FAKE RONJAN!!Dear fantasy football: you're doing it wrong!
The NFL recently ranked its top 100 players, league-wide, and 8 of the top 25 players were quarterbacks. The NFL also ranked their top 200 players for fantasy football. As of this post date, there were no quarterbacks in the top 25.
ESPN recently featured a mock draft by 12 fantasy football professionals, and we noted on Facebook how crazy it was that Tom Brady (3rd QB) was selected one pick after his teammate Shane Vereen (23rd RB).
What the hell is going on here?! Is it fun having marginal running backs be more valuable than the best players in the game?! Really?!
Fantasy football is awesome. I've been playing it since 1998, and it makes Sundays so much fun, even as my fanaticism for the NFL in general has sharply fallen off in recent years. And sure, it's not perfect. Touchdowns are probably overvalued, the importance of first downs is largely overlooked, many of the newer analytics around win probability aren't incorporated, and so on ... but ultimately, the best players tend to score the most, and that's what makes it work.
This is why the consistent undervaluing of quarterbacks is so appalling -- particularly because it's completely artificial.
Look -- it's one thing with defensive players. It's hard to take standard player statistics and make that meaningful for defenders, since they can heavily impact the game without ever touching the ball or creating a statistic. I get it. Fantasy football scoring is probably better when defenses are treated as a team instead of breaking out players by themselves.
But quarterbacks? Come on. They have the ball on every play. Their impact is measurable. Yards, touchdowns, interceptions, scrambling production ... it's all there. The top scoring fantasy quarterbacks last year were Drew Brees, Tom Brady, and Aaron Rodgers ... it just sounds right.
The problem is that we set up leagues that start only 1 quarterback, while we start at least 2 running backs and 2 wide receivers, thereby creating an artificial abundance of quarterback talent in our fantasy leagues that doesn't exist in the real NFL.
"Value Based Drafting" (VBD) is a drafting principle that has been around for a while, originally put forth by the fine folks at footballguys.com. I was once an avid reader of that site, even participating in (and "winning") a league with some of its key contributors and forum members. It's great stuff, and a lot of their learning and analytics have cascaded into mainstream fantasy football sites.
The basic principle of VBD, in footballguys.com's own words: "The value of a player is determined not by the number of points he scores, but by how much he outscores his peers at his particular position."
The traditional way to account for this is to take whatever you expect a player to score for a given season and subtract out a baseline figure for that position. Often, the baseline is the lowest ranked starter that will be in your fantasy league. So in a 10 team fantasy league, the baseline figure would be the 10th ranked quarterback, the 20th ranked running back, and the 20th ranked wide receiver.
Above is ESPN's standard projections for the 2013 season. The blue line shows the points projection for the #1 scoring QB, the #2 scoring QB, and so on down to #20. The green line is RBs, the yellow line is WRs, and the red line is TEs. So we end up with an illustration of where production falls off for each position.
As you can see, quarterbacks scale down little by little until they completely fall off at around the 13th ranked QB (who happens to be Eli Manning). Notably -- this drop from #12 to #13 is larger than the drop from #13 to #24. This makes sense -- today's NFL has 4 or so elite QBs, with several solid-to-great QBs that follow, and then a lot of shaky or inconsistent ones (at least from a statistical standpoint).
However, if the baseline level for quarterbacks is the 10th ranked quarterback (Tony Romo), then every projected starting quarterback is better than where the real dropoff in production takes place.
Meanwhile, the RB, WR, and TE positions see a couple of truly elite players followed by a long tail of players with slightly diminishing production. Once you're past the top two or three players, you're already out of the elite / great area.
Net -- quarterbacks are punished because there are too many good ones in today's NFL. That's crazy!
And to be precise -- even if there were only 3 or so "elite level" quarterbacks before a huge drop-off, QBs would still be at a disadvantage because we are not going as deep to calculate their baseline. But in today's pass-heavy league, the disparity is almost unprecedented because there are so many good passers!
The solution here is simple, and it's one that I've used in every league I've been in since 2004. Start two quarterbacks! In doing this, you're measuring the top quarterbacks against lower baseline players, thereby making them more valuable. Not to mention, the option of starting both Aaron Rodgers and Drew Brees creates legitimate and fun new strategies.
The counter-argument is always the same. "How are there enough quarterbacks in the real NFL, especially when bye weeks start coming around?"
This was once a legitimate concern, but now it's largely been eliminated with the emergence of true flex positions (called "OP" in ESPN leagues) that allow you to start a 2nd QB or a flex player at any other offensive position. Since quarterbacks tend to score more in general, it highly incentivizes a two-QB strategy, but it leaves open enough flexibility in case of injury or bye weeks (or indifference to the strategy). Problem solved!
I'd recommend playing around with different settings to see how they impact the stats. Footballguys.com (among others) releases an Excel program every year that allows you to input your league settings and have it customize your rankings "cheat sheets" according to VBD principles and calculations. If you run their standard scoring and settings (using the free tool with June projections), Aaron Rodgers is the highest-ranked quarterback at spot #20. If you change the settings to start 2 quarterbacks, Rodgers shoots up to #2 overall. Which one sounds better to you?
It's really that simple. Start two quarterbacks. Or you're doing it wrong.
-rj