While we don't question the underlying logic of the Stud Running Back Theory, we're on the record as heretics when it comes to the widespread belief that uninspiring RBs should trump quality players at other positions on draft day. But what comes next? What strategy makes sense for 2007?
Well, here goes: Unless you can get Peyton Manning, Carson Palmer or Tom Brady, don't draft a quarterback in the first seven rounds. And no, we're not smoking crack.
Instead, we suggest you spend the first seven or eight picks of your draft loading up on running backs and wide receivers. Bargain-hunting for elite tight-ends, kickers and defenses is OK after the fifth round, depending on your format, but smart owners generally will focus their attention on building good depth at RB and WR in these rounds.
But why shun the QB position? And why wait until the eighth or ninth round to start drafting them?
Three reasons: 1. Because we think there are few QBs capable of scoring at a competitive level consistently over 16 weekly matchups this year; 2. Because in most drafts, the 12th QB is selected in the 7th round; and 3. Because the differences in production between your league's 12th quarterback and its 20th quarterback are often slim and related more to weekly matchups than issues of talent.
In other words, we're proposing that smart owners can get better overall production out of their QB slot by treating it as a weekly decision. The old days of just plugging in the same slightly-above-average guy every week and hoping for the best simply don't satisfy us anymore.
EVOLUTION OF THE FANTASY QB
When I started playing fantasy in the early 1990s, most of the owners in my league were fixated on quarterbacks, not RBs. Randall Cunningham was the first player chosen in our inaugural draft (1991), and until 21st century rolled in it was common for 80 percent of our owners to keep a quarterback on their three-man off-season protected roster.
So why do I think the quarterback position is declining in its fantasy stature?
Obviously, the near-universal orthodoxy of the Stud Running Back Theory has had something to do with it. So does math. In a 12-team league, it's common to find at least three or four starting NFL quarterbacks on the waiver wire. Good luck finding a starting NFL running back in your free-agent pool.
We think there's a third dynamic this year: With only three sure-fire quarterbacks available, we're much more comfortable with a strategy that spreads around the risk while offering greater week-to-week flexibility. In other words, absent a true stud quarterback, we'll take a risk on our ability to get stud-level production by tracking the weekly matchups for three capable but affordable fantasy backups.
Consider these scenarios:
It's the 4th round. You've selected two RBs and a WR, and you see that Donovan McNabb is still on the board. Since he's the highest-ranked QB (6th overall) on the board and the available WRs are now second tier, you decide to pick him.
When your 5th-round pick comes around, you find yourself deciding between depth at RB, filling your second starting WR slot, or taking a flier on a tight end with Top 3 potential. Rounds six and seven unfold in a similar way, with needs steering your decision-making process.
Having invested a relatively high pick in McNabb, you're not that interested in selecting a backup, and because of McNabb's injury history, you want to save a roster slot for McNabb's backup on the Eagles. Finding a bye-week fill-in for McNabb becomes a relatively low priority, and carrying three QBs doesn't strike you as a particularly attractive option.
And let's say you get lucky: McNabb stays healthy all season (something he hasn't done in years).
But you've got some problems. Like the Eagles' tough matchup against the Bears defense on Oct. 21. Not to mention home games against Dallas and Miami on Nov. 4 and Nov. 18, plus road trips to New England just after Thanksgiving and a Dec. 16 game in Texas Stadium. Oh, and that cold weather matchup against Buffalo in Week 17 looms as a possible down week for the Eagles' passing game. There are reasons to anticipate dips in production in these games, but you'll start McNabb and hope for magic.
What choice do you have? When you invest a 4th round pick in a QB, you've decided to ride the guy as far as he'll take you.
It's late in the seventh round, the 11th quarterback has just been drafted -- and you still don't have one. You take a look at what's available and decide to take your favorite kicker instead, confident that the quarterback picture will look pretty much the same in the 9th round as it does in the 7th.
A couple more QBs get taken before you draft Green Bay QB Brett Favre in Round 9, following up that selection with 49ers signal caller Alex Smith in Round 10. In Round 11, you "reach" for Bears QB Rex Grossman (his average draft position puts him in the 13th round).
In Week 1 you have your choice between Favre at home against the Eagles, Grossman on the road at San Diego, and Smith hosting the Cardinals. You give Smith the start. In Week 2, with both the 49ers and the Packers on the road, you play a hunch and give Grossman the start against the Chiefs at Soldier field. And so on, week-to-week, tracking trends, looking for advantages, exploiting opportunities.
Along the way things will change, players will get injured, various teams will slump and surge. And from time to time (more often than not, it may seem), one of the two players you didn't start will have a great game on your bench. That will sting, but you'll trade that feeling for the sense of being in control of your weekly lineup.
THE MYTH OF THE QB STUD
Think of your starters this way: If you're in a 12-team league and your quarterback is scoring in the Top Five at his position, then that's a statistical "win." If your quarterback is scoring like the No. 6, 7 or 8 QB in your league, that's a "hold." And anything below that is a loss.
That's basic logic, but it's confusing to people who drafted the 12th QB overall and had him score like the 10th overall quarterback. That feels like moderate success, even though they're still losing ground to their competitors most weeks.
Fantasy teams are unlikely to be successful at each position, so in some cases owners don't mind holding steady at one position because the strength of their team is elsewhere. I don't mind breaking even -- or even limiting my losses -- at QB if I've got LaDanian Tomlinson and Chad Johnson wreaking havoc at the No. 1 RB and No. 1 WR slots. And breaking even at QB can be done with less investment than at any other position.
This year there are only three QBs who have that sure-fire stud smell to them (Manning, Palmer, Brady), which means that the next tier on most 2007 draft charts (Drew Brees, Marc Bulger and Donovan McNabb) come with question marks. And even if these quarterbacks actually go on to finish in this order in terms of fantasy points, whoever drafts the 6th one has done nothing more than break even in a 12-team league.
So for comparison's sake, let's stack up our hypothetical QB platoon (Favre, Smith and Grossman, all drafted in the 9th round or later) against the final QB to represent a relative statistical "win": Marc Bulger (2007 ADP = 8th pick, 4th round, in 12-team drafts). We'll use 2006 numbers and the CBSSportsline scoring format for comparison.
STUD vs. PLATOON
Bulger was worth 272 points in 2006, scoring 15 or more points nine times in 16 weeks. That's much better than any of the three guys in our hypothetical platoon.
Yet our three-man platoon, as a group, scored 15 or more points 19 times in 17 weeks, and at least one of the three QBs put up 15 or more points in 12 of 17 weeks. At least one member of the trio equaled or exceeded Bulger in 10 of his 16 starts.
In fact, if you took the best score from the platoon over 17 weeks, the resulting per-game average of almost 19 points would trail only Manning (almost 21 ppg) among fantasy quarterbacks. Carson Palmer's average? An impressive 17.75 ppg -- which still isn't up to our backup trio's best.
Of course the flaw here is obvious: Any old idiot knows to start Manning every week, which means you actually GET every single one of his 332 fantasy points. The odds against extracting all of the 322 potential points from our hypothetical platoon are overwhelming. So what's a realistic comparison?
Well, a conservative approach would be to add up the points you'd score if each week you guessed halfway right -- starting the quarterback who scored the second-highest point total of the three that week. That method yields an average of almost 12 ppg -- right at Grossman's average, just below Favre's, and almost 2 points above Smith's. Assuming this level of performance, you're looking at a quarterback position that is a slight loss in terms of competitive standing, yet it still offers other advantages: Your scoring is good across 17 weeks, not 16; you've been able to invest higher picks in other positions; and you've spread out your risk of injury and other QB maladies considerably.
But here's the thing to consider: While a smart, experienced fantasy owner won't be right all the time, one can reasonably expect that he or she is likely to be right slightly more often than not. So let's see what happens if an owner gets some percentage of the available points between the median (what you'd get by starting the 2nd best of the three each week) and the optimal (always starting the highest-scoring QB each week).
Capturing 30 percent of those optimal points produces a 14 ppg 17-week average. Boost that to 40 percent accuracy and the 17-week points-per-game average jumps to a highly respectable 14.7 points.
Is that reasonable? We think so. Janet and I tried variations on this theory last year with some success, in large part because we used the weekly report offered by Fantasy Guru, which breaks down the offensive performances allowed by NFL defenses, to target attractive matchups. The irony is, if you're heavily invested in stud players, you've got unprecedented high-resolution statistical information from professional fantasy services, but no decisions to make. When you're managing non-stud players, you're psychologically more likely to act on the insights you gain from that level of analysis.
In other words, the explosion of fantasy football information is really what makes this strategy possible. It's rational in 2007. Ten years ago a weekly QB strategy would have been nothing more than educated guessing.
WHY WE THINK IT WINS
Obviously, if I can get almost 15 points per game and a competitive "hold" at the QB position without spending a pick on a signal caller until the 9th round, then I've managed to create a solid advantage for my team.
It means my first eight picks can be spent on positions where platooning is a less-effective option. It means I'm more likely to have quality depth across my roster, more likely to acquire advantages at undervalued positions like kicker and defense/special teams. It means selecting my No. 2 WR at a point in the draft when many owners are still looking for their No. 1 WR.
Not only that, but I'm in a better position to adapt to a mid-season injury, and I'm less likely to just passively absorb the impact of a stud QB starter facing an obviously disadvantageous matchup.
15 QBs AVAILABLE IN 'THE PLATOON ROUNDS'*
13. Jay Cutler, Denver (8.01)
14. Eli Manning, New York Giants (8.10)
15. Ben Roethlisberger, Pittsburgh (9.08)
15. Brett Favre, Green Bay (9.08)
17. Jake Delhomme, Carolina (10.02)
18. Alex Smith, San Francisco (10.09)
19. Matt Schaub, Houston, (11.10)
20. Chad Pennington, Jets (12.01)
21. J.P. Losman, Buffalo (12.08)
22. Jeff Garcia, Tampa, (12.09)
23. Steve McNair, Baltimore (13.01)
24. Rex Grossman (13.02)
25. Byron Leftwich, Jacksonville (13.07)
26. Trent Green, Miami (13.09)
27. Jason Campbell, Washington (13.11)
*Average Draft Position information provided by Ant Sports. Specific data is for a 12-team league, High-Performance scoring system, normal lineup with TE-required, based only on Serious Mock Draft results from July 9th to July 30th, 2007. The first number refers to the player's ADP ranking at the position (i.e., Cutler is, on average, the 13th quarterback selected). The number in parentheses has the draft-round number before the decimal, and the selection within that round after the decimal. The author has removed Michael Vick from the ADP results.