For the past two weeks, laps led have been an emphasis in NASCAR DFS. With 267 laps, there was plenty of upside to justify using drivers at the front of the pack, and we saw that play out in perfect lineups for the races.

This week, laps led shift from an emphasis to a priority.

Sunday's Instacart 500 in Phoenix is 312 laps long, the lengthiest race to date in the 2021 season. It leaves us with 31.2 points available for laps led on FanDuel, an increase of 4.5 from last week. That may not seem like much, but history tells us it should shift our mindset.

Let's dive into that data, see how far we need to go to find these lap-leaders, and apply that to roster construction for Sunday's race.

A Front-Heavy Approach

Heading into each of the past two races, the only "must" of a FanDuel lineup was at least one driver who was starting near the front and had the ability to lead laps. One driver from the top 10 made the perfect lineup at Homestead, and two did so in Las Vegas.

Here, the requirement shifts to at least two, and you have the wiggle room to jam in a third at times, as well.

The evidence for this comes from last year's races at Phoenix and tracks like it. In the spring Phoenix race, three drivers who started inside the top 15 made the perfect lineup, and all three led at least 60 laps.

Both drivers who started on the front row making the perfect lineup isn't all that common, but it does illustrate the upsides in targeting lap-leaders.

The second Phoenix race is more unique. It's the championship race, meaning the four championship contenders tend to get extra leeway and dominate the race. So, we should take data from it with a grain of salt. But once again, both drivers who started on the front row made the perfect lineup.

In the two Phoenix races, the perfect lineups contained an average of 251 laps led across all five drivers, and both had at least 220. Unless one driver absolutely torches the field, this tells us we need multiple lap-leaders in each lineup.

One thing worth noting with the first Phoenix race is that there was qualifying for the race. That means -- in theory -- the fastest cars started at the front. That may not be the case this weekend with the starting grid being set by the same formula used to set the starting grid the past three races. That presents the possibility we could be more relaxed with our front-centric philosophy if the fastest cars aren't necessarily at the front.

There is some validity to that mindset, but it's not something that should alter our approach. Pushing the championship race aside, we can get other signals for our approach from looking at the playoff race in Richmond.

For that race, the starting order was set by the same algorithm in place for this weekend. Although it looks a bit different from a roster construction perspective, the perfect lineup for that race still involved two drivers in the top 10 who led a healthy number of laps.

Martin Truex Jr. finished second, helping explain his inclusion despite not leading laps. Finishing and place-differential points still matter here even if we're putting more emphasis on laps led.

The final race we can turn to for signals of what to expect on Sunday is the summer race in New Hampshire, the closest comp we'll get to Phoenix. That was a 301-lap race where the starting order was set by a similar-but-different method to this weekend. In that one, once again, two lap-leaders wound up in the perfect lineup.

These four races are the ones that best inform how we should play this weekend, and they all say the same thing: we need at least two lap-leaders in each lineup, and those guys are decently likely to start right at the front of the pack.

If you love the speed of Joey Logano, who is starting in ninth, you have the leeway to use him. The Richmond playoff race provided evidence of that. But you also have full agency to roster both Brad Keselowski and Kyle Larson on the front row if you think they're the two best cars in the race. There's more than enough laps-led upside to pass on place-differential for the studs in your lineup.

Two other things should be apparent based on the perfect lineups above: we should look for mid-range drivers who can run out front, and we should be willing to roster drivers in the $6,000 range.

It's very possible things play out in a similar fashion to the way they did in last spring's race. It wasn't the championship race, so a larger pool of drivers was pushing to get the win. And in the new rules package, passing was much easier than it had been in 2019. We could very well see three drivers soak up points via laps led. So, if you find a couple of drivers with salaries between $9,000 and $11,500 who you think have the speed to be in contention throughout the day, there's a lot of value in plugging them in.

In order to jam in more of the mid-range plays, we should be willing to roster true values. These aren't drivers who will up the win expectancy of our lineup themselves, but they allow us to squeeze more win expectancy and laps-led upside into our other four slots. You don't want to get too crazy down there because you do still need finishing points, but there is increased viability of legit values here versus what we've had the past two weeks.

As you can see in the charts, most of those value plays should start deeper in the pack. The average starting slot for drivers with salaries below $8,000 is 22.3, and the average for those under $9,000 is 21.4. Laps led are a priority only for the studs and a secondary focus for the mid-range plays.

We will have some place-differential juice available to us, too. Four of the top 15 drivers in my model are starting 20th or lower, and all of them are likely to have mid-range or value salaries. The starting slot shows that they've had rough starts to the season, but with Phoenix being the first oval race with the 750-horsepower package, it's fair to expect a rebound from those who have lagged if they were good in this package last year.

All told, our focus for this week is pretty clear. We need to prioritize lap-leaders, shoving at least two into each lineup. If we can snag a third in the mid-range, that's a viable option, as well. We can feel free to spend down if we need help jamming in the lap-leaders, and in doing so, we should look toward drivers in position to scoop place-differential points. This type of construction showed up in each of the perfect lineups above, and we have no reason to expect anything different this weekend.