The NFL starts in a few days, which means it's time for millions of Americans to once again fret about their fantasy football lineups.

Analyzing and using data has become big business just about everywhere you look. How, then, has it been incorporated into the evolving world of fantasy football?

First take: Michael Rand

I see it playing out in a couple of different ways. First, if we talk about daily fantasy games, which involve big money, the most consistent winners are essentially professionals who incorporate complicated algorithms into their lineups to gain edges.

In more traditional fantasy football, which incorporates full seasons and either a draft or auction — often with friends — it becomes a question of who has the most time to invest and who knows where to look for the numbers that matter vs. the ones that don't.

Some people still draft players based on hunches or because they like certain teams, but there is an alarming amount of information at everyone's fingertips when it comes to both drafting and setting lineups.

Chris Hine: I listen to a few different podcasts that focus on season-long fantasy football, and each one of them can find you an advanced statistic that will argue for or against a certain player. It can be information overload and ultimately not very helpful for your preparation.

In the league I've been in the longest (with a bunch of college buddies) one friend of ours, Mike, who doesn't watch much football, has won the league three times in 11 years with a draft strategy that boils down to "If I know their names from Madden, I'll draft them."

So I think you can forget about yards per target, target share, opportunity share and just go with Mike's strategy.

Rand: Really? Because every week I see on Twitter dozens of questions flying around about which player is best to start based on various scoring iterations in different leagues.

I think people actually prefer the information overload — or at least they like the information and the feeling that it gives them an advantage and/or makes them look smart.

(And as someone who puts very little effort into my league and has had generally mediocre results, I am not the "Mike" in question.)

But to your point, not all information and data are created equal. Some of it — maybe even most of it — is best discarded.

Here's what I think is relevant: Positional data that helps you avoid drafting certain positions too early; matchup data for opposing defenses during the season; and of course injury data. Not every "questionable" status on an injury report is created equal.

Hine: I think the format stuff is valid. Taking a pass-catching running back like the Patriots' James White is more important in a point-per-reception league than a standard scoring one. I think you're on to something with how people feel, and perhaps that's how these places make money — and why they try coming up with more stats.

They sell a feeling to people, much like Gwyneth Paltrow and her Goop lifestyle brand. It may not amount to much, but you'll feel better after spending the money.

Rand: Just don't draft Andrew Luck, no matter what his 2018 stats are telling you.

Final word: Hine

Maybe Mike from my league hasn't heard about that one yet.