4 Keys to Dominate NASCAR DFS

NASCAR DFS is one of the most fun sweats in DFS, but it can be a tricky sport to get right. And it's easy to make expensive mistakes if you're not careful. My name is Jordan Chan. I'm the head coach here at SaberSim. And in this video, I'm going to be giving you my four keys to crush NASCAR DFS for the 2024 season.

Whether you're brand new to NASCAR and looking for something to sweat on Sundays, now that football is over, or Or you're a seasoned veteran ready to come back and crush the 2024 season. These four tips will have you well on your way to building winning lineups for this year. Let's go ahead and dive right in Now the first key is to understand what upside looks like at a given race or at a given track.

In NASCAR DFS, racers primarily score points in one of two ways, or they get upside in one of two ways. The first is by finishing a lot higher than they started. This is often called place differential. So if you start in the back, you're starting at 35th or 36th and you end up finishing in the top five, you're going to make a ton of points for that finishing at differential there.

The other way you can score a lot of points in NASCAR DFS is leading the race for a long period of time, racking up points for laps led and fastest laps. Now the reason this matters is because different races throughout the year are going to benefit different strategies or different kinds of upside in different ways.

In some races, it's very hard to pass. There's not going to be as much action. It's much harder for drivers in the back to make up a lot of place differential and drivers starting in the front of the race are in a really good position to dominate the race from start to finish, leading the majority of laps and racking up a ton of fastest lap bonuses.

In others, like Daytona coming up this Sunday, the race is chaos. It's very easy to pass. There's going to be a ton of changing. The race becomes very high variance and a lot of times the optimal strategy becomes stacking the back, playing a lot of drivers in your lineups that are starting at the back because they're the most likely to rack up a lot of place differential points.

It's very hard for any of the drivers starting in the front in Daytona to maintain that position for the entire race anyway, so you might as well play the drivers that have the upside from place differentials starting at the back of the race in your lineups. What this means is that every event and every slate in NASCAR DFS is very different from each other.

You can't apply the same strategies throughout the entire season for every race and expect to be successful. You need to be building lineups that have upside for that particular race on that particular track. At SaberSim, we solve this problem by using lap by lap simulations of each race. If you're building on SaberSim, as long as your settings are set to sim mode, and you're building on SimDiversity 10, every lineup you get using our tools will be the optimal for an actual way that race can play out on that particular track.

for that event. We're not just optimizing your lineups based on average projections, that can be misleading and underestimate the different types of upside for that particular track. Each of your lineups is actually the optimal for a given race simulation, which means we're going to be picking up on the races where it's a little bit better to shoot for those place differential points and stack the back like Daytona, and we're also going to be picking up on those races where it's very easy for the guy that starts in the first, second, or third position to dominate the entire race.

If you haven't used SaberSim before for NASCAR DFS, it is an extremely powerful tool to build you optimal lineups right out of the box that have upside for that particular track in ways that other optimizers cannot. But even if you're not using SaberSim for your NASCAR DFS lineups this season, understanding the track and race dynamics of a given event is the most important thing in NASCAR DFS, which is why it's my first key to success this year.

Now, my second key to crush the NASCAR DFS season in 2024 is to avoid duplication in your lineups. NASCAR DFS is a lot like NFL Showdown. You're building a lineup of six players, using a relatively small player pool, and often competing in some large contests, which means people are going to play the same lineups as each other.

We almost always want to avoid duplication when we're playing GPP in DFS. The more people sharing our lineup, the more people we have to share the first place prize with when our lineup takes the contest down. But this generally becomes a bit of a balancing act. The field is often directionally correct with what lineups they identify.

are a little more likely to take first place down. So as lineups get more and more duped, those are lineups that are also a little bit more probable to win first place in your contest. We don't want to play terrible lineups that have no shot at winning just to say we played unique lineups, but we also don't want to play lineups that are being duped hundreds or even thousands of times just to get a little bit of a higher percentage of that lineup taking first.

We want to play lineups that have win equity, that can win the contest. but are also not going to share the first place prize with very many lineups when they do win. Now, there's a couple different ways that you can handle duplication in NASCAR DFS. One of the ways is to cap your maximum salary you're allowing for your lineups.

Most optimizers and people building by hand have a natural tendency to use the maximum salary or very close to the maximum salary when building their lineups, but there are plenty of lineups in NASCAR DFS. Just like NFL showdown and other heavily duped sports that can still win. And those lineups are a little bit disproportionately showing up in your contests.

Now, one note on NASCAR specifically is salary can often be very loose in different events. The top projected lineups might already leave a ton of salary on the table. So if you sit down to build your lineups, and you cap your maximum salary at 49, 000, for example, that might not actually reduce your duplication at all if the top cash optimals are only using 48, 000 or even 47, 000 salary already.

What I would recommend doing is when you sit down to build your lineups for a NASCAR slate is first just build the top cash optimals, the top lineups by projected score, and get a sense of how much salary those lineups actually use. You can do this on SaberSim by changing your build mode over into optimizer mode and just building say a hundred or so lineups.

So you get a sense of what the top projected lineups. actually use in terms of salary. From there, if the top projection lineups are using 48. 5 salary, you can drop your maximum salary down to 47, 000 for example, so that you know you're still getting a little bit of a salary disparity there between the top projected cash optimals, the lineups that are likely to be very popular in contests, and the lineups that you're choosing to enter.

Another way to avoid duplication in your lineups is to limit the overall product ownership or the geometric mean of ownership in your lineups. Now, bear with me, this can get a little bit mathy. I'm going to link a video in the description of this one that goes much more into detail in the math side of how product ownership or geometric mean ownership work, but I'll show you everything you need to know here.

To implement this in your process, the idea behind ownership product or geometric mean is that you can roughly calculate how often a lineup is likely to be duped by thinking about the ownership projection of each player in your lineup as the probability of that player showing up in a given lineup.

You can then take the product, multiplying all of those different probabilities together of each driver in a lineup to assess the probability of that lineup getting played. Then you can multiply that probability of that lineup getting played by the number of lineups that are going to be in the contest and get a rough approximation of how likely that lineup is to be duped.

Now, I'm going to give you the formula and show you exactly how to use this in your lineups on SaberSim. Obviously, this is going to look a little bit different for other optimizers out there, but this will show you exactly how to implement this in your game on our tools here. The equation that you're looking for first starts with the number of desired dupes or expected dupes over the number of players in your contest.

So if I'm playing for the Millie maker here this weekend, it's about 106, 000 lineups here. And I want to try to make each of my lineups unique if possible here. So my desired number of dupes is one, and I'm going to start by dividing that by 106, 000 here. I'm going to take this number here and I'm going to put it to the power of one over the number of players that you play in your lineup, which is six.

You play six drivers in a lineup. So I take this number here and I put it to the power of in parentheses, one divided by six. This is the desired, the maximum geometric mean of ownership I want to play in my contest. And on Saberson, we represent this, uh, as the percentage here. So it would be 14. 5 percent geometric mean of ownership is my maximum.

Now, again, if you are interested in more of the math side of how this equation comes about here, you can check out the video in the description, but let's go ahead and implement this here into our lineups now. Now, the first thing we need to do is create a custom metric here that shows the geometric mean of each lineup in your pool.

Now, one thing to note, this feature is only available on pro plans or up. So pro and ultimate on SaberSim here. So you will need to upgrade to one of those plans here to get access to these custom metrics, but we'll go ahead in our sorting method here and add a metric. I'll call this GeoMean. So we know what it is here.

And we're just going to say that this is the ownership projection. It is the geometric mean of ownership, and we want the actual value of that for each lineup. We can save this metric here, and that will allow us to filter on these metrics here. So we can add a filter, and we're going to say, hide all lineups where the geomean is greater than the number we just calculated.

And this is now going to filter out a little over 1100 lineups from my pool. Those are all lineups that per geometric mean of ownership are expected to be duped or played by somebody else at least once here. And now we've removed those lineups from our pool. And finally, the best way to avoid duplication or to manage the problem of duplication appropriately in your NASCAR contests is to use a contest sim.

The way a contest sim works is it takes your lineups that you're considering playing in a given contest and it simulates those lineups in what we expect the field to look like on race day and see what is the actual expected ROI over of those lineups over the long term if the race played out over and over again.

The reason this is the best tool for navigating the problem of duplication is that instead of trying to make an arbitrary rule of thumb of about how many dupes is acceptable in a contest, this just directly sees the ROI of a lineup given the number of dupes that are expected. Some lineups are still going to be profitable if they're duped three, four, five times, and other lineups are only going to be profitable if they're completely unique.

A contest sim goes straight to the source and directly calculates the dupes. It's the expected ROI of your lineups and uses an actual expected field to determine how often a lineup is likely to be duped. SaberSim Ultimate allows you to run a contest sim on each of your optimal NASCAR lineups that we've built using the lap by lap simulations.

We've got what we expect the field to look like for a given race, you can program the contest size and the payout structure of the contest that you intend to play, and then run your contest sim to see how your lineups would perform in this contest with what we expect the field to look like over the long term.

Once your contest sim is complete, you can sort lineups by the highest expected ROI, which is already capturing how likely a lineup is to be duped, and how often a lineup is winning in your contest. When it comes to navigating the problem of duplication in NASCAR DFS, I'd recommend using some combination of all of these different tools that I've presented here.

If possible, use a contest sim, to simulate your lineups against an expected field to see which are the most profitable, limit your maximum salary to avoid playing lineups that hand builders and other optimizers are disproportionately likely to favor, and use geometric mean or product of ownership to reduce lineups that are just mathematically a little more likely to be duplicated based on the ownership projections.

Now my third key to crush the NASCAR 2024 season is to embrace the variants. NASCAR DFS is a very high variance sport, which you can probably expect, we're racing cars here. Expect wrecks, expect engine failures, expect other reasons that your drivers don't finish a race, and honestly, expect a lot of things that you just don't really have a lot of control of and that are very hard to predict.

You're not going to profit every race, and you're certainly not going to win every race. But think about it like NFL Showdown. Your goal is to play in a way where you can survive the season, giving yourself time to have that one big race where all the chips fall in your favor, and you're able to realize most of your profit on that given event.

You should be using strong bankroll management principles and contest selection frameworks to make sure that you're playing at NASCAR DFS in a way that allows you to weather the swings of this high variance sport and make it to the end of the season to realize your EV over the larger sample size. If you haven't seen it already, you should check out our DFS profit plan video, which is a bankroll management and contest selection framework you can follow for the NASCAR DFS season to make sure you're playing sustainably.

Now, one NASCAR specific tip to add to the DFS profit plan, if you've already seen it before, is to play the Trucks and Xfinity races. There's often two other races for most weekends here, Trucks on Friday and Xfinity on Saturday nights, which are just smaller races, generally smaller DFS contests, but our sims and tools on SaberSim are just as powerful for these races, And it allows you to play three races a weekend instead of just one playing more races over the course of a season will lower the variance.

It will give you more data in your sample, more opportunities to have a big win in one of these contests. Now, one quick note on trucks and Xfinity races, the qualifying races for those actual events often take place in the same day. The qualifying race is a pre race basically that NASCAR uses to determine the starting positions for how racers are going to order in the start of the actual race.

Now we need the starting positions to actually run simulations and generate projections for these races. So if you sign on, on Friday morning, To build your lineups for a NASCAR trucks race, you're probably not going to see any projections just yet. That's because qualifying still has to take place. A lot of times, sims and projections are up within an hour of lock of trucks and Xfinity.

So if you register for these contests, make sure you're going to be around, around lock to build your lineups around that time for the cup races, the big races that take place on Sunday. This isn't as much as a pro of a problem because qualifying will take place either the day before or a little bit earlier in the week.

And if you sign on on Sunday morning to build your lineups. The projections there are likely the ones that we're going to have into lock. Now the fourth and final key to crush this season of NASCAR DFS is to understand the tools that you're using here and understand where your projections or your simulations are coming from.

SaberSim had a huge update to our NASCAR model for the 2024 season, improving the individual race sims and race dynamics. Our percentiles are correlations and more understanding how the underlying projections or simulations you're using to build your lineups and NASCAR DFS is absolutely crucial to getting the most out of them and playing the best lineups possible.

The best way to understand how the SaberSim NASCAR sims work is to hear it straight from the person who built them. I interviewed our data scientist, Eric, who is the architect of the SaberSim NASCAR model about how our NASCAR sims really work and what the big improvements are for the 2024 season. You can check out that interview in the description of this video.

And as you sit down to build your lineups for Daytona and the rest of the NASCAR DFS season, remember my four keys. The first, understand the track dynamics and the race dynamics and what upside looks like. The second is to avoid duplication in your lineups. The third is to embrace the variance of the sport.

And the fourth is to understand the tools you're using to maximize their value to your lineup building process. Thanks for watching. Good luck this season. And we'll see you next time.

‍