It's a question we tackle every NFL offseason -- what should we do with Bill Belichick's backfield in fantasy football?

This year is no different.

In 2017, the New England Patriots' backfield work was divvied up into roles that were easier to predict than the narrative would have you believe. When everyone was healthy, Dion Lewis was the lead guy, and Rex Burkhead got meaningful work as a do-it-all number-two back.

But Lewis is in Tennessee now, and the Pats took running back Sony Michel in the first round of the NFL Draft. There's uncertainty again with this backfield, and it appears to have created a situation where we can find some fantasy football value.

Let's try to dissect what this backfield could look like for 2018.

Belichick's History

The fabled Patriots running back curse -- does it even exist? To find out, let's take a quick look at the historical rushing attempts for the backs who have played under New England's Hall-of-Fame head coach.

I know, that's a lot of colors. So let's break it down.

The Patriots have produced a 200-carry back in 7 out of 18 seasons under Belichick (plus a couple years in which a player on that pace was derailed by injury). The carry-count leader in New England has averaged a respectable 210 carries over that span, with a high of 345 in 2004 and a low of 96 in a flukey and injury-riddled 2014 season.

While 200 carries isn't a number typically indicative of an elite fantasy producer, 8 of the top-24 backs in standard formats last season -- including Dion Lewis -- had less than 200 rushing attempts.

As for the committee curse, there have surprisingly been only five seasons under Belichick in which the top back in the offense and number-two runner have been separated by 15% or less of the team's total carries. This 15% benchmark cropped up in 2000 (10.6%), 2003 (1.0%), 2006 (5.6%), 2013 (5.8%), and 2014 (0.5%).

It's worth noting that while 2000, 2003, and 2006 were more or less true timeshares, the more recent occurrences were not. The 2013 numbers were a result of Stevan Ridley's fumbling issues and the consequent split in production between he and LeGarrette Blount. The 2014 market shares were marred by Ridley's season-ending injury after just six games as he was on pace for 256 carries that year.

So, what about the aforementioned fumbling concern? The 2013 mess was probably the worst -- or best, I guess -- example of the Belichick fumble hatred as Ridley lost a huge chunk of his prior year's 290-carry role as a direct result of his ball-security struggles. We also saw a fumble result in Mike Gillislee being summarily benched in Week 6 last year -- and then essentially scratched from the lineup for the second half of the year -- but he had also been largely ineffective since Week 2.

On the flip side, we saw little to no ill will toward Rex Burkhead after his Week 11 fumble last year. The versatile back totaled 40 touches for 209 yards and 5 touchdowns over the next three weeks.

All in all, the data is a bit of a mixed bag. Under Belichick, it's been, essentially, an even split between seasons of fairly typical volume for a lead back and true committees. Which will we get in 2018?

Depth Chart Dive

Following the loss of last year's lead back, Lewis, the Patriots were quick to add veteran Jeremy Hill in free agency and rookie Sony Michel in the draft. That makes for a crowded running back room (as of July) of Michel, Burkhead, Hill, Gillislee, James White, Brandon Bolden, and undrafted free agent Ralph Webb. Bolden has always been more of a special-teams guy, and Webb will have to do a whole lot to make the team, let alone impact 2018 fantasy drafts, so we'll focus on the other five guys.

One of Hill or Gillislee will likely be cut before the season. They essentially profile in the same role of short-yardage pounder, so it wouldn't make much sense to keep both around. Regardless, neither should be a threat to top the depth chart, and both feel like depth players, at best, although Gillislee did shine in a limited role with the Bills in 2016.

Meanwhile, White is a Patriot pass-game staple, like Shane Vereen and Danny Woodhead before him. He has topped 40 receptions and 400 receiving yards in four straight seasons, and he has been a stud throughout the NFL playoffs. His role should remain steady in 2018. He's not a fantasy option except in deep PPR leagues, and he shouldn't impact the carry count of the lead backs much at all see that he's averaged 34.6 carries per game the last three years.

The rubber really meets the road in the faceoff between Michel and Burkhead, and the outcome of this depth-chart battle should have a serious fantasy impact.

The Patriots' Jackknife

Burkhead was efficient in limited action last year. In only 10 games (chest, knee injuries), he shouldered 64 carries for 264 yards and 5 touchdowns, adding 30 receptions for 254 yards and 3 more scores. And he also contributed on special teams.

In Week 10 against the Broncos, Burkhead had his signature do-it-all game. He totaled 13 touches for 63 yards, snagged a receiving touchdown, and blocked a punt.

Oddly, Burkhead's rare versatility may actually serve as a ceiling to his role as a fantasy running back. After all, when is the last time you saw a star fantasy running back who was also asked to regularly contribute in punt coverage? Burkhead is the quintessential Patriot player and a jackknife of a weapon -- which means the Pats will likely have him running routes out of the slot and chasing down returners on special teams.

The Gifted Rookie

As for Michel, the Patriots took him with the 31st pick in the draft, only the second time they've landed a running back in the first round in the Belichick era. Obviously, Michel is talented if he went that early in the draft, but the one knock on him -- and this could be a killer in New England -- is an infamously poor fumble rate in college.

While it is possible Michel fumbles early in 2018 and quickly cedes the job to Burkhead, we may need to pump the brakes a bit. Michel improved his fumble rate every year in college -- from 3.1% as a freshman to 1.3% as a senior. And it stands to reason that if Belichick thought Michel had a serious fumbling issue, the Pats probably would've opted to take a different running back at some point in what was a draft loaded with quality backs.

Additionally -- and, crucially, from a fantasy standpoint -- Michel, on paper, outclasses Dion Lewis as a goal-line back, measuring in at 3 inches taller and 20 pounds heavier. Burkhead enjoyed hefty red-zone usage last year, splitting work with the undersized Lewis, so we could see that balance out with Michel in town.

Clearly, Belichick wanted Michel -- and paid top dollar for him -- for a reason. Unlike Burkhead, the reason the Pats value Michel so highly figures to be entirely founded in his value as a running back. He should be the better pure runner and higher-upside talent. Burkhead, for sure, should see a healthy share of touches, as we'll discuss below. But, barring a collapse by Michel, the high-pedigree rookie figures to get more than a fair crack at the lead role.

Projecting Production

First off, we have to consider the value of a running back carry in the New England offense. Since a relatively down year in 2004, the Patriots have finished in the top 12 in rushing yards in 9 of the last 14 seasons, and they've been in the top 12 in rushing touchdowns every single year in that span -- including eight finishes in the top five.

Based on the historical data we've combed through, the players we've profiled, and the fact that it was just a year ago, the best possible blueprint for this backfield seems to be 2017. The key figures were Lewis, Burkhead, and White. Throwing out any games Burkhead missed due to injury, as well as the first two weeks, when the Gillislee Experiment was still online, leaves us with an eight-game sample of this trio (Week 7 through Week 14).

From those eight games, here are the stats and half-PPR fantasy points for each guy.

Just about everything in this table works fairly well when trying to project forward to 2018, except the touchdowns. As we mentioned above, Michel could bring more to the ground game in the red zone than Lewis did. Additionally, White happened to score all his 2017 receiving touchdowns in this stretch, and he has averaged four receiving touchdowns per season over the last three years, so his total above may be a bit of an outlier.

Of course, it may not be this simple, but if the Pats' backfield carries this usage over to 2018 with Michel subbing in for Lewis, then it'll make Michel a guy we'll want to get our hands on.

In an effort to see what the year-long numbers could look like this upcoming campaign, let's plug Michel into the 2017 template above and calculate the 16-game totals (so multiplying everything in the above table by two). Touchdowns can be fickle, so even though the scoring numbers from our eight-game sample in 2017 may be a little misleading for Michel, we'll leave them as they are.

Again, it's probably not going to be as easy as removing a Lewis-sized puzzle piece, plugging in a Michel-shaped one, and expecting everything to stay the same. But this will at least give us a rough idea of what this backfield could look like if Michel gets Lewis' lead-back role.

Those are some pretty excellent numbers for the top two backs. Comparing the per-game averages to those of all fantasy backs last season, Michel would rank as the RB18 in half-PPR leagues, and Burkhead would shoot his way up to the RB14. Again, the finishes are highly-dependent on that touchdown split, so if Michel does outshine Lewis in that capacity, he and Burkhead could swap. But clearly both have impressive upside in this projection.

And if you think that seems unrealistic, note that three teams -- Saints and Falcons, and the Pats themselves -- produced a pair of top-24 backs on a per-game basis in 2017.

Plus, the Patriots have one of the easiest schedules in the league, per our metrics, and Vegas has their win total set at 11, highest in the NFL. It certainly helps to draw six matchups with AFC East foes, all of whom ranked in the bottom 10 in fantasy points allowed to opposing running backs last year.

And then there's the fact that Tom Brady, while still elite, will be 41 years old this year. We shouldn't expect his game to regress on a per-attempt basis. But you do have to wonder if the selection of Michel in Round 1 -- rather than a new quarterback heir -- might spell the beginning of a shift in the Patriot offense, much like we saw in New Orleans last year.

Conclusion

There's some potentially awesome value here, especially if you think Michel is going to get a role similar to Lewis' in 2017.

Michel is currently being drafted as the RB24 in 12-team standard leagues, per Fantasy Football Calculator. Burkhead is falling all the way to RB34. Both players have a chance to outdo those average draft positions by a longshot.

Don't fear the Bill Belichick backfield.

Embrace it.