This Sunday's race in Atlanta is what happens if you put the past three races in a blender.

It has elements of Homestead and Las Vegas because it is a 1.5-mile track. This allows passing to exist, meaning drivers can work their way forward from poor starting positions to finish well. It also overlaps with Homestead in that there is heavy tire falloff within a run.

Last week is more about the race composition than the track itself. After running 312 laps in Phoenix, there are 325 scheduled for Sunday's Folds of Honor QuikTrip 500. That now makes this the longest race thus far in 2021, a key driver of our decision-making in DFS.

So, we have a 1.5-mile track with a bunch of laps to be led. What kind of strategy does that lead to?

Let's look at past races of a similar composition and see what we can cook up for our NASCAR DFS lineups.

Prioritizing Lap-Leaders Again

If you're a regular DFS player, you know what the race length this weekend means: we need to once again sell our souls to get lap-leaders into our lineups.

The 325 laps means there are 32.5 FanDuel points available for laps led, which is a big number. If we miss that boat, we won't have enough juice in our lineups to compete for tournaments.

Oftentimes, it'll wind up being one driver who toasts the field and becomes a can't-miss option. Last year in Atlanta, Kevin Harvick led 151 laps. In the first Texas race -- another at a 1.5-mile track with more than 300 laps -- Ryan Blaney led 150. If one guy dominates, you can squeeze by having just two lap-leaders in a single lineup.

But there is plenty of leeway to jam in a third. And even if that guy doesn't lead laps, it can still pay off.

There were three total races at 1.5-mile tracks last year with either 325 or 334 scheduled laps. All three starting orders were set without qualifying, which will also be the case this weekend.

Each of the perfect lineups from those races included three drivers with a salary of $10,000 or higher. For each one, the third driver had a salary in the $10,000 range.

This makes abundant sense. If the race is longer, the appeal in using drivers starting closer to the front goes up. Those drivers will typically carry bigger salaries. If they can scoop a top-five finish and lead some laps in the process, the odds they're in the perfect lineup are very high.

This doesn't mean the perfect lineups went full stars-and-scrubsy. Across the three perfect lineups, only one driver with a salary lower than $7,000 made it, and that one driver drove for a good team with Stewart-Haas Racing. The route to a three-stud lineup was not via punting but rather via balance in the studs.

Only one driver in the three perfects came with a salary higher than $12,200. That was Harvick when he dominated in Atlanta. The others peppered the $11,000 and $12,000 range, leaning on the lower end of the elite tier to get good finishes and laps out front.

This absolutely does not mean you should ignore the high-salaried drivers. As mentioned, we often wind up having a can't-miss stud in these races, and the odds that can't-miss stud comes with a large salary are high. It's more to say that we should give an extended look to drivers on the lower end of the stud range, ask ourselves if they have the upside to lead laps, and be more willing than usual to use them.

Because the race is so long, you could -- in theory -- get your lap-leaders from deeper in the pack. There's plenty of time for them to work their way forward and be a contender the second half of the race. That's just not as likely this week.

Of the top 11 drivers in my model, only one will start outside the top 11 (Kyle Busch in 19th, who ranks 9th in the model). The most realistic lap-leaders for this race are starting at the front. So for both our studs and mid-range plays, we should be willing to use them regardless of where they're starting thanks to the increased upside for those at the front.

We can afford to drift further back with our value plays, though it's far from a must. In last year's Atlanta race, Cole Custer made the perfect lineup after starting 31st. Three drivers made the perfect lineup in the first Texas race while starting outside the top 20, a group that included the top two finishers in the race, Austin Dillon and Tyler Reddick. As long as the car is fast enough, place-differential is fully on the table.

There are some intriguing drivers starting further back, especially Austin Cindric in 39th in a Penske car. But we should be receptive to mid-range and value plays with higher starting spots, too. The second Texas race last year -- the best analog to this one from a starting order perspective -- had no drivers in the perfect lineup who started outside the top 15. We need good finishes here, and if we can't get them from those starting further back, we need to move closer to the front. This will require a close look at the speed of each driver, something we can get a good idea of by looking back at the Homestead race due to the similarities in tire wear between the two tracks.

In all, this is an event where our roster construction will be different from usual. The uniqueness of the event requires that. But it will position us to exploit all the upside available to us.

Primarily, we'll need at least two lap-leaders per lineup, likely starting near the front. In a healthy number of lineups, we should look to pump in a third with a mid-range salary who can at least get a good finish if they don't lead laps.

The place-differential will come from the value plays, if we wind up targeting it at all. We'll just have to dig deep into those drivers' performances on similar tracks and decide if they can push to get us a top-10. If not, we'll have to be comfortable with a front-heavy lineup that forgoes place-differential entirely.

It's a far riskier strategy than we usually deploy without the low starting spots to boost our floor. Failing to make these tweaks, though, will keep us short of the requisite points to cash in a tournament, much less take one down. So at least for this one week, it's wise to embrace that risk and hope we hit on the right drivers.