Lol if all we have is the harden step back then we are the most predictable offense. The spurs have the player and ball movement and pick and roll offense as well as their defense. The heat have the best iso mid range players and pick and roll game. The celtics have rondo who can create off picks and penetration and defense, and the lakers have the triangle. I think the reason our players suck at he mid range is because they don't get any picks and they overthink it too much. Thinking is this a good shot that i can take or not?
The Offense is great almost all of the time; but sometimes it stalls dramatically. Like in that 19 point OKC game second half. What is worrying is that the Rockets will have similar breakdowns in the playoffs if they don't find better ways of re-setting themselves when the offense bogs down.
So what evidence is there that setting screens increase the efficiency of the offense? I don't see any correlation here. I mean, so what if the Rockets don't set a lot of screens. Any evidence to suggest that low-screen setting teams are doomed to always fail?
Nope. Would be interesting to see if there's a correlation on 3pt FG% and screens, but there are so many variables there it may be hard to tease out. What's super interesting is not that we run fewer screens but how much fewer we run. We are 10% below the league average and a huge drop from memphis. That's gotta have an impact somewhere? Not sure if it's positive or negative here.
I think this issue was touched upon in the link provided in the original post. It is pretty long so if you don't have time to read, I'll quote the main point: "Who Cares About Screens? As the only legal method for blocking in basketball, screens are vital to creating open shots, favorable matchups, and forced rotations. Brand new research using Vantage data is verifying what coaches (and smart fans) have long understood: with a league full of world-class athletes, teams that don't screen effectively cannot score efficiently. The guiding star for any offense is "Offensive Efficiency" or, in other words, Points Per 100 Possessions. Research by Vantage Contributor Lorel Buscher verifies that two Vantage metrics, "Received Screens Per Chance" and "Set Screen Points Per Chance," are significant predictors of Offensive Efficiency." In another post on that site, the actual data showing the correlation is provided (I think?), although I will admit I don't really understand some of that stuff... for example they show correlation between setting screens and points generated from setting screens, but then why do I see a negative correlation between offensive efficiency and setting screens? In that chart (shown below), you see the "pie" where set.scrn.pc intersects with Off.Eff, and it is in pink (suggesting negative correlation ) I think I may be misunderstanding the data (could someone maybe clear this up?) I think the evidence (assuming I'm just completely misinterpreted the chart) is supposed to show that correlation. But in a way, you still do have a point, since correlation in and of itself does not say everything. A direct quote from that site: "Does setting a lot of screens make teams more efficient? Or do more efficient teams just set more screens? Which metric causes the other?" Anyone care to elucidate this question? Does it matter where the comments are coming from? If it is established that setting screens can significantly improve an offense, it would benefit the entire team, not just Lin or Lin fans
There is no correlation between setting screens and offensive efficiency? Can you explain the posts in that link for me (I must have entirely misunderstood it )
So in other words, the Rockets front office/coaches are behind the times when it comes to analytical research AND they are incapable of knowing common sense that even fans know? Fire Morey?
Or maybe there is something they know that no one else does? Torocan presented the question that maybe the lack of setting screens was part of the Rockets' offensive system (and maybe by design?). So while setting screens is highly correlated to offensive efficiency, maybe some teams like the Rockets have figured out an even better system that necessitates less screen setting or something. I honestly have no clue. And as mentioned earlier, correlation =/= causation. There is the possibility that a team that sets a lot of screens ends up being offensively efficient because of that, and also the possibility that already offensively efficient teams sets a lot of screens (believing that is what is causing the efficiency). Maybe Morey knows the answer to this chicken-egg problem, and realizes that setting screens doesn't actually cause efficiency, just that many teams set screens believing they would. But this doesn't answer the question why more efficient teams tend to set more screens. Anyways, yeah I know what you're getting at. I doubt that Morey and those stats guys working for the Rockets are behind the times. The question really is that if the lack of setting screens is by design, what are the possible reasons, and what are the specific benefits?
I think we have a secret play book that we've kept secret all year only to surprise teams in the playoffs with. Jk but really when we get stuck in a half court offense we have to set more screens and learn to move without the ball or else we are in for a short ride.
Hakeem Olajuwon, a mid range player? Maybe the source of our disagreement is what 'midrange' means. I don't think there's a coach in the world that would call Hakeem that. If you want to call his baseline fadeaway a midrange shot... ok, I guess, but he gets that shot off by posting lower first. He had a good mid-jumper, but that obviously was not key to the Rockets success. He hardly ever shot that. You could say Malone's jumper was key to the Jazz's success because that was an important component of the p&r with Stockton. But Olajowon's jumper? Quite unimportant in the overall scheme. I'm reminded of JVG: (paraphrasing) "our opponents would be delighted if I relegated Yao to jumpers." Why especially the midrange? For two full seasons we've had a novel system of offense that largely ignores midrange jumpers, and we've been tops in the leagues in offense both seasons. In two seasons, the rest of the league hasn't figured it out? Could you explain why we are 'easier to defend' with our lack of midrange, cause if we're easy to defend we wouldn't be the scoring juggernaut that we are. If/when we lose in the playoffs, it will be because of our defense, not our offense. Harden has shot multiple game winners with his step back 2 when the defense has taken away everything else. I'm glad we have that option, but I can see why the focus of the offense is elsewhere for most of the game. All of our shooters take occasional shots off of screens from Howard/Asik/Dmo if the defense goes under- at the 3pt line.
To repeat- if you chase someone off the three, that defender's momentum has carried them out of the possession for a couple of seconds. It's 5 on 4. The bigs sag off, because the perimeter defense has been penetrated, which is a good thing. You're presented with options to -go to the basket (results: missed layup, missed layup with offensive putback by our big who is now uncovered due to penetration, foul, or foul and one) -pass to the open big, or -pass to the perimeter. How is the penetrating wing/guard easier to defend if you know the player won't pull up for the mid jumper? He has already burned his man; his defender is temporarily out of the play. Do the bigs and wings who are sagging off their men have an easier time defending this penetrating player if they know he won't pull up for the mid J? If it does, does that ease make up for the numerous advantages of layups over 2pt jumpers? As for the shooter who has been chased off the 3pt line- which is better: to pull up for a less-contested long 2, or drive to the teeth of the defense? The data says the latter is better. It's weird that we've adopted a system that almost entirely ignores one over the other, but that's what we've done, and in two years we are a proven offensive juggernaut. There's nothing easy about defending the Rockets, this season or last. I share your concern that about living and dying by the three. Sometimes it seems that all we need is a player to two to enter a shooting slump, and our ship is sunk. And yet the championship Rockets (and Lakers and Heat and Spurs) depended on great 3pt shooting as well. I disagree that our system isn't our identity, it is. Teams prepare for us and they still can't stop us. Even in our losses since Jan 1 we score well. Our system is novel and relatively untested in the playoffs. I guess we'll find out if it is playoff worthy or not. The two seasons we've had bode very, very well for the offense. I agree that Harden gets flustered if the refs call a loose game and allow a lot of contact at the rim, or don't call the arm hacks, but even then I wouldn't call him useless. Also, there are definitely playoff games that are all about star calls. You're right, they all had great mid range shooters, but really only the champion Celtics had a system where that shot was important. Everybody is pretty much inside-out. So are we. I'd be more inclined to see causality between defense, rebounding, and star play to championships rather than great mid range game. Defensive coaches like JVG and Thibs and Pops would agree that if you've succeeded in forcing your opponents to "settle" for jumpers instead of getting to the rim, you're defending well. Why is that? Why is it called "settling" for a jumper?
Where's this correlation anyway? NBA top-10 offenses and their screen% ranking 1. Clippers, 5th 2. Heat, 25th 3. Dallas, 1st 4. Houston, 30th 5. SA, 12th 6. OKC, 3rd 7. Phoenix, 21st 8. Toronto, 4th 9. Minnesota, 22nd 10. NY, 23rd. So among the top 10 offense, 4 are ranked in the top-5 in screens. 5 rank in the bottom-10 in screens. 1 is ranked near the middle at 12th. Seems to me the entire basis of the BR article is pretty BS. Not that I expect any better from BR, but it seems many do believe their drivel now more than one probably should.
Ah, good catch. This is consistent with my stated confusion about why setting screens in that correlation chart did not appear to be positively correlated with offensive efficiency (the "pink pie" at the intersection of screen set per chance and offensive efficiency). I'm curious then how the author came up with the statement that "teams that set a lot of high-quality screens have more efficient offenses." One possibility is that maybe the emphasis is more on the "high-quality" of screens rather than number? Although if that were the case, then the Bulls should be much higher in offensive efficiency as Joakim Noah is supposedly the best screen setter in the entire NBA. But in your list, they don't even break top-10. Maybe the implication is that without Noah setting high quality screens, they would be even worse or something? Maybe already highly efficient teams can be even better if they set more screens? But how would you even determine that? In any case, if your list is correct, my mistake for taking the author's word for it (or completely misunderstanding his post and making a false assertion ). This is why I'm hoping someone else can take a look at the link torocan provided and provide a better explanation for how the author went from the vantage data to the conclusion that good screens -> efficient offense.
After another quick read through, I think I understand where the discrepancy between those blog posts and your list comes from. That site is looking at "set screen points per chance" rather than "set screens per chance," in other words, it is looking at number of points generated by screens, rather than number of screens set. The blog writer said, at one point, "Even a team like the Heat that can rely on a dominant one-on-one player is still in the top eight in Set Screen Points Per Chance." So even though your list has the Heat at number 25 in setting screens, they are still number 8 in getting points from screens. The implication is that they may set low numbers of screens, but their screens are such high quality that they generate enough points to put them at number 8 in points due to screens. Same with the Spurs, they may be number 12 in raw number of screens, but their screens are high quality and tend to result in a lot of points, in fact tied at #1 with OKC. Maybe Phoenix, Minnesota, and NY, are also the same way? The author doesn't address those teams, so this would be a good thing to check. I'm assuming Houston may still be an anomaly, since setting good screens aren't Howard's strength, but that's another thing to look at. The correlation is thus better described as between offensive efficiency and points generated by screens. So you are right that raw number of screens may not necessarily result in more efficient offense, but the combination of setting more screens and setting higher quality screens that can lead to more points is positively correlated with offensive efficiency (assuming the data provided by that site is correct). PS. There were 2 links provided in the original post, the one I've been referring to is the non-BR one, so IMO warrants some more examination.
If so, we have a lot of room for improvement once we start setting more screens. This could be a catalyst for us to go deep in playoff.
Right, unless, as some have suggested, the lack of setting screens is by design, and Morey and co is seeing something that the rest of us aren't. There is no way they would be entirely oblivious to the existing data in support of the significance of screens. The question posed in this thread is why specifically is the Rockets offense geared this way. Does placing less emphasis on setting good screens provide some benefit that we aren't seeing? Is it because Howard doesn't set the highest quality screens, so they figured it's best not to waste energy on setting them in the first place (if this were the case, do the Rockets set more screens when Asik is starting, assuming he sets better screens?) The other possibility is that prior to these Vantage metrics, there was not that much good data on the significance or effectiveness of setting high quality screens. As torocan pointed out, Synergy data only accounted for raw screens set, which was NOT correlated with offensive efficiency (as Meh's list deftly demonstrates). Vantage, on the other hand, can directly measure the quality of a screen by tracking if the the screen "makes contact with a defender, or re-routes that defender" (direct quote). It also tracks whether the screen leads to points scored. The resulting data thus suggests that screens may not increase offensive efficiency perse, but good quality screens that result in points definitely do. Since this data is all pretty new, only made available with Vantage, maybe Morey is taking the time to verify it, or maybe it just takes time to implement a new screen-heavy offense? Lastly, I will say there is one criticism I have of regarding the whole points-generated-by-screens idea (this is briefly mentioned here as well), which is that offensively efficient teams may simply have good players who score regardless of having screens or not. If you have a player who is offensively gifted like Tony Parker, how do you know that a good portion of his points are "generated by screens"? Maybe he was going to score regardless, which would artificially inflate the points generated by screens variable. But then, assists are the same way. You can't really know if the assist is what *caused* a made field goal, and having offensively gifted players would naturally inflate assist numbers (while having players who can't shoot would deflate assist numbers, no matter how good your ball-movement is). But no one is arguing that assists are not important for an efficient offense. The assertion made in the vantage blog is that "Set Screen Points Per Chance" is basically a "a non-passing assist". So if we regard assists as important for an efficient offense, then the argument can be made that setting high quality screens that can lead to easy points is significant as well.
I would not recommend changing the Rockets' overall offensive philosophy. What they do, works. But 23 percent is low, when the league average seems to be about 32 or 33 percent. Even the Heat run screens on 29 percent of their set plays. (OKC - 37.9 percent; Spurs - 34.8 percent) And also, we have -- Howard, Asik, Harden, Lin, Parsons.
