
As I keep saying, IMO the value of scoring in not fixed. It depends on the circumstances. Butler is a fairly high usage inefficient scorer that IMHO is usually not so significant, but with Dirk out too, it was a problem.
Italian Stallion
January 29, 2011
Cool hand luke,
Had I gotten around to reading your notes before I started responding to comments to me, I wouldn’t have bothered. :-)
Bulls Fan in Krakow
January 29, 2011
Cool Hand Luke,
I’ll reiterate one more time before moving on to more important things, while not wishing to diminish you with that statement.
The author of this post stated quite clearly that Wade, Rondo, and Fields are “much better” guards not limited to this particular statistical method (WP), but for the purpose of choosing them for the All-Star game.
So, given that, you wish me to explain Derrick Rose’s value off the spreadsheet compared to Landry Fields?
Well, I propose that the All-Star game is for the fans, in order to raise the interest of the Association, in general, around the world.
Therefore, I can simply state that Derrick Rose received 1,225,575 votes for the All-Star game, trailing only Kobe Bryant, Dwight Howard, Lebron James, Dwyane Wade, and Kevin Durant in the entire NBA. Landry Fields, meanwhile, received fewer than 156,230 votes, as he is unlisted below Darren Collison of the Pacers on the NBA’s website. Therefore, the world as a whole believes that Derrick Rose is at least 8x more worthy than Landry Fields to participate in the All-Star game, and more worthy than Chris Paul (949,049) and Landry Fields combined.
See how easy that was? In the statistical method known as “Who does the public believe is more exciting, deserving, and worthwhile to watch in the All-Star game”, the answer is a resounding Derrick Rose. :-)
But if anyone here actually believes at the core of their NBA-fan being that they would rather have Landry Fields on their team than Derrick Rose (assuming that their team doesn’t have Chris Paul, Deron Williams, or possibly Rajon Rondo, mind you), I simply am not going to be able to convince them otherwise, regardless of how many intangibles and off-the-spreadsheet real-world variables I list that aren’t accounted for within this particular statistical metric.
Bulls Fan in Krakow
January 29, 2011
One last note….
If the Bulls had picked Landry Fields in the NBA draft, and he was starting in place of Keith Bogans in front of Ronnie Brewer, I could see them winning the NBA title this year.
I’m not arguing the value of Landry Fields, to the Knicks or any other team, in the real world.
Italian Stallion
January 29, 2011
Evanz
“Let me ask you this simple question…Boozer has a 58.3 TS%. Why doesn’t he take all the shots? Clearly, in a world where the assumption is that every shot is an independent event and treated the same, that would be the best strategy. So, why doesn’t that happen? And why doesn’t James Jones take all of Miami’s shots?”
I don’t think anyone disagrees with this point or many of the others you have been making.
The thing is, it usually doesn’t get that extreme.
What usually happens is that the loss of one scorer is partially offset by his weaker replacement. Then he and the other 4 players divide up the remainder of the usage over all the possessions. So each player is only taking on a minor incremental usage burden. That limits the damage. It is limited further if the player being replaced is not particularly efficient to begin with.
IMHO the value of that incremental burden is a variable. It depends on the circumstances.
If the team is loaded with good and under utilized scorers, it can practically be zero.
If the team struggles to score because it has a lot of players that contribute a lot of value in other ways (defenders, shot blockers, rebounders etc..) without scoring much, then the value of the burden can be much larger.
IMHO, this model tends to underrate/overrate the value of scoring at the extremes of usage because it doesn’t take that issue into account.
IMHO, most others have flaws because they don’t account for the variability of the value or dramatically overvalue it.
I might also even tweak the value of rebounding (especially defensive) on this model , but I still think the general perception among fans and ratings like PER wildly overrate athletic high usage scoring and scorers.
You still haven’t seen “me” say that Fields is better than Rose.
The only thing I feel certain about is that the gap is not nearly as large as perceptions and varies depending on the makeup of the team.
Italian Stallion
January 29, 2011
I want to add one other thing and then I’m going to let this one go.
I enjoy these intellectual debates because I always learn from them. They force me to examine my own thinking more closely and almost always offer new insights. There is one difference between me and many other here.
I bet money on my opinions and I have to handicap tonight’s games right now. :-)
Philip
January 29, 2011
The “this person has a TS% of XXX…why doesn’t he take all the shots/stop shooting entirely?” argument seems absurd to me. WP, like other box-score-based metrics, is designed (duh) from the box score. It observes player performance in a normal setting. It doesn’t purport to answer questions about how a player would perform in some kind of inane hypothetical setting.
It’s like asking “why doesn’t Ray Allen only shoot 3s” or “why doesn’t Lebron only shoot at the rim.” Those are good questions, and there’s a lot of reasons for them. But the point of WP isn’t to answer the “why” players do things. It’s to provide analysis about what they do do in a normal setting.
And normally, players shoot the ball a certain amount, score at a consistent efficiency. Scoring efficiency is one of the less consistent variables across time. But players more or less remain the same player, and drastic change is pretty rare across a season.
Should we analyze players differently based on their role? I believe there’s a possibility that we should. Though I don’t have, nor have I seen, any hard evidence to support this believe.
Bull Fan in Krakow,
Rose was mediocre in the Celtics series any way you slice it. His first game was great, but after that he trailed off big-time, and Rondo absolutely ate his lunch.
And the fact that he’s lionized for his performance in the series is indicative of why we use stats at all: people’s neither people’s memories nor eyeballs are infallible.
marparker
January 29, 2011
Here’s the more appropriate question. If I already have a 50 win team and I can add Derrick Rose or Landry Fields who would I want to add?
My answer would be a resounding Landry Fields.
Now if I have a 20 win team who would I add? Derrick Rose.
Again context matter and I think the more important question is whom to add to your 50 win team.
reservoirgod
January 29, 2011
The best comment was the one that made the point that voting starters for the all-star game is not about the most productive players but the most entertaining players. I’m never paying money to see Landry Fields play. I MIGHT pay money to see Rose play (under the right conditions).
Reas
January 31, 2011
“If the team struggles to score because it has a lot of players that contribute a lot of value in other ways (defenders, shot blockers, rebounders etc..) without scoring much, then the value of the burden can be much larger.”
you basically made the point for Rose over Fields. That’s exactly how the Bulls are built, and the knicks have the complete opposite composite. The knicks are built to score, anything else is just nice to have for them. Fields rebounds well on a poor rebounding team. When Noah and Boozer play together, Rose doesn’t really have any chances to rebound.
Also people keep asking why Rose just doesn’t pass to Boozer more and give him a little more opportunities. I agree, he should, but you also fail to mention that these numbers are over the entire season, and some of Rose’s most inefficient games have come with Boozer out, because as good as Deng is, hes a better defender and rebounder than scorer, Bogans only defends and Noah/Thomas get theres off putbacks or in Thomas’ case, assisted jumpers (usually from Rose’s penetration). From “actually watching games” you would notice that Rose doesn’t always play with Boozer, Noah, Korver, Deng. He’s more often playing with either Boozer or Noah, Taj Gibson/Kurt Thomas, Deng and Brewer/Bogans. I understand the need for advanced statistical analysis but it shouldn’t be taken as absolute, it’s still just a model used to compare players.
I also don’t think players with extreme differences in usage should ever be compared, because its fairly obvious at that point that they don’t have the same roles on their team. Is Landry Fields a better 4/5th option SG than Derrick Rose? I hope so, because that’s not the role or position Rose plays. Could Landry be a 1st/2nd scoring option on a playoff contender while having to create for all of his teammates? I doubt that, because he’s not that type of player, Derrick Rose is, and thats why the comparison is so absolutely ridiculous. Could Landry Fields up his usage to Rose’s and continue with the same efficiency? I have a hard time believing that, looking at his stats and watching both players play.
Stats are great, but please use them in context of what each player is being asked to do. Rose is asked to take on the burden of creating the offense and scoring and taking tough shots when the team is struggling. Landry is asked to play within himself and lets Amare or Felton do most of the heavy lifting. That means he’s hardly taking the last shot, or having to take a shot from a double team because someone just kicked him the ball with 2 seconds left on the shot clock. Star players can’t be compared to role players in efficiency because they’re not taking the tough shots that star players have to take just to make it so its a shot attempt instead of a turnover. So yes please, compare Rose and Wade all you like, but Rose vs Rondo in shooting efficiency or Rose vs Landry is just an absurd comparison. Everyone thinks in what might be, if Rose had all his players healthy, if Landry upped his usage, whatever, but its still just guessing, so rely on the stats and what you see happening on the court. Rose is light years ahead of Landry as an all around basketball player, or Fields would be doing what Rose is doing, since he’s not, he’s not better. Simple.
Bulls Fan in Krakow
February 17, 2011
Okay, I’d like to revisit this entire thread, now that ALL the games before the All-Star break have been played.
Anyone have a problem with Derrick Rose being the starting PG on the East squad – after 4 victories in 8 days where he outplayed Deron Williams, Chris Paul, and Tony Parker? Or was he too inefficient in these victories?
:-)
Tonight against the ‘best’ team in the NBA:
Derrick Rose: 42 pts (18/28, 6/6), 8 ast, 5 reb, 1 to
Parker + Ginobili: 42 pts (17/39, 7/8), 7 ast, 2 reb, 2 to
Pretty good. Pretty, pretttty good.
Landry Fields? Really? Landry Fields?
Last 5 games:
Landry Fields: 36 mins, 59% FG, 67% FT (1.2 att/gm), 12 pts, 7 reb, 3 ast, opponents combined .488
Derrick Rose: 38 mins, 50% FG, 94% FT (7.2 att/gm), 30 pts, 8 ast, 3 reb, opponents combined .587
My favorite quote from above: “Here’s the more appropriate question. If I already have a 50 win team and I can add Derrick Rose or Landry Fields who would I want to add? My answer would be a resounding Landry Fields.”
Memo to ‘marparker’: If you already have a 50-win club and you add Derrick Rose, you now have a 65-win or 70-win club. Add Derrick Rose to Miami, Boston, Orlando, Dallas, or the LA Lakers, and they win the NBA Championship in a sweep. Only San Antonio wouldn’t benefit as much because they already have Tony Parker (17 & 7). Rose replacing Rondo? Another Celtics banner – book it! Rose replacing Jason Kidd? You go from 9 & 8 to 25 & 8!
The Derrick Rose haters are welcome to crunch as many statistics necessary in order to make them sleep more soundly at night. But the question at the All-Star break is NOT: “Is Derrick Rose better than, or more deserving of an All-Star berth, then Landry Fields?”
The question is now: “Is Derrick Rose the best PG in the NBA?” Oh, and throw in another question: “Is Derrick Rose in the hunt for the NBA MVP?”
Landry Fields? :-D
Chadreius
February 24, 2011
What I meant was stats are useless at determining who is better in the nba. Stats don’t account for killer instinct, a smart shot, an important stop, etc. Stats tell me Kevin Love is a bonified beast with only a 15 win team. He has no help? When D Rose was dropped on the bulls they had inconsistent Deng (Still do) and An underdeveloped Noah. They got 41 wins. K Love has a decent point in Ridnour and A proven scorer in Beasley. Not to completely knock the man’s game but I’m sure he and drose are on two way different levels.
William
April 11, 2011
I’ve witnessed performances by all the “top players” in the league. Kobe, Lebron, Wade, etc. and most of the time, they lose games for their teams by clinching onto the ball way too much (sticky fingers). I’ve also seen Derrick stomp on the entire league this year. No one can stop him, as he comes through in the clutch all the time. You guys wouldn’t know because you don’t watch the games.
It’s funny how you stat nerds want to write your comments like scientific data. Also can’t believe there is a community like this, basketball is meant to be watched and played. So sad, I feel sorry for you guys.
He is the best point guard and is a top 5 player, simply watch the games vs. Jazz, Celtics, Hornets, and the rest. Derrick always shuts down Paul, Williams, Rondo, etc.
How can you even think about deciding who is best by stats alone? You can’t tell how a player scored or assisted.. Double team? Splitting the defense? Breaking down the defense with a hard drive? And 1 over a center? With 20 sec. Left? Your eyes prove this, not stats. You sound like Colin Cowherd, who favors Lebron and his high fg%, but where is that fg% when it matters? Nowhere.
Fields better than Rose? You’re an idiot. People like you cannot appreciate the game, the drama, or an up and under layout to avoid a block.
Enjoy Derrick’s game and his MVP season, he embarasses the competition with superior speed, strength and quickness, can finish with either hand, he is so humble and rarely complains about obvious fouls that never get called.
Bulls Fan in Krakow
April 18, 2011
Here here, William!!
Shall we revisit this now? :-)
Landry Fields first playoff game of his career, age 22 years 10 months, vs 2011 Boston Celtics:
19 mins, 0/3 fg, 0/0 3pt, 0/0 ft, 1 reb, 0 ast, 0 blk, 0 stl, 0 points!
Derrick Rose first playoff game of his career, age 20 years 6 months, vs. 2009 Boston Celtics:
12/19 fg, 0/1 3pt, 12/12 ft, 4 reb, 11 ast, 0 blk, 1 stl, 36 points
Derrick Rose first playoff game this season, age 22 years 6 months, vs. 2011 Indiana Pacers:
10/23 fg, 0/9 3pt, 19/21 ft, 6 reb, 6 ast, 3 blk, 1 stl, 39 points
I know, I know – you’d still rather have Landry Fields because Derrick Rose was way too inefficient on his threes, right? :-D :-D :-D :-D
Derrick Rose v. Landry Fields! The bad thing about blogging and the internet: Your mistakes are always out there for others to recall, remember, and laugh about.
Get a clue!!