Welcome back, as I continue the journey from the top down. Last week, I covered roles, formations, positions, etc. in the modern NBA. That provides the groundwork for just about everything else I’m writing, so if you missed it, go back and check it out. I will be writing an updated version of the post at some point in the future (next week?) thanks to the suggestions I’ve gotten from people on reddit and twitter, so thanks to you guys for that.

So, if last week was the NBA on a team level, it only makes sense to look at the NBA on an individual level. Basically, the goal of the draft is to find quality NBA players, right? Well, define a quality NBA player. It’s not easy. Some people default to advanced statistics like win shares or adjusted +/- or on/off. Some people default to more basic stats, like points and rebounds. This is not going to be about the best way to determine player quality, because quite frankly, ultimately, it’s pretty subjective and even I don’t think I have anything close to a decent grasp of it. That being said, I do think there are certain factors that are often lost in player evaluation, especially at a micro level, starting with the one most related to last week’s article. There’s also some lessons to take away here for the draft. This is a draft series after all, right?

Team Scheme Really Really Really Matters

I really cannot stress this enough, which is why I’m starting with it. Each player brings individual skills to a team, but basketball is very much a team game on both ends of the floor, and it simply isn’t possible to evaluate players in a vacuum. Very few teams are equal to the sum of their parts. Through roster construction, offensive and defensive schemes, and on-court lineups, teams can find themselves far better or far worse than expected, with the individual stats suffering and reflecting the same.

A team that does not have enough shooting will make everybody look much worse. Players who feast on inside shots will not be able to get clean open looks. Players who can’t create shots for themselves won’t be able to get open. Contested shots are much harder than uncontested shots, midrange shots are much harder than shots at the rim, etc. If a team can simply pack the paint and leave one or two guys glued to the only shooter(s), the quality of shots goes down, the offense goes down, and the individual numbers plummet. A player who does not fill a clear role in an offense is automatically a problem, because a team needs to figure out how to create shots in spite of that player.

Keep in mind that if you can’t figure out what a player’s role is on offense, he’s probably not going to be particularly good. Michael Beasley and Jabari Parker come to mind as two highly ranked draft prospects who really strugged/are struggling because they have no clear role on an offense. But playing outside your role or not being allowed to utilize your strengths can be just as damaging. A 3 point shooter forced to play in the Doug Collins/Byron Scott 2 point jamboree will look like a really bad player, even if he is extremely good. A driver asked to play off the ball is going to look much worse than when he’s allowed on the ball. This misutilization of talent occurs far too often, and identifying when it’s happening is a great way to identify talent that may be undervalued.

The effect is the same on defense. Whether you like advanced stats or just plain old conventional wisdom, it’s clear that it’s easier to play defense with a good rim protector behind you. If you play on a team without a rim protector, your individual defensive effort is just going to look worse, because every mistake is magnified, as there’s nobody behind to fix it. If you play on a team with an elite perimeter defender, your defense will look better because it’s easier to defend weaker players. Schemes designed to limit opponents to mid-range shots and avoid fouls will make individuals look better than schemes designed, well, any other way.

Draft Lessons: A potentially great player drafted to the wrong roster or scheme will look terrible. A low-ceiling player drafted to the right roster or scheme can end up looking far better. There’s a reason certain teams are great year after year after year and certain teams always fail to live up to expectations. Players need to be put in to positions where they can succeed, otherwise, they can be failures through no fault of their own. Basketball is so cruel, ain’t it?

Consistency and Opportunity

Every player on an NBA roster, as well as many on NBDL rosters and top-class non-American rosters, is capable of having a good game in the NBA. It’s true. These guys are all amazing athletes who have spent thousands of hours of their life getting where they are. A retired Brian Scalabrine, the human victory cigar himself, can still completely crush even the best non-pros. So what separates the good players from the bad? Consistency.

Far, far, far too often an argument is made that a player “can” do something. Andre Drummond can shoot 3s, and he can shoot them far better than 99+% of the planet. He can make 3s. But you don’t want him taking 3s, because he can’t make them often enough in game situations. And ultimately, we have lots of stats that tell how well a player does something – FG%, 3P%, FT%, FTr, REB%, and so on and so forth. But strange things happen in small samples – whether that sample is one game or ten games or even thirty games. Now it’s certainly fair to judge whether a player was good or bad within a given sample, but be very careful when using small samples to project future performance or making value judgments about a player. It is NOT always reasonable to say “well, if he can do it over X games, he can be that player all the time!” It just, well, doesn’t work that way. Players will go hot and cold, players will have good games and bad. An inconsistent player is simply not a starter, and an inconsistent player who is bad more than good is not a rotation player.

Player A: 36 points, 7 assists

Player B: 35 points, 6 assists

Player A is Jordan McRae on 4/13/16. Player B is Andrew Wiggins’ highest point total and highest assist total of the year. There’s like a billion ways you could poke holes in why those aren’t equivalent, but it makes a simple point: it takes more than talent to be good. It takes opportunity, the opportunity to get playing time, the opportunity to get to shoot, the opportunity to get the ball in your hands. Remember LINSANITY!?!?!? He played 300 mainly terrible minutes for the Warriors, who cut him. He got a second opportunity and made the most of it.

See that above part where it’s really hard to judge players off of small samples? Well, a lot of players simply never get enough of an opportunity to show their talent level. And quite frankly, many players get far more opportunity than they deserve, even after proving they’re not any good. The NBA is not a meritocracy – factors such as where a player was drafted, how much they’re being paid, and name value influence playing time, driving teams in to the ground as they try to force things that aren’t working. Just remember when evaluating players – look for enough sample size, but be quite wary of accepting or dismissing a player solely because of minutes.

Draft Lessons: Don’t judge or project a player based off only good games or bad games, judge on the whole body of work when possible. Don’t judge on draft position, judge on what a player does on the court. Don’t judge on whether a player gets opportunities to play or not, teams often make baffling personnel decisions based on lots of factors other than talent, and some players who never get a chance are better than some highly drafted players.

Let’s Get Advanced

So, gonna take a trip in to the world of advanced statistics here, which requires a stop in baseball world. I think most people reading this article have at least heard of baseball’s WAR stat, a kind of catch all stat to reflect the value of a player. There are four components to WAR: Offense, Defense, Baserunning, and Positional Adjustment. It’s easy to forget about that last one, and it makes a huge difference. In FanGraphs WAR, a catcher is worth three full wins more than a DH just due to the positions they play. Unsurprisingly, the same type of adjustment is needed for the NBA. Surprisingly, no advanced stats actually seem to utilize one. Maybe it’s because of the difficulty in determining positions. I really don’t know why, I don’t create the stats, I just use them.

(All stats and positions here taken from basketball-reference with an 820 minute minimum except for RPM taken from ESPN)

PER Top 25 2015-16: 10 C, 6 PG, 4 PF, 3 SF, 2 SG

OWS Top 25 2015-16: 7 PG, 5 SF, 5 SG, 5 C, 3 PF

DWS Top 25 2015-16: 7 C, 6 SF, 5 PF, 5 PG, 2 SG

WS/48 Top 25 2015-16: 8 PF, 7 C, 5 PG, 3 SF, 2 SG

OBPM Top 25 2015-16: 12 PG, 6 SG, 6 SF, 1 C, 0 PF

DBPM Top 25 2015-16: 12 C, 7 PF, 4 SF, 1 SG, 1 PG

BPM Top 25 2015-16: 7 C, 6 PG, 5 PF, 4 SF, 3 SG

ORPM Top 25 2015-16: 11 PG, 6 SF, 5 SG, 2 PF, 1 C

DRPM Top 25 2015-16: 17 C, 5 PF, 2 SF, 1 SG, 0 PG

RPM Top 25 2015-16: 8 C, 5 PG, 4 SF, 4 PF, 4 SG

You can dig around the numbers a little more yourself, but certain things will hold true. Plus/Minus stats give centers far more defensive value and other bigs more defensive value than guards, while giving primary ballhandlers far more offensive value than any other type of player. Also, it’s just really hard to be a SG that any advanced stat likes in any way. Remember this when comparing NBA players using advanced stats – only compare within positions. When I have more time (aka, sometime over the offseason), I hope to complete some type of positional adjustment chart for some advanced stats. If you’re a mathy type of person and want to help out, let me know.

Anyway, I think it would be short-sighted to say “we need positional adjustments” without asking why. The data certainly points in certain directions – centers make the biggest impact on defense by far, primary ballhandlers make the biggest impact on offense by far. How does this change how to value positions? Well, as far as the draft is concerned, you’re more likely to hit on a primary ballhandler or a big over a wing, but wings have more value over replacement because there’s less high impact ones. There’s also just more midlevel bigs than there are non-bigs, so after the top bigs, there’s just not a lot of value to finding mid-level bigs, whereas there is significant value in finding mid-level non-bigs.

But it’s also worth realizing that, if centers make the biggest defensive impact and primary ballhandlers make the biggest offensive impact, it’s quite problematic having a bad defensive center or an inefficient primary ballhandler. Going back to baseball statistics, while WAR uses positional adjustments to compare players between positions, defensive metrics such as UZR and DRS compare players within positions, such that a player who is good for his position gets a boost and a player who is bad for his position gets knocked down. We will have to settle for a simpler statement: your center better be able to play D and your primary ballhandler better be pretty good on offense or you’re gonna have a bad time.

A quick note on On/Off since I mentioned it above – if you look team by team, you’ll find that very few rotation players have big on/off numbers one way or the other.  It’s mainly true superstars (positive) and their backups (negative), good players (positive) with no backups, players that don’t fit in to modern offensive and/or defensive schemes (negative) and two players who play the same nominal position but have very different roles, such as a stretch big (positive) and a non-stretch big (negative).  There’s also a lot of noise, especially comparing starters with backups.  That being said, it’s still a useful stat for a variety of things, especially providing some context to other stats.  Just be careful with it.

Draft Lessons: Wings have higher risk but higher reward. Bigs have lower risk but less reward. Lead guards are somewhere in the middle. It’s generally not worth taking mid-level bigs because there’s so many of them. When using advanced stats to look at prior draft classes to see how teams drafted, be very careful because you cannot compare between positions using current advanced stats. Pay much more attention to defense for centers and offense for primary ballhandlers, their opposite is far less relevant to their quality at the next level.

Well, that brings us to the end here. As always, please send suggestions, thoughts, criticisms, or whatever else my way, I really do read them and take them into account. Til next week!