3 Secrets to Better NBA Projections

3 Secrets to Better NBA Projections


Hey guys, this is Max Steinberg. I’m a partner at SaberSim and a Daily Fantasy professional, and I’m here to bring you a video about Daily Fantasy NBA or specifically about three ways you can easily add value to SaberSim’s build process. So our motto at SaberSim is build better lineups faster and that’s not just a clever saying. Our models and tools are developed so you can easily build optimal lineups for any contest and you do that in a speedy process without having to jerry rig an outdated lineup optimizer to put tons of players in groups so you have to make sure that they’re correlated or that you have the specific players that you want in your lineup to get game stacks or whatever sport you’re doing. You don’t have to do any of that. We’re going to use our simulation data to set that up for you.

And so we have simplified the process to what we call a three step build process. First you adjust your projections and ownership projections from SaberSim’s default or you can upload your own. Then you choose your build settings and we’ve made this process much easier by just … you select your contest style, the entry limit to the contest and how many entrants in the contest. And for this video, I’m probably going to just build for the standard low stakes 20 max contest that DraftKings offers every day. So this is called the four point plan NBA. And it’s different for different sports, but they usually have a pretty big 20 entry max contest. And then the last thing is you build your lineups and do what we call the quality control process, which is make sure that you’re not getting any players that you think might be misprojected.

Tune your lineups to get the proper team stacks and the proper range of exposures that you want. And then you have your lineups, you can just export them to DraftKings or FanDuel and have your lineups built. So this video is actually going to be about the first step of the process, which is adjusting projections and ownership projections. So this I think is the most fun part of playing daily fantasy.

And I think that’s what makes SaberSim so great is this is the only part that I really like to focus on. The other stuff is really tedious, right? If you’re using a lineup optimizer, trying to take every team and try to group players correctly or try to take into account correlation or fine tune these settings in a way every day is just tedium that I really don’t like. I like doing research, I like trying to figure out what players I can [inaudible 00:02:46] higher, lower, try to adjust the ownership projections.

I just want to focus on the things that I can add to SaberSim or add to the line of building process that’s going to actually make my lineups better. And so if new to NBA, this might not be that clear, right? Say if you’re thinking, okay, well how do I even adjust these projections? How do I know if the projection’s too high or too low? How do I adjust an ownership objection? And this video is mostly for people like you, right? How do I make these adjustments? How do I make them quickly and easily? And there’s actually three really simple ways in NBA that you can add values to step one of the process without actually having that much knowledge, right? So I’m going to show you them and you can do it only using SaberSim and free tools and websites out there, right?

So the first thing that you can do is look for foul prone players. So high minute players that might be foul prone, right? So how do I do that? So the first thing is you’re just going to organize this by team, right? And so we’re going to look through these teams and see which teams have a lot of players who are out. Because this is going to be where it’s going to be harder to project minutes because it means that there’s new situations on these teams and there might be something that might be off. So Boston has Kemba Walker out. That is definitely something, but we’re going to look for a team that … where there might be a lot of players out, right? So Chicago is one of these teams, right? They have quite a few players out so that’s a team that I might want to focus on.

Detroit has a couple of players out. Golden State has a lot of people questionable but not quite out yet. Indiana has Jeremy Lamb out and it’s a recent thing. And then did we miss Sacramento, where is Sacramento? There’s Sacramento. Sacramento has Richaun Holmes and Marvin Bagley out so they have a lot of bigs out. And so what I’m going to focus on, and I sort of did some pre-research here, so I would have researched Chicago a lot and I did but didn’t find anything with this specific thing, but I’m focused on two teams. One is Sacramento and one is Indiana. So with Indiana, if you look at the detailed projections of the team, we’re going to look for players that might have really high minutes. If you look at the team, the one that sticks out to me is T.J. Warren, right? He has the highest minute projection on the team.

And clearly his projection is being affected by … whoops let’s see, I’m sorry, this got unorganized, is being affected by the fact that Jeremy Lamb is out, right? And so getting 34.5 minutes requires a lot of things to go right. It requires the team to have full lack of depth, which they do, but it also requires other things like Warren cannot get into foul trouble, right? And so we want to look at this and say, well wait a second. Is there a chance that this minute projection is too high because of purely just foul trouble? And so if you look at Warren, you just go to his basketball reference page and type in T.J. Warren and we already have it up right here, but I’ll reload it and we can look at his per 36 minutes stats and as you can see … and we’re looking for players where their personal fouls for 36 minutes are over three.

That’s what I consider someone who is foul prone and T.J. Warren is certainly on that edge, right? He averaged 3.2 fouls per 36 minutes last season. The average is 3.1 for 36 minutes this season. That means to me that there’s a chance that T.J. Warren is going to get into foul trouble. And if he does, he almost certainly is not going to get 34 and a half minutes. And so because of that, I can confidently lower his projection because I think that his minutes projections is too high, right. And that’s just a really … in one minute we can do this research and discover that I can confidently lower his projection. Another player that stands out is someone on Sacramento and that is Harry Giles. So Sacramento, again we’re seeing two players are out so there is some effect here but it’s going to be harder for a model [inaudible 00:07:01]

And Harry Giles who is a starter is projected at 25 minutes, which is not that high, right. However, Harry Giles is incredibly foul prone. He averages about 6.5 fouls for 36 minutes. So most of the time that he plays over 30 minutes, he’s actually going to foul out. So he has a hard time getting high minutes. So even at 25 minutes he might be a little over projected. And so he’s someone as well that you can lower a bit and confidently say, okay, I think there’s a good chance I’ll [inaudible 00:07:35] So those are just two really quick things you can do to just see if someone might be a little over projected. Another thing you can do is look at questionables, right? And make some adjustments at the late games where there’s going to be questionable players that you might not actually end up knowing if they’re out or not.

And if these players are out, it’s going to benefit the players that are on his team … some of the players on his team quite a bit. So Victor Oladipo’s questionable, but he’s an early game so we’re going to know about that by lock time. Bruce Brown, however, is questionable and his game starts at 6:00 pacific time. And what that means is that players on Detroit, if we don’t know if he’s ruled out or not, are going to be affected if he is and there’s some chance that he’s going to be ruled out, he’s a questionable player. And so a quick and easy thing you can do is just look at players who play the same position and sort of boost them based on the possibility that Bruce Brown is out. So Bruce Brown is kind of like a combo guard [inaudible 00:08:42]

And obviously I have more knowledge of NBA so I know intuitively who he backs up and … but you don’t even need to know this. You can say, okay, he’s a point guard small forward. He’s some sort of wing player. So this might possibly affect Derrick Rose, this might positively affect Svi Mykhailiuk, this might positively affect Langston Galloway. And we can just adjust these player’s projections a little bit. And now we’re just taking into account that Bruce Brown might be out and this absolutely is going to add value to your projections. It’s accounting for something that our projections just simply aren’t accounting for because when we have a player as questionable we just pretend that he’s in, but you can add value to our map model by taking into account the possibility that he’s out. So a very easy way to do that is just by boosting these players a little bit when there’s a player who is questionable and when you take that into account that can guarantee you you’re adding value to the player projections by doing this.

The last thing you can do is do something with ownership projections. So ownership projections for a model base have a lot of factors. But they … it’s really hard to take in these sort of intuitive factors that human beings who play DFS take into account and a couple of those intuitive factors have to do with narratives, right? So if there’s a specific narrative or there’s just a high team total or something of both, that’s going to affect what a player’s ownership might be. And something that our ownership model almost certainly is not taking into account is the fact that the Lakers were [inaudible 00:10:25] the New Orleans Pelicans and it’s an Anthony Davis revenge game. And because of this, people probably are going to own Anthony Davis more than our model’s expecting because we’re not taking into account people caring about this revenge game. And it also helps that his over-under is high.

I think star players when the over-under is high usually are going to do … are going to garner higher ownership because people see that narrative and think, okay, yeah I think this player is going to do well because of this and so you can do one of two things, right? You could actually boost his projection based on the quality of that narrative and you might want to do that. I think there’s probably some factor which will make Anthony Davis a little more motivated to play a few more minutes, maybe get … maybe an extra block from extra motivation so maybe his projection is too low, right? So we can actually boost his projection. But we can also boost his ownership projection, right? He probably is going to be owned more than 60% by the field. I think he’s going to be owned 24, 30, maybe even more so we can boost this ownership projection. And add value to our model in that third way is saying, okay, this is something that obviously our model is not taking into account.

We don’t take into account narratives, but that’s something that I can take into account so I can adjust his ownership projection. I can adjust his projection. And again, it’s just another very simple way that you can add value to the model. So this all took about five minutes, right? And if you’re playing daily fantasy NBA and you’re not playing as a professional you don’t have time to research all these teams, look at box scores, so on and so forth. And that’s fine, right? That’s why we have a projection system. That’s why we have an ownership model.

The great baseline [inaudible 00:12:17] is all you need to worry about is just to find those little ways that you can add value to our projection system and ownership. Let SaberSim build those lineups for you from those projections and ownerships. And then you’re going to have great lineups that have a good chance to win [inaudible 00:12:38]. So, I hope you enjoyed this video. I think adjusting projections ownership can be daunting, but it can actually also be really simple if you keep it simple. And even if you’re more of a football person or a baseball person, I think you can find ways to actually be good at daily fantasy basketball, even if it’s not your main thing. Thanks for watching.


Start building better lineups right now!

Start My Free Trial