Before we jump into everything we’re going to cover in this course, let’s first define a clear goal we wish to achieve by the end of our sessions. The goal of this course is to provide the framework needed to be a successful NBA DFS GPP player, built for those who have either never played before or are hoping to take NBA DFS more seriously moving forward. We should treat this course as our first building block, our foundation, upon which everything else is built and added to for our NBA DFS journey. For a more in-depth breakdown of specific strategy as it relates to cash games and GPP play, I invite you to check out Xandamere’s course available in the OWS Marketplace.
The goal of this course is to provide the framework needed to be a successful NBA DFS GPP player
In NFL DFS, we routinely talk about things like running back opportunities, targets, pace of play, DVOA, matchup, etc., as these are the metrics that most commonly correlate to DFS scoring and points from a predictive standpoint. Even on a full PPR site like DraftKings (while we’re here, the remainder of this course will be written utilizing Draftkings scoring and roster construction; the overarching strategy themes remain relatively constant from one site to another, but the scoring differences and roster requirements are tangible enough to where there are very clear differences in the most optimal ways of attacking a roster), touchdowns provide the most points for a single act. In NBA DFS, there are a vast multitude of ways for players to accrue points, including points scored, assists, rebounds, steals, blocks, and even losing points through turnovers. In this course, we’ll first talk about the basics of NBA DFS (roster positions, scoring, bonuses), the similarities and differences of NFL and NBA DFS (the importance of a bottom-up build, utilizing roster builders/optimizers, ceiling and floor), dive into the predictive metrics used in NBA DFS, talk about the importance of lineup flexibility and breaking news, discuss optimal roster construction, and wrap it up with some NBA DFS Game Theory.
So how do players accrue points in NBA DFS? As we talked about earlier, there are more ways for a player to accumulate points in NBA compared to NFL, including one point (1) for every point scored, two points (2) for every steal (change in possession credited to a single player), two points (2) for every block (a blocked shot from a player on the defensive side of the court), one and a half points (1.5) for every assist, one and one quarter points (1.25) for every rebound (gaining possession after a missed shot), half a point (0.5) for every three-pointer made (so for every three, we get 3.5 points total, one for each point and half a point for the made three-pointer), negative half a point (-0.5) for every turnover (loss in possession, either through a steal from the other team, an offensive foul, or being the last player to touch the ball prior to it going out of bounds) and bonuses that are awarded for a player reaching certain milestones (double-double and triple-double, one and a half points (1.5) bonus for a double-double and three (3) points bonus for a triple-double, which are aggregate, so if a player achieves a triple-double they would receive four and a half (4.5) additional points through bonuses; a double-double is defined as reaching double digits in two of the five main statistical categories (points, assists, rebounds, steals, blocks) and a triple-double is defined as reaching double digits in three of the same categories). Finally, similar to NFL, a team’s players do not lock onto our roster until their respective game tips off, which we will discuss further in a later lesson.
Let’s talk quickly about the different positions on the court at any given time for an NBA team (we’ll do this in order of importance from a DFS perspective) and compare those positions to the NFL. NBA teams utilize five primary positions: Point Guard (the one), Shooting Guard (the two), Small Forward (the three), Power Forward (the four), and Center (the five).
Key Examples :: Joel Embiid, Nikola Jokic, Rudy Gobert, Karl-Anthony Towns, Bam Adebayo
Centers are typically the biggest players on the court and the player most commonly closest to the basket. They derive most of their value primarily through points scored, rebounds, and blocks. Because the majority of centers in the league rarely contribute to the other two categories, those who do are of high importance, leading us to think about centers as we do running backs in the NFL. NFL slates are commonly won and lost (at least, it’s much harder to cash if you don’t get the position right) at the running back position, and we place a premium on workhorse running backs who can accrue points in more than one way. The same goes for centers in the NBA, as we should place a premium on those who can earn points in more ways.
Key Examples :: Stephen Curry, Kyrie Irving, Damian Lillard, Luka Doncic, Russell Westbrook
The second most important position on the court (and this may be contrary to what other analysts will say, as most will argue point guard is the most important position on the court, which is true from a real-world perspective, but not for daily fantasy!) is point guard because it is typically the position that sees the most usage (we’ll define this predictive metric in the next lesson!). Point guards can be compared to quarterbacks in the NFL, as they are the player who touches the ball the most (if not every possession, damn close to it). Point guards typically accrue fantasy points through points scored, assists, and steals, with a few select PGs who are capable of consistent rebounds and/or blocks. Again, a premium is placed on those who can contribute to more than three statistical categories. Think of these players as the quarterbacks in the NFL who bring rushing upside to the table, as they typically carry a higher raw floor and ceiling compared to those who don’t.
Key Examples :: Giannis Antetokounmpo, Anthony Davis, Jayson Tatum, Kristaps Porzingis, Zion Williamson
Next up in order of importance are power forwards. The NBA game has evolved to where PFs come in a few different shapes, sizes, and skill sets. The most common are tall, athletic big men who do most of their work in the paint, deriving most of their value from points scored and rebounds. As the game has evolved, we’ve seen the emergence of perimeter power forwards who have a soft shooting touch; these players are commonly referred to as “stretch fours,” or power forwards who stretch the court. Again, emphasis is placed on the power forwards who can contribute to more statistical categories, with some adding the ability to consistently contribute to blocks and assists. Power forwards are akin to wide receivers in the NFL, where some are high aDOT, splash play wide receivers and some are low aDOT, possession-style receivers (PFs who contribute to points and rebounds and work primarily in the paint typically have a great chance at the double-double bonus, making them high floor plays, while PFs who are used to space the floor typically have more of a boom-bust nature to their game; throughout the season (especially as you hang out in the OWS Discord as you prep for each slate), you’ll quickly gain an understanding of which players fit into which specific categories here)).
Key Examples :: Kevin Durant, LeBron James, Kawhi Leonard, Jimmy Butler, Paul George
Small forwards can be thought of as a mix between PF and SG, typically seeing low usage. They derive most of their value from scoring and rebounds, but normally have a lower percentage chance at the double-double bonus, making them lower floor plays. Again, emphasis is placed on those with higher usage and/or those who can contribute to other statistical categories. SFs can be compared to tight ends in the NFL, where the elite ones are slate-breakers and the rest are rather tightly compacted from a floor and ceiling standpoint. As such, decisions on how to handle the position as a whole need to be made on each slate.
Key Examples :: James Harden, Jaylen Brown, Bradley Beal, CJ McCollum, Klay Thompson
Shooting guards regularly carry the lowest usage on the floor, primarily skilled at pull-up shooting and off-ball offense (meaning their movements on the court are designed to get them open shots). There are very few SGs who have high usage (the ball in their hands a lot); these players are typically high-priced as they are few and far between. Because SGs typically carry low usage, their scoring is mostly high variance, relying on actual scoring to derive value. Shooting guards are akin to defense in the NFL (highest variance position in NFL DFS), as every so often you’ll see a shooter get hot and post a ceiling game seemingly out of nowhere.
The standard lineup utilized by Draftkings is one of each position (PG, SG, SF, PF, C), one “Guard” utility position, one “Forward” utility position, and a flex (can be any position), which looks like this:
Big picture differences for NBA are we now are playing only eight roster spots, compared to nine for NFL. That means that instead of $5,555 available per roster spot (as we have in NFL), we now have $6,250 to spend per roster spot (same $50,000 salary cap, divided by nine for NFL and eight for NBA). Since floor is much flatter in the NBA when compared to the NFL, we can think of players priced above $6,250 as the “studs,” and players priced below $6,250 as second or third tier when it comes to potential ceiling (we will discuss why this is the case and how to leverage this truth in a later section!).
Similar to NFL DFS, where we start the lineup building process from a bottom-up approach: identifying what value is available first and foremost is of utmost importance in the NBA. By identifying the value available to us on a given slate, we can then decide where else to allocate salary. One thing that is different in NBA when compared to NFL is that in NFL DFS, player pricing is typically a reflection of floor (the higher the price, the higher the floor), whereas in NBA, player pricing is typically more closely tied to ceiling. This idea is primarily linked to the fact that players have a multitude of various ways in which they can contribute points in the NBA, and almost every player brings a “standard range of outcomes” floor of a 4x multiplier. 4x puts you on pace to take down GPPs in NFL, but in NBA DFS, you’re looking for 6x at a bare minimum for every player on your roster, as the scores required to take down GPPs are regularly between 350-400 points (quick side note: the idea of salary multipliers can be applied to players priced between $5,000 and $9,500, generally; players at pricing extremes should be treated a little differently, which we’ll cover in a later lesson).
When we’re talking about values in the NBA, typically we’re referencing players at sub $5,000 in salary, and a general rule for any player in this range of pricing is we need 30+ points for them to make sense for GPPs (again, due to the heightened scores required to finish in the top 1% of GPPs in NBA). So, when we have a value player open up, like NFL, the first thing I’m doing is classifying them as either “Good Chalk” or “Bad Chalk” based on the chances of them hitting 30+ points, regardless of their salary as long as it falls below $5,000 (any player priced from the bare minimum of $3,000 up to $5,000 I disregard salary multiplier and instead reference 30 points; if their range of outcomes is biased to hit 30+ points, they are considered good chalk, if it is not, they are an easy fade).
The point of this section is not to lobby one way or the other when it comes to using solvers to generate lineups, but to instead highlight the importance of being familiar with them in the NBA as a means of differentiation. Because median projections in the NBA are more reliable than in other sports (basketball can be considered a lower variance sport when compared to football, baseball, hockey, and soccer, as there are fewer random ways in NBA for a player to get a bunch of points at once (think a home run in baseball, or a touchdown in football, as plays of outsized importance compared to other means of scoring)), NBA was the first to be “solved” by lineup builders and optimizers (as in, the people who were able to build high confidence projections and implement them using solvers were the first to grind consistent profits in NBA cash games, a process that has remained to this day). The carryover into today’s NBA GPP scene has meant high overlap amongst rosters in high-dollar and/or low max entry contests, so being familiar with the median projections fed into lineup optimizers gives us some solid leverage in these contests. Creating leverage through this understanding can be done manually or through manipulating median projections prior to running optimizers, and either way has their own merits. How you ultimately choose to create this leverage is up to you, but it is one of the best and quickest ways to gain leverage on the field in NBA.
If you’ve followed me or my work for any period of time, you should recognize this next bit, but I cannot stress enough the importance of looking at floor and ceiling through the lens of your roster as a whole as opposed to individual players. The shift to this mindset creates a multitude of benefits, the clearest of which gives us a stable cash rate without sacrificing our opportunities to take down GPPs. Explained a little more in-depth, we know that in order to hit top 1% scores in GPPs we’re going to need some differentiation, but instead of introducing suboptimal plays (which would by definition lower the floor of our roster as a whole, reducing our cash rate), we piece together players that give the roster the best mix of floor and ceiling. This process takes time and practice to master but is a viable carryover from NFL to NBA! In the following lessons, we’ll discuss some of the more optimal ways to approach floor and ceiling in NBA DFS.