So, Brian did a really good job defending the Anderson/Davis combo here:
With this information, the questions about the defensive capability of the Holiday-Gordon-Evans-Anderson-Davis lineup pivot towards Holiday and Evans and away from Davis and Anderson. It's unclear whether Evans will be able to use his limited stature (but enormous wingspan) to effectively corral opposing small forwards. And it's equally debatable whether he can even be bothered to make an effort on the defensive side of the court.
- Brian Ball
He basically looked at the lineup combinations that the Pelicans/Hornets had last year. The Anderson/Davis combo was "bad" in terms of raw numbers. (Around 120.8 points per 100-possession). But if you dug a little deeper about who those other three guys were, it makes sense there might be some hope regarding the RyNo/AD combo.
And then last night, as I was reading my usual sense of basketball (APBR forum, Pelicans Report), I came to a question - has anybody actually discussed the importance of shooters on a numerical scale?
I mean people say that shooters are important because they help spread the floor for other guys (and in a symbiotic way, they are helped by the attention this other slasher/post player gets). And it got me thinking -- has anybody really looked at this relationship?
I'm sure someone has and it may have been among the many research papers that I haven't read in SSAC. But even if there were, I'll still go through with this because I really feel like it's a topic that we as Pelicans fans need to discuss.
Shooters gonna shoot, Slashers gonna slash
The league average 3PT% this season is 35.9% and generally hovers around the 34~35% range. So, here is a list of all the players I'd consider as "shooters". It's a list of players who attempted at least 2 3PT per game, shot at or above 35% and qualify for the 3PT% leaderboard (i.e. I think 200 3s, not sure). If you sorted those according to their teams, there would be a total of 26 teams with two or more shooters: Atlanta (4), Boston (4), Brooklyn (3), Chicago (3), Cleveland (2), Dallas (3), Denver (2), Detroit (2), Golden State (3), Houston (3), Indiana (3), Clippers (3), Lakers (4), Memphis (3), Miami (5!!!), Milwaukee (3), New Orleans (2), New York (4), Oklahoma (3), Philadelphia (3), Portland (3), Sacramento (4), San Antonio (4), Toronto (3), Utah (3), Washington (5).
Among those teams, I'm going to focus on the following: Boston, Milwaukee, Sacramento and Washington. Why those four among the 26 teams? Well, the purpose of this exercise was for us to identify how a shooter helps a slasher. With the other 20 teams, their primary "slashers" were also "shooters" (in a sense) or they depended on post player (something that the Pelicans won't probably have this year, so no use in discussing this here. I might extend the study and find another place to post it): Lebron James, Chris Paul, Marc Gasol, Lamarcus Aldridge, Deron Williams, Stephen Curry, James Harden, Paul George, Kobe Bryant are all formidable post up players OR capable off-the-bounce 3PT shooters. I decided to not include Oklahoma because, well if your any slasher (like Westbrook) and Durant is off-the-ball, you're PROBABLY going to get a LOT of space, yes? Plus, Westbrook's played a lot in the post this year. I also decided to remove San Antonio because, well, they are San Antonio.
Also, those four teams have something in common with the Pelicans' projected 2013/14 lineup: Boston with Rondo, Milwaukee with Ellis/Jennings, Sacramento with Evans (YEAH!) and Washington with Wall - all are players who are not really respected for their jumpers (like Evans is). So, we can take a look at how that teams' offensive rating changed with lineups and how the players shots near the rim (and equivalently, percent assisted on 3PT shots) goes up as they play with shooters.
So, how do we do this?
I'd first like to thank Evan Zamir for an incredible job with the website, NBA Wowy. It has slowly become among my go-to sites (and one of my favorite NBA stat websites).
Rajon Rondo was injured midway through the year. When he played, Boston scored at a rate of 100.9 points per 100 - that's way below league average.In fact, over that time frame, Rondo attempted 41.1% of his attempts from 0 to 3 feet, where he made 57.6% of his attempts.
But if you isolated his minutes to those times when he played with at least two shooters (Pierce, Terry and Lee), their offense suddenly doesn't look as bad:
|Lineup||Total Minutes||Offensive Rating||%0-3 FG||FG% from 0-3|
As you see, outside of that Rondo-Terry-Lee trifecta (one that barely played the equivalent of one game), Boston played way better with at least two shooters in the lineup with Rondo. In fact, if you looked at the Rondo-Pierce duo, a lot of their minutes with big men were swallowed by Brandon Bass and Kevin Garnett -- two very effective midrange shooters. That might have contributed heavily towards their "inefficiency" (since Rondo has a knack of finding both of those guys for shots). Whatever it is, there's a clear increase in Rondo's % of shots taken near the rim when he plays with shooters.
Ellis and the Bucks are a funny case. Despite Milwaukee having three really good three point shooters playing a lot of minutes (Jennings, Ilyasova, Mike Dunleavy), the team was still a horrible team at spacing the floor. Ellis, for his part, played a lot of minutes with those three players. How did it fare?
A lot of Milwaukee's inefficiency stems FROM Ellis -- his shot distribution is a sight for sore eyes. Ellis took 27.3% of his shots from 16+ feet (non-3PT shots), that's a HUGE number of really long 2's.
However, suprisingly, the four man team of Jennings-Ellis-Dunleavy-Ilyasova scored a whopping 116.3 points per 100-possession -- that's a surprisingly big jump from how Milwaukee fares with Ellis (104.7 points per 100). Yes, yes, that four man combo played just 378 minutes together (the equivalent of around 8 games) but still, those numbers are mind blowing. Even if you isolate those four into whatever slice you can get, the team is still efficient:
|Lineup||Minutes||Offensive Rating||%0-3 FG||FG% from 0-3|
|Jennings-Ellis||2402||103.9||23.6% for Jennings, 29.7% for Ellis||50% for Jennings, 58.7% for Ellis|
|Jennings-Dunleavy||1185||105.5||28.6% for Jennings||46.4% for Jennings|
|Jennings-Ilyasova||1760||105.8||23.9% for Jennings||50.8% for Jennings|
|Ellis-Dunleavy||1332||109.2||28.8% for Ellis||66.7% for Ellis|
|Ellis-Ilyasova||1866||107||31.4% for Ellis||59.9% for Ellis|
|Jennings-Ellis-Dunleavy||724||111.1||29.1% for Jennings, 28.4% for Ellis||51.6% for Jennings, 72.7% for Ellis|
|Jennings-Ellis-Ilyasova||1540||106.3||23.4% for Jennings, 30.4% for Ellis||50.3% for Jennings, 59.5% for Ellis|
|Jennings-Dunleavy-Ilyasova||511||112.5||28.5% for Jennings||49.1% for Jennings|
|Ellis-Dunleavy-Ilyasova||566||113.4||29.9% for Ellis||63.4% for Ellis|
(Note: in those combos where Jennings/Ellis are not together (example, in a Jennings-Dunleavy-Ilyasova), it does not mean that it was exclusively Jennings-Dunleavy-Ilyasova WITHOUT Ellis. There are other lineups where Jennings, Dunleavy and Ilyasova are featured but without Ellis. So take that into consideration).
As you can see, like in Boston, both slashers have a clear priority to attack the rim when playing with a shooter. For reference, Ellis took 29.9% of his attempts from 0 to 3 feet, while Jennings 24.2% of his attempts from the same distance. Of course, the number tends to go down when both players play together (combos that have both Jennings and Ellis in them tend to lower the percentage of shots taken from close distance by only one or both of them). Both also experience slight bumps in their FG% from 0 to 3 feet in most of these combos (for reference, Jennings makes 47.8% of his shots and Ellis makes 59.9% of his shots).
Should the Jennings/Ellis combo scare us with those Holiday/Gordon or Holiday/Evans or Gordon/Evans duos?
Sacramento is a weird bunch.The biggest problem with Sacramento wasn't that their offense was horrible the team actually scored at a rate that would rank them 12th best in the league - scoring 106.3 points per 100. That remains static even with Evans (Kings scored around 106.3 points per 100 with Evans).
Nevertheless, it's interesting to see that when Evans plays with shooters (i.e. Fredette, Thomas, Thornton and Salmons according to the link above), Kings barely get any better (unlike the previous two).
|Lineups||Minutes||Team Offensive Rating||%0-3 FG||FG% from 0-3|
Just like the previous two teams, Evans does attack more playing with shooters, the most promising of which is the 55.5% in the 913 minutes together by the Evans-Thomas-Salmons trifecta. That's gotta be a source of happiness for Pels fans, who saw a ton of contested midrange shots, because we didn't have anybody who could get to the rim.
Also, excluding the Evans-Thomas duo, most of Evans' lineup with shooters features an offensive rating that was an improvement on the team's overall offensive rating. Good signs, yes?
We've finally arrived at the last team. Technically, Washington has five (FIVE) players in the list - Webster, Beal, Martin, Price and .... TREVOR ARIZA?
Kidding aside, I'm going to take Price out of the equation. That leaves us with four players to evaluate (which means so many more combinations to look at). I was skeptical at leaving Ariza there but you know, the numbers do point out he had a "shooters" season.
With Wall, the Wizards were not the Wizards - they actually scored at a pretty decent clip (105.1 points per 100). In those 1602 minutes that Wall played, he shot just 31.6% of his attempts from 0 to 3 feet (a pretty horrible number for someone like Wall), converting 57.8% of his attempts there. Did his behavior change playing with shooters?
|Lineup||Minutes||Team Offensive Rating||%0-3 FG||FG% from 0-3|
|Wall - Beal||460||110.5||33.7||69.5|
|Wall - Webster||1015||108.4||32||59.6|
|Wall - Martin||154||104.5||28.8||47.8|
|Wall - Ariza||658||103||36.8||53.4|
|Wall - Beal - Webster||303||
|Wall - Beal - Ariza||212||107||30.2||65.4|
|Wall - Martin - Webster||15||100||50||80|
|Wall - Webster - Ariza||303||110||37.3||50|
|Wall - Martin - Ariza||9||79.2||22.2||0|
|Wall - Beal - Webster - Ariza||58||125.6||26.1||50|
I can't tell for sure what type of improvements John Wall showed in 10 combinations. In 6 of them, he attacked the rim more and in 6 combinations, he made more at the rim shots but only two instances where he attacked the rim more and made more shots. You have to consider that there are two lineups that barely played: the Wall-Martin-Webster combo and the Wall-Martin-Ariza combo. So if we took those out, the numbers begin to show better results - 5 combinations out of 8 where he attacked the rim more, 5 combinations out of 8 where he made more near the rim and 2 combinations out of 8 where attacked the rim more and shot better there.
In comparison (removing low minute lineups):
- In three combinations, Rondo attacked the rim more in all three cases, made more shots from near the rim in one case and subsequently attacked the rim more and shot better near the rim in one case.
- In six combinations, Jennings attacked the rim more in three combinations, made more shots in five combinations and attacked the rim more and shot better near the rim in two lineups.
- In six combinations, Ellis attacked the rim more in two combinations, made more shots in three combinations and never attacked the rim and shot better from that distance.
- In five combinations (I removed the 15 minute stint of the Evans-Thomas-Fredette combo), Evans attacked the rim more in all five combinations, made more shots in just one combination and attacked the rim more and shot better near the rim in just one lineup. It is interesting to note that in all five cases, Reke still finishes at an insane level. Just an FYI.
I think it's painfully obvious that shooters do help open the lane for slashers - in 27 combinations, the aforementioned slasher/passer (whether it's Rondo, Jennings, Ellis, Evans or Wall) attacked the rim more than he usually does on 18 combinations, made shots better on 15 combinations and was just better overall (more shots + better efficiency) in 6 combinations. Those are crazy good numbers. I'm sure if we expanded this to a league-wide study, the number won't move as much.
And it's ok - attacking the rim more usually means the probability of creating a good shot goes way up and the probability of extending a possession also goes up (through offensive rebounds). I think that's pretty self explanatory. However, when we start talking about efficiencies, it becomes a question of efficiency vs usage - i.e. more shots near the rim means it's harder to maintain a healthy number of makes.
What does this mean for the Pelicans?
Between Anderson (career 38% makes on 5.1 3PTA/game), Gordon (career 36% makes on 4.7 3PTA/game), Morrow (career 42% makes on 3.6 3PTA/game) and Holiday (career 37.4% makes on 2.7 3PTA/game) not to mention Brian Roberts (38.6% last year) and Darius Miller (39.3% last year), the Pelicans can employ a number of lineups to surround any of the three slashers/passers (Holiday/Gordon/Miller). In fact, the Pelicans can have at least two shooters in at the same time.
Don't even get me started on how good Anthony Davis is becoming both in attacking space, reading the pick and roll and improving on his inconsistent jumper last year.
If the numbers bear anything out, the Pelicans' offense will probably be even better than last year (ranked 16th). The question, once again, shifts back to the defensive side. Can the Pelicans find a way to improve their defense to respectable numbers, to allow their probably budding offensive team into playoff contention? And can they improve it (maybe in year 2) to the point that it's a title-worthy defense (to borrow a quote)?
That's the biggest question.