I don't know if comparing today's stats to historical ones serves much of a purpose to prove anything. Today's game isn't the same as the game being played even 10 years ago. Rules are different and the emphasis has changed. All you need to do is look that the number of 3 point attempts across the league and the percentages. Today's players shoot the 3 much better than prior generations and because of that they shoot it much, much more. If all we did was attempt to shoot 3pointer then I might agree about the turnovers. While we certainly do need to cut down turnovers, I don't think that our turnovers are caused by us shooting 3's. We're just generally sloppy with the ball much of the time. We could run the Princeton offense and we'd still have turnovers. We have a few guys that aren't good ball handlers, Harden is always going to force the action and that will lead to turnovers and Howard is going to get doubled and that will result in turnovers.
In addition to the team having a historically bad 3pt shooting year, the Rockets are ALSO missing a ton of layups. Maddening. Further, this team NEEDS fast break points to put them over the top. Those fast break points aren't materializing because of the poor defense. I feel fast break points should be treated as "value add" to a team's offense instead of PART of the offense. But with Houston they are a PART and not a value add. Rockets half court offense hasn't evolved and that is a problem. DMo being out is a huge blow. It all ads up.
I didn't say that. I said compare our 3FAs to TO ratio to *any* successful playoff team, including now. I just said you can also go back all throughout history. Do it. It's shocking. We are an outlier. Heavy doses of 3FA should help you control the ball and get more quality possessions. That's the real reason behind less emphasis on 2s. 2s are more contested and are known to cause more TOs, too. We also fail at the 3s vs TO measure because we really suck at the layup part of the 3s and layups vs motion offense like SAS, ATL and GSW. We suck at the motion aspect of 3s and layups to get layups ( not saying motion is the only way, but it's what others do). Since we suck at this, because we don't have the players and coach suited to the system (and that's on Morey), the defenses have less to worry about. They've gone to school on our offense, and they are causes us to shoot contest threes at 10% of all our shots, which is second only to the Lakers. This isn't really up for debate. This is analytics behind the analytics of 3s and layups.
Thanks for the updates Batman. Agree 100% about the 3's and layups. We are leaving so many points at the rim. It's unconscionable. I don't know which is the chicken or the egg; we aren't hitting our 3's, so teams clog the paint. We are ineffective at getting points in the paint, so we force passes and rack up turnovers. When we take 3 point shots, they rarely seem to be a catch and shoot as nobody is moving well off the ball and creating that extra space. It's like the basics of basketball took a back seat to this mess that we have right now.
Not optimal. <blockquote class="twitter-tweet" lang="en"><p lang="en" dir="ltr">Of players shooting at least 100 shots this year... Corey Brewer has the lowest FG% in <a href="https://twitter.com/hashtag/NBA?src=hash">#NBA</a>. Ty Lawson 2nd lowest. <a href="https://t.co/NtLde29sCB">pic.twitter.com/NtLde29sCB</a></p>— Daren Willman (@darenw) <a href="https://twitter.com/darenw/status/668879035456462850">November 23, 2015</a></blockquote> <script async src="//platform.twitter.com/widgets.js" charset="utf-8"></script>
The quantitative approach is just a methodology that has many different subfields and levels of application. Some of the larger issues with quants in general is that they believe that they can measure everything. For example, durvasa on ClutchFans actually tried to tell me that you can measure leadership by how early somebody comes to practice. That's obviously ridiculous, because there is no correlation or causation between coming early to practice and having leadership and authority among peers in a locker room, especially when it's mainly the guys on the end of the bench that show up early most often. Intangible components like leadership and heart are not quantifiable and this is something that advocates of the quantitative approach reject. Thus, they ignore those components and feel that they're not relevant. In particular, as it relates to Maury, some of the issues with his quantitative approach is not valuing team chemistry and continuity, feeling that height in basketball is unimportant, and that coaching doesn't matter. Also, the biggest problem with his approach is instituting a "philosophy" and "system" that is highly dependent on 3 pt shooting, but not having any good 3 pt shooters on the team.
This is an excellent point. But it's up to Morey to evaluate the results of analytics and make the correct strategical decision. I know this will be a very unpopular opinion around here but Harden's habits promote Iverson / Young LeBron / Kobe ball as opposed to Moreyball. Harden is most comfortable with the ball in his hands on the attack. He'll pull up for 3 nine times out of ten strolling up the court if he just made a nice play previously or feels in the slightest bit of rhythm. Otherwise he'll attack the defense and try to draw a foul or kick out to an open guy. That's not moreyball, it's not what San Antonio nor Golden State is running. The only play with even some semblance of ball movement that we currently run is a Harden/Howard pick and roll or Harden/Capela pick and roll. And that's why so many posters are clamoring for an offense with H&H pnr + weakside ball movement. Because we don't even know what it's like to see more complicatd motion offense or what it would look like here. For better or for worse, as long as James Harden is our franchise player we will not be running a motion offense (that as you stated and I agree with, is most conducive to 3s and layups when paired with a skilled coach). Our best bet is Harden iso + Harden pnr and surround him amazing defenders, shooters, and rebounders. Kobe ended up winning 2 rings due to this formula so why not take a shot?
I wouldn't say leadership isn't quantifiable. We just don't have the right data or enough data. Also, causation isn't nearly as important as correlation. What you can do is say, "leaders" (however you define it, which is important too) have x,y,z traits in common. Yes, those traits aren't causing leadership but it can still predict behavior. Like how they predict voting behavior based on things like where you shop, where you spend money, who you're friends with, what you do, etc.
I would think it wouldn't be too hard to determine what the metric equivalent is to "intangibles guy" or "leader". Basically you pull together a big database of all basketball teams and all metrics for how they effected their team's chances to win, and then you rank your top subjective "leaders" or intangibles guy for the various years and let the computer AI determine what statistical similarities exist between the players on your list. This was covered during the whole NY Times "Shane Battier just helps teams win" article where they talk about how his stat line was underwhelming, but the net effect he had on his team's fortunes was overwhelming and then discuss Morey's use of sabermetrics to identify that.. Morey was widely praised for that but now that the team is losing, people are acting like he hasn't cracked that code. I'm not buying it. The difference is MORE teams are utilizing sabermetrics now as well so the intangibles guys & locker room leaders are no longer being undervalued.
It's like saying 90% of team leaders (however you define it) show up early to practice, but 90% of people who show up early to practice aren't team leaders. Though this might sound like you can't draw any conclusions from it. You can. You can eliminate people who don't show up to practice early as potential team leaders, though it doesn't necessarily mean those who show up early to practice are leaders. It's still useful. But the big thing about stats and behavior and predicting behavior is, the inputs can change. Which is why no one could predict Draymond Green would be as good as he is, else he wouldn't have lasted to 35. It's why the draft is still an imperfect science. You can probably predict the odds (based on a host of variables) whether a guy "figures it out" or matures. But the key to sabermetrics isn't to predict outcomes necessarily. It's simply to tilt the odds in your favor.
Things said in friendly conversations don't hold much water, particularly if the individual's actions are in direct contrast to his words.... A man can preach about the virtues of chastity all day long, but if his actual behavior mirrors that of a skirt-chasing lascivious lech... then his words mean nothing...
We disagree about Morey's behavior. I'm sorry you don't trust your friends when they talk to you. I do.
One man's friends is another man's acquaintances... How loosely or firmly are you using the word...? I have another question, when did this "always" in the statement "he's always been more concerned with heart and chemistry than analytics" start coming into play...? Was it when he assembled the trio of JLin, Parsons and Harden under the guidance of one "I don't know, they should know what they need to do (referring to his young, immature team)" Kevin McHale...? Was it when he paired Dwight Howard with James Harden and along with Les gave Mchale an extension and called it a winning foundation...? Now, me, personally... I don't think Howard by himself is a problem (injuries not withstanding), but Howard with Harden is a problem... especially under the helm of a coach that has neither the credentials (Bickerstaff/Mchale) or the personality (McHale) to be a leader and hold all players accountable... Nothing about that arrangement says heart and chemistry is a priority to me...
I don't mind the Rockets shooting 20-30 threes but bear in mind you are most likely to get to the free throw line attempting a lay up or a mid range pull up jumper.
Are you asking me how close I am to Morey? Kind of a weird question but okay. We're good friends, he's on my board of directors, and we're creating a new musical together. And he doesn't have any reason to mislead me about the team. He knows he can say (and often does), "I can't talk about that." If he doesn't want to share something with me, he doesn't. But he doesn't lie to me. That would be ridiculous. Still it makes me a little uncomfortable to post about a friend like this so maybe I'll stop here. I haven't violated a confidence but it still feels awkward. I wouldn't have posted at all but the public perception of Morey is almost a parody and I felt it was important for me to share the fact that it does not square at all with what I know of who he is and what he values.
Sometime in '07, my last year of grad school, Morey presented to us on campus @ Rice. I asked him about Bonzi Wells, "how do you asses the unquantifiable aspects of Bonzi's addition to the team?" The answer was honest albeit unimpressive. The saying that you're only as strong as your weakest link applies. If players were emotionless chess-pieces, Morey's team construction acumen and approach would be flawless. As is, the weakest link in the team composition process partially devalues the analytical iterations of his team construction model(s). In the financial world, Citadel is well known for their "edge" on team construction. They spent probably 25K just to interview me with, exorbitantly paid, industry leading consultants. That's a very practical application of a revolutionary company identifying and then relentlessly pursuing an area - or process - where they've identified low hanging fruit or optimization which creates value. It's clear to me that teams like the Spurs, and more recently GSW, have bridged the subjective gap that still remains with (almost) entirely quantitatively dependent team composition. This isn't a knock against Daryl. I'd hire him before the printer finished his resume. But specific to this discussion, there is clearly a gap in the process, I'm not convinced that anyone internal with the Rox org can fix it. Now, the beautiful thing is, if Ty regresses from terrible to merely average and things like TO's and 3% moderate (embedded w/in Ty regressing), this isn't even a discussion.
