JULIET:
…What’s in a name? that which we call a rose
By any other name would smell as sweet;
–Romeo and Juliet
One of the many famous phrases from the bard and a favorite of those attempting to pitch woo. But the story goes that Shakespeare was making a joke at the expense of the Rose Theatre, a local rival to his Globe Theatre, who’s batrooms were not up to par. This may be a legend or it may be true, much like the sweet nothings being whispered in a ladies ear. The point of this is that what our eyes and ears tell us isn’t always the case. Sometimes perception doesn’t match reality, sometimes it does and sometimes it tantalizingly might. That is how we come to perplexing case of Derrick Rose.
Derrick Rose is an all-star. Derrick Rose was the rookie of the year. Derrick Rose is a World Champion. These are all facts.
Derrick Rose was not really playing like an all-star. Derrick Rose was not the best rookie in his class. These are also facts.
But what is the truth? Is he a star or is he overrated? The truth is never quite so straightforward.
A Rose by the numbers
- Truth is truth
To the end of reckoning. - William Shakespeare, “Measure for Measure”, Act 5 scene 1
For newcomers, go here for the Basics.
When I did my Bulls review, A lot of readers were surprised that Rose, while the leading scorer for the Bulls was not the most productive. Let’s look at Rose’s number for last season (versus the other 148 guards who played more than 400 minutes in 2010) to explain:
I looked at basic statistics per 48 minutes as well as sho0ting % (Field Goal, 3 Point and Free Throws) as well as points per attempt (just Points divided by Field Goal attempt as I want to give credit for free throws drawn and converted). Compared to other guards in the league last season, Rose was a better shooter in overall FG% but worse at 3P% and FT%. This meant that he generated points at just slightly above league average. He did score almost eight more points a game than average but it took him 6.5 more attempts to get it done. He was above average in assists. The possession stats (Rebounding, Steals & Turnover), however where not good placing him very near the bottom of the league for guards. So overall he was good with assists, average at converting attempts to points and bad at getting/losing the rock for his team. If we use the composite stats (Wins Per 48 (WP48) and Wins Score per 48 (WSP48)) we see that he was slightly above average and just about the level for a starter but very definitely not a star.
To drive home this point, let’s look at how Rose compares against the very best point guards in the league:
Rose scores more than them but he does it much less efficiently. The difference in overall scoring is more than made up by the elite Point Guards because of superior passing (as reflected in assists). All but Nash are better at posession stats as well (but Nash is just a ridiculous passer). At this point you may be thinking that this comparison is unfair, he’s young and a star for his age. We can look at his peers though to get some perspective:
There were three other 21 Year old guards in the league last year who played more than 400 minutes: Russell Westbrook, Stephen Curry and Rodrigue Beuabois. Westbrook was a worse scorer but better at everything else. Curry was the superior shooter, better at possessions and almost across the board except for raw scoring. Beaubois (granted in limited time) was better at everything but fouls,assists and offensive rebounds. Westbrook and Beaubois were better in both the overall metrics WP48 and WSP48. Curry was superior in WP48 and behind in WSP48. So against his peers Rose was overall a slightly below average player.
The final argument for Rose is that he creates more points than other point guards. To counter this point I made up a statistic on the spot :-). I calculated :
- Net possessions generated by the player per 48 minutes = (REBS+STL-TO).
- Net Possesions Spent by the player per 48 minutes=(FGA + Assists)
- Net Possesions Used by the player per 48 minutes = (FGA + Assists) -(REBS+STL-TO)
- Points Generated by Team as a result of the player per 48 minutes = (Pts + 2.17*Assists) (2.17 is the avg points per field goal converted for 2010)
- Points Created per Possesion Used equal to Points generated by team/Net possessions used
The results look as follows:
This table is extremely illustrative. Rose generated the 12 most offense of any guard in the league on a per minute basis last year but he was in the bottom twenty percent in converting possessions into points. In 2010, when you saw the Bulls play against the majority of the league, Rose looked like the most offensive player on the court but the numbers reveal that this offense had a great cost. He was a 300 million dollar movie that generated 300 million dollars last year.
So in the 2010 regular, Rose was a good but inefficient player that didn’t watch the hype but don’t worry Bulls fans cause that’s not where the story ends.
A Rose in Bloom?
Derrick Rose is twenty two years old. Derrick Rose raised his WP48 from .053 in the first half of 2010 to .153 in the second half. If we look at point guards at 22 years of age over the last five seasons (13 in total):
Eight of 13 have improved, five have improved significantly. Rose is giving indications of doing just that. Yes he’s been inefficient but he got better. Bulls management has certainly given him all the tools (better coach, better teammates) to move him in the right direction.
The reckoning is at hand for Rose. To be a star or not to be a star,that is the question and for the Bulls that may make all the difference between disappointment and a ticker tape parade down Michigan Avenue.
Arturo's Silly Little Stats v2.0 
some dude
10/05/2010
Considering he had a nasty ankle injury the 1st half of the season and clearly looked slower, I take his .153 WP48 mark more serious that the season average. I think that was the real Rose.
That would put him above Curry and Westbrook (with WB’s defense far better) and just below Roddy (but I’d like to see him play a bit more).
This is one case where “you have to watch” hold up because early season Rose was running at 60% capacity and everyone was saying as much.
arturogalletti
10/05/2010
sd,
That’s my point and the Eastern conference hinges on it. If Rose is the .153 guy that we saw in the second half (and not the guy who didn’t show in the playoffs) and he grows like the top half of that list? Then Chicago is every bit as good as Miami . Rose is the key. He also needs to pass more and shoot less.
Chicago Tim
10/06/2010
I don’t expect another CP or Rondo, but I think it’s realistic to hope for another Deron Williams. We shall see.
arturogalletti
10/06/2010
CT,
You’ll know it by the ring count.
nerdnumbers
10/06/2010
Arturo,
Love it! I think we agree on this but Chicago and Portland are both the teams with such a high ceiling hinging on a few factors. Portland has the health thing to worry about and Chicago needs Rose to not turn into Melo (score at all costs to get that max deal) and Boozer to stop punching walls.
Raspu10
10/06/2010
Very nice analysis. Love the on-the-spot point guard (or playmaker) stats. Any chance of seeing the full ranking?
Re:some dude – if you’re going to go for second half, you have to do the same for Steph. Overall, that’s a very talented group of 21 year olds, and I suspect they’ll all get better.
arturogalletti
10/06/2010
In the 2nd Half Curry was .196, Westbrook was .188 (and a ridiculous .328 in the playoffs against the eventual champs , which is actually the reason I have OKC as a contender), Rose was .143 (but he pooped his pants in the playoffs at .002), Rodg was .228 in the second half in 441 minutes.
I’d worry less about Rose if he hadn’t been that bad in the Cleveland series.
As for the Playmaker stats, I might be able to do something about that (with some refinements). I think I’ll include all the players.
Austin
10/06/2010
Good stuff. Two suggestions:
1. TS% basically gives you points per shot, including free throws. It’s a fantastic box-score statistic, and besides turnovers it instantly tells you how efficient of a scorer a player is. Personally I would use it instead of showing FG%, 3pt% and FT% separately.
2. I believe Dean Oliver’s ORtg is a more, shall we say, thoughtful attempt to calculate something like your points created. It’s pretty darned complicated though. Basketball-reference.com attempts an explanation but it mostly just refers to Basketball on Paper (which I have not read). Also, I think it’s important to show USG% in situations like this: if Rose has a greater portion of his team’s offense run through him, he’ll probably be less efficient. Now that’s not necessarily a good thing, but efficiency and usage complement each other in telling you how good of a scorer a player is.
arturogalletti
10/06/2010
Austin,
I’m familiar in both TS% and Ortg (and I agree they are both good). I was trying to keep everything mass audience friendly and TS is unfamiliar to most and ORtg is very hard to explain in a thousand words or less. By using FG%,3Pt% and Ft%, I kept it in terms a more casual fan can understand which is also what I was going for by doing a very simplified and flawed version of Ortg.
The Usage/efficiency ratio is a function of diminishing returns and it seems to be something that the great players have to get to be great. As a point guard, You have to id teammates with higher percentage shots than you and pass them the ball rather than taking a contested shot (and do it in a split second). That way you can raise the your teams return on possesions (you also force the defense to have to account for that which’ll get you better looks). Obviously you need your teammates to be good offensive options for this to work. Rose has the teammates now for this equation to work. If he forces the shot less and looks for his teammates (who can very definitely convert shots into points) his team will score more efficiently and win more. If he’s going to get it done the time really is now.
(By the way, I’m kinda speaking to the lurkers and casual fans more than the sharps with the piece and this comment)
Austin
10/06/2010
Cool. No surprise that you’ve already thought through all of this. Thanks as usual Arturo!
brgulker
10/06/2010
Awesome post, Arturo. This was incredibly well done.
arturogalletti
10/06/2010
Thank you for reading and the kind words.
Almighty J
10/06/2010
Derrick Rose is the most fascinating player alive.
He’s the evolutionary Allen Iverson, expect more complex and polarizing, if that’s even possible.
Arturo, have you compared Rose’s college productivity to the Freshman year productivity for the other point guards mentioned (Paul, DWill, Rondo, Westbrook)? I think that it takes players a couple years to learn how to play the point and am interested to see whether part of the disparity between Rose’s productivity coming out of college and some other notable point guards is partially due to the years each had to learn how to play the position at a college/pro level.
I think you can make a similar argument for Rose’s NBA numbers if you consider that a) all of the other players had more time to learn the point in college and b) none of them were asked to shoulder as much of a burden in the pros (be the primary scoring threat for a team with no outside shooters or low-post scorers), but that argument seems to devolve faster than a facebook discussion about abortion.
Great post as always and I applaud you for noting the difference between measuring past productivity and predicting future productivity. Keep up the great work.
arturogalletti
10/07/2010
J,
I have looked at college productivity for Rose as compared to the other PGs (go take a look at the Wizards posts for some of this work). The difference between the 19 and 20 Year old version of a player >> the difference between the 24 and 25 year old. Guards in particular are skill based (handling,passing,shooting) and it takes a while for them to develop. The point that’s been made for Rose (and now for Wall) is that while he could be a great player he didn’t show it as a freshman. What this means is that with these young point guards you’re getting unskilled labor (with potential) but your initial investment will be very high for little or no return in wins.
However, Rose now has the hours and now has the talent around him. His employer has every right to expect him to brig the goods. I think he has it in him but it’s up to him to get it done.
You’re right about the argument typically devolving faster than a Palin post on FARK or an internet discussion of evolution. The derp is powerful with AI,Kobe, Rose, Murphy and Lee.
Almighty J
10/08/2010
Arturo, I couldn’t find the Wizards’ post with the college point guard analysis. Can you point me in the right direction?
I’m specifically interested in comparing Rose’s college production to the first year production for other point guards. I want to know whether other point guards were similarly unproductive during their first year in college.
Also, I know we men of science balk at the notion of finding significance in small sample sizes, but it bares mention that Rose was incredibly productive during the NCAA Tournament.
Regardless, “The derp is powerful” is the funniest thing I’ve heard today. Classic.
arturogalletti
10/08/2010
Hurr?
🙂
The infamous piece is here
I’m putting up a full rookie model in the next few hours.
Almighty J
10/08/2010
Ok, that was ALMOST what I was looking for. Full Disclosure: I’m a stat-minded Bulls fan trying to square my Derrick Rose holy-balls-look-at-him-score man crush with the overwhelming statistical evidence that he’s a solid, yet unspectacular producer of wins.
My crossing-my-fingers-and-hoping-it’s-correct hypothesis (and it’s potential significance): Learning to play point guard has a steep learning curve and current productivity measures do not account for this in their evaluations. The historical comparison for Derrick Rose (or John Wall or next uber-athletic-one-and-done-omg-I-need-to-stop-using-hyphens point guard) shouldn’t be Chris Paul the junior or Rajon Rondo the sophmore (interesting quick observation, if Rondo’s rebounding rate jumped from 2.9 in 25 min his freshman year to 6.1 in 30 his sophmore year, a huge difference). We should compare apples to apples.
Maybe the conclusion is that one year of a developing point guard doesn’t give you enough information. It’s a crap shoot. But, I’d like to know how point guards stack up early on in their development.
So…. ummmm….. go do all the work for me. Ha ha ha. 🙂
Have a great day, Arturo.
arturogalletti
10/08/2010
Check the new piece 🙂
The problem is in part that the college experience helps pg get his reps and learn the game. A pg who’s lacking as a freshman is a better bet than one who’s lacking as a senior. He can be taught in the right system/coach (DelNegro not the guy, TT I think should be)
some dude
10/06/2010
Raspu, good point. I have no problem with the claim that those guys might be better. They are all talented.
Arturo, Rose was a .002? That doesn’t seem right to me. He scored more and assisted more without much of a change in rebounds or TOV in the playoffs. His shooting efficiency dropped, but not that much. I remember that in the series Cleveland was struggling to defend him. They only won 1 game because Cleveland was much better, but it seemed like Rose played well. Not as good as 2nd half Rose, but his numbers seemed to be better than total season Rose which was around .1, no?
arturogalletti
10/06/2010
I’m probably going to post this later rose’s pts per attempt went down to 1.07 which was 109 out of 141 players in the playoffs who played more than 60 minutes and he took more shots than anybody on a per minute basis. 2 thru 5 in shots were Kobe, Carmelo, Durant and Wade at 1.31,1.33,1.22,1.51 per attempt. You have to go down to Joe Johnson at 12 to find someone who made less points per attempt.
That .002 was valid (but can be explained in part by the opposition).
Raspu10
10/07/2010
Didn’t think you would. 🙂
I’m actually kinda fascinated by the parallels between the current Chicago starters and the Warriors. Chicago’s about a year older on average, but both units have the same number of years in the league. Curry vs Rose is obvious as a fun matchup…. but so are Noah v Biedrins, Lee v Boozer, and Deng v dWright. We should be seeing excellent games between these squads for years to come – or until one or the other team gets torn apart.
some dude
10/06/2010
Oh yeah, like I said he did have a drop in efficiency. But a lot of it seemed to be in just 1 game. What were the individual game WP48s?
I mean, with only a 5 game sample, there’s room for a lot of variation. The average doesn’t tell us much.
And he played against a very good defensive team (as you mention) whereas say melo did not. WP48 in the playoffs bother me, especially for one series, because it doesn’t adjust for opp. Well, it’s not fair to say that. Pretty much all stats in the playoffs suffer from this.
After the WCF, the lakers looked like a juggernaut offense through the playoffs, but I knew this wasn’t the case. It’s just that Phx’s D was that horrendous it made them appear to be much better than their true value. The low scoring Boston series was something I predicted and happened.
My memory is foggy, but did they at one point move Lebron onto Rose because Mo Williams was struggling with him? I can’t remember. I know they did that with Rondo in the next series, at least for a bit.
Anyway, this is how I see it. Would you rather take a .15 WP48 by Kobe vs Boston or a .18 WP48 of Melo vs. Phoenix. (made those up). The answer isn’t so easy, is it?