My Twitter feed is full of people arguing that the Giants taking Saquon Barkley over one of the available QBs would be malpractice because quarterbacks impact their teams so much more than running backs - or players at any other position. I've argued on the podcast that while the second clause of that sentence is partially true - more on that below - the first part might well be false.

Let's concede from the outset that quarterback *play* is far and away more important than say running back play or wide receiver play, but how much of your quarterback play is due to the player himself and not the circumstances in which he plays? Put differently, we all thought the Eagles were done last year when Carson Wentz tore his ACL, but Nick Foles had no problem lighting up the best defense in the league in the NFC title game and hung 41 on the Patriots with flawless play in the Super Bowl. Is Foles, who played like peak Joe Montana, really that good, or was it just great to play in Doug Pederson's system? And what does that say about Carson Wentz?

Let's grant there are some transcendent QBs that can overcome a mediocre system - Aaron Rodgers and Russell Wilson come to mind. But if Case Keenum can be a borderline star one year after looking awful his entire career, if Jared Goff can go from almost-certain bust to rising star under Sean McVay, how much can we really attribute to the quarterback himself rather than the system? And it's not just players freed from the clutches of Jeff Fisher, either. Andy Dalton played at a near MVP level in 2015 and has been just average since. Matt Ryan under Kyle Shanahan had one of the best QB seasons of all time, a huge departure from his usual, solid but unspectacular output. Yes, quarterback play is the most important factor, but if the quarterback is typically only half of that equation or less, how important is it to land a particular one in the first round rather than a coachable one later or through free agency?

Next, even if we were to concede it was the quarterback himself largely responsible for the team's quarterback play and not the system, that would only be worth ponying up for in the draft if we could predict with some reliability which QB would actually be good. Because so much of a quarterback's success is mental and psychological, i.e., decision-making, judgment and poise, rather than size, speed, strength and athleticism, the most relevant skills are hard to measure. That's not to say we have no idea who might be good, only that it's very hard to tell who will merely be adequate and who will be great (see the table below.) Remember, adequate is Keenum, Foles, Josh McCown, Tyrod Taylor and other QBs a new crop of which are almost always freely available every season. Great are the ones who largely transcend system or function at an all-world level within a good system.

When we say QBs are the most valuable players we tend to mean, most valuable over replacement level, and that's undoubtedly true if we count the dregs of the league - DeShone Kizer, Trevor Siemian, Tom Savage - as constituting replacement. But if we move replacement to adequate - the Keenum/Foles level - then it's unclear why having Marcus Mariota, or Jameis Winston or even Goff is so enormously valuable. At least 20 teams have quarterbacks that are adequate or who project as such in the future like Pat Mahomes and Mitch Trubisky, so it can't be that hard to find a player to consider your franchise QB.

There are two other problems with drafting a rookie QB early. First, if you miss, you're probably sending your franchise into the wilderness for at least two and maybe up to four years because it's hard to give up on such a large draft-day investment. If your high-end defensive tackle busts, it won't have nearly the impact because it's not going to delay your search for other good players at the position, and it's not going to drag down an entire unit of your team. The QB pick comes with a major free-agent and later-round-QB-pick opportunity cost. If the Seahawks took Ryan Tannehill, for example, they probably wouldn't have drafted Russell Wilson (though oddly the Redskins did draft RGIII and Kirk Cousins that year, but that is unusual.)

Second, the NFL operates under a salary cap and once your QB, who might take a couple years to hit his prime, graduates from his rookie contract, he's going to clog up more of your cap than the star receiver you passed on in his class. So even if it largely works out, you're not getting a major bargain if you're putting an unduly large portion of your budget toward Matthew Stafford or Derek Carr. Sure, you pretty much can't overpay Wilson or Rodgers, but essentially, unless you hit the pick out of the park, it's unlikely to be a major advantage for the franchise long term. And keep in mind Rodgers, Wilson, Brady and Brees were all taken outside of the top-20 picks.

I ran out of time to get the salary info, but I went back and tracked the first-round picks for the last 20 years to see how they panned out as a group. Here are the results:

You can see from the table the vast majority of value from first-round rookies is vested in a handful of players: Peyton Manning, Ben Roethlisberger, Aaron Rodgers and Drew Brees, and since 2000, no player in the top-10 has been that kind of difference maker. That doesn't mean the Giants should necessarily pass on one of this year's crop at No. 2, but it cuts against the notion that taking one there is a no-brainer.