Thursday, Sep 12th

Sunday, Sep 15th — Early

Sunday, Sep 15th — Late

Monday, Sep 16th

This course will focus on NBA strategy — how to approach both cash games and tournaments. The primary focus here is on cash games and on the factors that go into projections, to help build a deeper understanding of what a projection system actually is and does, and how to think about it critically. There are a lot of great projection systems out there that can be used to play NBA DFS, but what isn’t really in the market right now is much content about what to DO with those projection systems. How should we approach cash games? Just shove in the highest projected lineup and cross our fingers? Or are there other ways we can gain an edge? And in tournaments, should we just tell the optimizer to run 150 lineups and then print money? Of course not…so how do we find edge in tournament play?

I’m going to assume here that you already know the basics of NBA DFS — what the positions are, how scoring works, etc. If you’re an NFL DFS player looking to try your hand at NBA DFS, I would recommend Hilow’s “Transitioning from NFL DFS to NBA” course as a great introduction.

I’m also going to assume that you have an optimizer. Unlike NFL, playing NBA DFS without an optimizer is much harder. NBA DFS is all about math, and you have to be a lot more price sensitive (at the end of the day, a guy’s points projection boils down to “expected per-minute points production times expected minutes = expected points”), which makes it much harder to build optimal lineups by hand. If you want to try, I wish you good luck (and I do know a couple of guys who’ve had sustained success by doing this) — but for most of us, it’s far, far easier to use an optimizer. OWS, of course, has a deal with FantasyLabs to get access to their projections and optimizer for NBA DFS and I personally recommend it. I used DailyRoto for years, but when we started exploring this partnership with Labs they gave me access to check it out, and I got hooked. I switched over and haven’t looked back.

NBA cash games can, in my mind, be a predictable source of profit throughout the season. There’s a slate just about every day, and on some days there are multiple slates. NBA scoring is more predictable than any other sport, which makes cash a much more consistent way to profit. The reason for this is sample size: in baseball, a hitter is going to get 4 or 5 at-bats per game. In football, a receiver might see 6 or 8 or 10 targets. But in basketball, a player can touch the ball far, far more often; a stud player is likely to take 10+ shots and also have opportunities for assists, rebounds, steals, and blocks. Each of these “events” only scores a small number of points — there’s no 90-yard touchdown or grand slam in NBA. Because of this, the variance in scoring is lower. It’s primarily driven by opportunity: “how many minutes is a guy on the floor?” is the most important factor (and who’s on the floor with him — after all, there’s only one basketball, so sharing the court with a bunch of other high-usage guys means there isn’t as much opportunity to go around). You don’t need the “big play” to go your way, you just need consistent per-minute production.

Cash games, of course, are all about floor. Because of how predictable NBA scoring is, they can be VERY chalky, with the “best plays” often coming in at 70-80% ownership. What’s important to understand here is what the field will do: most NBA DFS players will run their optimizer of choice, take the highest projected lineup, and put it in cash games. Maybe they’ll do a tweak here or there, but that’s how a lot of the field approaches cash games. You’ll see ownership play out this way — the guys who consistently show up in the top 20 or so lineups for the major projection sites will be the highest-owned cash plays. But, there are ways to think about cash in a more nuanced way to give yourself an edge (and if you don’t believe me, go look at the cash game lineups of some of the top DFS pros; you’ll see that it’s common for them to have some “off the board” plays in their cash lineups).

Projections are fantastic tools for NBA DFS (in fact: along with an understanding of how to play NBA DFS (which you’re picking up in this course and Hilow’s course), and with a community to help carry you to lock (which you’ll have through OWS Discord), projections and ownership projections are about 90% of what you need). What you need to understand, however, is that projections are median outcomes. This means that, even if the projection is perfectly accurate, half the time a guy’s score will be under that number and half the time it will be over. And, most importantly, not all medians are created equal. If I can get into stats-geek mode for a minute: what projections try to model is a range of outcomes for a player, and then the number they show you is the median. But, let’s do some math. Let’s say a player A’s range of outcomes is between 20 and 30 points. And let’s say player B’s range of outcomes is between 10 and 40 points. What are the median outcomes for each player? Well, 25 is the halfway point in between each of those projections — but which player would you prefer in cash? The guy who might get you 40 but also might get you 10? Or would you prefer the guy whose ceiling isn’t as high, but who is almost certainly getting you at least 20? I know that in cash lineups, I’d want the player with the higher floor, because in cash we should care more about floor (note that here I am talking about double-ups and 50/50s, where your goal is to beat half the field; if you play a lot of head to head, more ceiling is useful!). In addition to thinking about ranges of outcomes, we also need to think about where these projections can be wrong — and if they’re wrong, in which direction are they more likely to be wrong? Up or down? *The key question that you need to answer in order to successfully approach NBA cash games is “how much median projection am I willing to sacrifice in order to decrease the variance of my lineup?”* That is it, in a nutshell: you can just play the highest projected lineup and you’ll win sometimes, but if you can sacrifice a very modest amount of median projection in order to decrease your lineup’s variance, you will win at a higher rate. And, the right answer is different depending on contest type: in double-ups or 50/50s, your only goal is to get over the cash line, so you should be willing to sacrifice more median in order to reduce variance. In head to heads (or, say, triple-up contests or 3 or 5-man leagues), you’re rewarded for finishing higher, so you should be willing to sacrifice less median; you WANT the variance in those types of contests, because you get rewarded for finishing higher. Master this, and you can master NBA DFS cash games.

Next: Demystifying Projections »