Photo: Imagn Images

I'm taking a hiatus from dynasty rankings to reassess my philosophies about managing teams in these leagues. This week, I'm examining quarterback data through the lens of something I call career windows. I define career windows as a three-year period in a player's career because that span of time is known as the average career length of an NFL player.

When we draft players we believe will be viable fantasy contributors, I think we have the expectation that their careers will have more than three years of starter production. Many players have careers three to five times longer than the average.

I think we assess a player's talent and situation every year, but I like the idea of have some longer perspective about players at each position. I want to know the lay of the land:

• If and how long I can expect to get starter productivity from players at each position.
• How does a player's draft status fit into the scope of these career windows?
• Are there any basic ideas I can draw from quarterback careers to help me create a basic philosophy for rankings and ranking adjustments in dynasty leagues?

What I hope to gain from these exercises is a set of basic ideas that help me understand when I'm going with or against the grain and to make those decisions with awareness of the dynamics at play.

Tiers Defined

First, I need to establish how I define the categories I use for the position when talking about its fantasy production. Most of you are familiar with how these terms are defined, but I still need to cover my bases. I'm using a 12-team league as the standard for these tiers.

• Elite - In my view, these are the three most productive productive quarterbacks in a given season.
• No.1 QB (or QB1) - The 12 most productive quarterbacks in a given season.
• No. 2 QB (or QB2) - The 13th through 24th quarterbacks. First-tier reserves with potential for playing time on your roster.
• Bench - The 25th through 36th quarterbacks. Depth for your roster - some developmental; others emergency only.
• FA (Free Agents) - Quarterbacks lacking talent, situation, and/or opportunity to prouduce in the present.

These tiers are general estimates. There are some years where four quarterbacks post elite-level fantasy production and others where only one makes the cut. I'm not using fantasy points to drive tiers because the point values have changed - especially at the quarterback position - in just 20 years. Chase Stuart posted a graph of this change, but using VBD as a more refined data point.

Here's what the 20-year change looks like through the lens of quarterback tiers:

Fantasy Points By Quarterback Tier 1992-2012

You can see similar changes in Chase Stuart's VBD Trends piece or Heath Cumming's Quarterback Career Arcs piece - both recommended reads. Quaterback production has climbed significantly during the past 20 years. At the same time, the difference in fantasy points between each tier of quarterback play has remained stable.

Average Difference in Fantasy Points From Tier to Tier

The elite quarterback tier is on average six points better per week than the QB1 and the QB1 is five points better than the QB2, and so on. This chart underscores the importance of having at least a QB1 on your dynasty roster every year. It's also a reflection of why the more competitive dynasty leagues I've seen tend to hoard quarterbacks and demand what might seem like an inflated price for QB2s.

There's an inherent understanding from watching the game that if a fantasy owner lands a quarterback capable of elite production, they have a good shot to have no worse than a QB1 for 5-7 years, which is roughly twice the length of the average career window. This list of quarterbacks from the past 20 years with NFL careers spanning at least a decade in length tells the story.

Quarterbacks with at least 10 Years in the NFL

The first four names are the poster children for this point. If you drafted Manning, Brees, Brady, or Favre in a dynasty league and didn't trade them, you've had between 9-14 years of QB1 production. That's 3-5 career windows - or dynasty team lifespans - that you didn't need to address quarterback play. It's the kind of dynamic that allows you to allocate your draft picks, trade focus, and waiver wire options to other positions.

It also saves you roster spots and lessens the likelihood of trading away players you liked in order to acquire a position you would have otherwise needed to address. Whether these four quarterbacks (and Dan Marino) are an anomaly or the beginning of a trend for elite production remains to be seen, but the table demonstrates that you don't have have an elite passer to possess a valuable anchor for your squad.

Drew Bledsoe may have never been an elite fantasy performer, but eight years of QB1 production made him a solid anchor to build a team around during the front half of his 14-year career. Mike Vick's career has been filled with injury and turmoil, but he has totaled two career windows as a viable fantasy starter.

There are eight players on this list with careers of at least a decades in length and at least six years of QB1 prouduction - five of them with the bulk of their career taking place after 2000. Half of this list is comprised of quarterbacks with more than one career window (three years) as a fantasy starter. Of the 14 players with only one career window or less as a QB1 had at more than one career window as a QB2. I think the information gives some indication that if a player has three years of fantasy QB1 production under his belt, he's healthy, and his team is confident in him as the starter, then he has a good chance to produce at least another career window of QB1 value and worth an investment.

Here's a list of quarterbacks with 5-9 years of experience and how their production in terms of fantasy tiers.

Quarterbacks with 5-9 Years in the NFL

The dynasty implications of this list is the most fascinating of the two. Some people think Joe Flacco has turned the corner as a quarterback. He played at a high level in the playoffs last year, but I'm not convinced it will translate to elite fantasy production. On the other hand Jay Cutler has been a QB1 twice in his career and in an offense that make him with the gamble to acquire. Even if he falters and the Bears let him loose next year, I like the risk-reward because he's finally paired with a coach that has a great offensive track record and the offensive line should be better. If it comes together for him in 2013, it could begin a run of QB1 production that puts him into that 10+ Year category.

Carson Palmer is by some accounts old and worn out. I'm not buying it. I see the Bears and Cardinals dynamics simiarly and it could give Palmer another three-year window of starter production. Considering how close he was last year with an offense that was horrible, it's another nice investment risk compared to spending a high-round pick for a quarterback this year or next in rookie drafts.

However, do I rank Flacco, Cutler, and Palmer high enough to value them as possible QB1s? I think the answer ultimately rests with team dynamics and surrounding talent, but there will be a significant historical element that warns me not to do it. The reason I'm beginning to see is that the risk of trying to acquire a quarterback before he hits might not be worth the reward. At the same time, the risk of dumping a quality fantasy quarterback before the bottom falls out might be an equally hasty approach.

For all the fears that Jerry Jones is wrecking his team by meddling too much in the daily affairs of his coaching staff, I seriously doubt Bill Callahan will damage the Cowboys offense so severely that Tony Romo doesn't meet QB1 expectations for another year. I'd pay a premium for another three years of Romo - especially with Dez Bryant turning the corner in his personal life.

Bill Barnwell wrote a Grantland piece about the Gang of Four and rookie contracts are revolutionizing the landscape of quarterbacking. I see the logic, but Barnwell's inclusion of Matt Ryan as a player he foresees as part of the impending middle class upheaval in a couple of years might be a stretch. If Sean Renfree plays to the potential I saw at Duke it could happen, but don't count me among those waiting for the Falcons to dump Ryan in 2-3 years. That said, I have compared Ryan to Drew Bledsoe and I think most would have considered what played out between Bledsoe and Brady impossible until it happened.

I wouldn't be surprised if quarterbacks look at the new salary rules and realize that less money and a longer career with a strong team will be the new paradigm. Some will make the mistake to go for the money and lose as the environment changes. Others will be looking ahead (or their agents will) and make the smart bet. I think Ryan is a smart-bet kind of guy. I'd pay a preimium for him as a QB1.

An analysis I often see focuses on draft spot and productivity. I just don't think there's enough good data to make a definitive case. I think there are too many variables to take into account that make the analysis limited. For the sake of examining draft position of an NFL player and his productivity, I took a fantasy angle and like other types of analysis I've seen, the sample size skews results in directions that make the information sketchy.

I examined 93 quarterback careers in recent decades and I narrowed the number to 73 players to elminate those who never managed at least QB2 fantasy production. Then I determined the average years of productivity by fantasy tier per round of the NFL Draft.

Average Quarterback Years of Production By Fantasy Tier Per Draft Round

Tom Brady, Marc Bulger and Matt Hasselbeck make the sixth round a strong one. However, there are also players selected in rounds that no longer exist (Trent Green and Brad Johnson) and it makes these rounds appear to be some of the most productive. I'm showing this data because the spread of the information indicates to me that unless I'm really sold on the talent of the player and the situation, the safer route to acquire a good starting quarterback in dynasty leagues is free agency; not a rookie draft.

I think Andrew Luck, Robert Griffin, and Russell Wilson were exceptions to the rule. I'm even on the fence with Ryan Tannehill, becuase these four players are a part of an extraordinary class and you have to be willing to veer from the confines of data boundaries to generate a huge profit. The conundrum is when to leave the safety of these boundaries and no one can really teach you that kind of judgment - you have to live it.

These four second-year quarterbacks might be the next Brady, Manning, Brees, and Rodgers, but that's a tough bet to make if you have any knowledge and experience of the dynamics of the game beyond possessing a typical fan's infatuation for one of them. As I said earlier, paying the piper once generally saves you roster spots and lessens the likelihood of trading away players you liked in order to acquire a position you would have otherwise needed to address. There is often an emphasis on monetary savings or value of investment. However time, effort, and roster spots also impact your bottom-line performance.

Career Window Totals

The last thing I did with this data was convert the number of years each player had at each fantasy performance tier and caculated the percentage of his career producing at these tiers. For instance, Peyton Manning has 14 years with starter stats. He has been an elite fantasy quarterback eight of those years - 57 percent of the time.

10+ Years

What I like about seeing the percentages is that it's clear that betting on acquiring an elite quarterback is a losing proposition and the safer way to go is to pursue QB1s with "stronger upside". This is why I'm a proponent of Matt Ryan's long-term prospects.You might have to pay a little more for production that may never reach the elite tier, but I think it will remain close enough for the next 2-3 years to be worth the investment - see below.

5-9 Years

Ryan also has a good chance of having a career of 12-14 years, which means he's only halfway through his high-production cycle. These tables also got me thinking more about the concept of career windows. How many career windows do we get out of quarterbacks within each fantasy tier? If I want to begin viewing dynasty rosters in three-year increments then this Career Windows concept could be a nice shorthand. We'll see if it has a place after I begin studying other positions.

Average Years of Production Per Fantasy Tier

What is worthwhile is examining the performance of proven fantasy starters based on the percentage of their careers as no worse than a QB1. The table above shows that the average quarterback that has been a QB1 at least 35 percent of his career also had a long career average length (12.6 years) a high average of years as a QB1 (7.4) and nearly one full Career Window (0.9) as an Elite passer (2.7 years). This continues to solidify my feeling that paying a little more for an established starter might still be a worthwhile investment considering how many years you get out of quarterback that "hits."

A quarterback with at least 50 percent of his career as a QB1 has an average career length that's more than a season longer than the 35 percent tier and his yeas as an elite passer are a half a career window longer (1.5 years) and he has 1.6 fewer years as a non-starter (combined QB2+Reserve+FA). It's becoming more apparent to me that the general approach to quarterbacks in dynasty leagues to save time, picks, and long-term expense might be "win now" mode rather than "home grown."

The top of the heap - the future Hall of Famer - is a quarterback with at least 75 percent of his career as a QB1. Thus far that list in the past 10-12 years is Peyton Manning, Brett Favre, Tom Brady, and Drew Brees. Since I'm looking at tiers of production relative to each season and not straight production, it also allows me to look deeper into the past if I wanted. Dan Marino was a QB1 82.4 percent of his career; Jim Kelly a QB1 72.7 percent of his career; and Steven Young 52 percent of his career. Johnny Unitas was a QB1 67 percent of his career. Considering that these top-notch players have career lengths of nearly 15 years; 12 of those are QB1 quality; and 6 are elite in nature with only 2.3 years as non-starters (QB2+Reserve+FA), why not buy a top QB at age 28 or 29 and bet on another 6-8 seasons of QB1 production?

Matt Ryan looks better and better at a premium price.

Another thought is to make deals for quarterbacks past the age of 32 and bet on 4-6 years of quality production at a what usually comes at a lower price. What I have paid to acquire Tom Brady and Peyton Manning in dynasty leagues the past two years has been worth it to my team - a first or second-round pick plus a starter. Generally my picks weren't going to net me a shot at Robert Griffin or Andrew Luck and even so, if I want these two bad enough, I'm beginning to think three years from now I can still leverage a deal that is worth the price long-term.

Even Tony Romo looks pretty good here because no one reasonable is going to demand the same premium as Brady, Manning, or Brees but you should get steady production that can make a dynasty team competitive for 4-6 years.

Re-Draft Implications

I think this info should help readers see that players with quality production for a stretch are reasonably good options to rebound with other teams. Kurt Warner, Drew Bledsoe, Peyton Manning, Drew Brees, and Michael Vick fit this description to some extent. I think it makes Carson Palmer a player to watch this year. I also think good producers with bad years are also worth considering for rebound seasons. Philip Rivers might be worth consideration. On the other hand, Matt Schaub had two good years in a row surrounded by non-starter seasons the rest of his career. The bolstering of the Texans receiving corps is promising, but has the system really changed from a run-heavy attack? It doesn't seem worth the risk.

On the other hand, the Bears have been trying to be pass-heavy for a few years with Jay Cutler and finally have the coach, quarterback, receivers, and system aligned to make Cutler's season look promising. The big obstacle is the offensive line, but how much? Cutler's production history (see above) places his career squarely at the crossroads among all other fantasy passers in 2013.

Dynasty Rankings Implications

I still have the other positions to study, but if there are no surprises I'm inclined to think that the idea of placing elite quarterbacks at the top of any dynasty rankings list may still have merit. Why should I move Drew Brees or Tom Brady down if either one dips below the fifth spot at their position for a season after displaying multiple years within the top 3-4 at the position? I may have to view these players as more immune to short-term changes relative to other positions. Certainly there will be reasons to go against this idea, but it makes me think that losing the likes of Wes Welker or Marvin Harrison doesn't matter much for a player like Brady or Peyton Manning. Look at Drew Brees, whose best receivers are slot guys like Marques Colston and Lance Moore. When has the Saints quarterback had a receiver like Harrison, Reggie Wayne, or Andre Johnson?

I also think this information could elevate the way I value mid-range QB1s like Matt Ryan, Tony Romo, and perhaps Cam Newton. It could also make them less immune to a one-year dip in production. I think factors that will be required to consider a drop in rankings will need to be a substantial change with the offensive line, significant injury, or team change. Even so, none of those things have hurt Manning.

Where I may downgrade quarterbacks is the tier that falls below the top 12-15 at the position. Promising reserves (Ryan Mallett) or starters everyone is waiting to maximize their fantasy potential (Ben Roethlisberger, Jay Cutler, and Josh Freeman) no longer have the same value to me. They may not drop that far in the way I reformulate my dynasty valuations, but I can envision rankings that aren't waiting on passers to be something full-time that they've only revealed in snippets.

I'm beginning to think quarterback - more than any position - should be regarded as a "prove it to me before I pay for you" part of your dynasty rosters.

Matt will be writing about running back values through the dynasty lens in the next 4-6 weeks.