Print and Go Back ESPN.com: 2012 [Print without images]

Monday, November 19, 2012
Updated: November 20, 10:58 AM ET
Schedules can help trade decisions

By Tristan H. Cockcroft
ESPN.com

It's the age-old question during the critical fantasy football weeks: go with the best players or the best matchups?

Fantasy football owners are notorious for overthinking, especially in the latter stages of the season. They cling dearly to superstars in October, only to doubt said standouts' matchups in December. They'll cast aside matchups in Week 2, only to put too much emphasis on "good on paper" matchups in Week 15.

The delicate balance between skill and matchups should have influence on your trade plans. With two days remaining until the deadline in ESPN standard leagues, what better time to garner some insight on which is more important?

History helps provide the answer.

Using the past three seasons' (2009-11) worth of data and statistics from Weeks 12-17 -- Week 12 kicking off with Thanksgiving in each of those years -- I examined each week's highest-ranked players and most favorable matchups to determine which provided the greater probability of fantasy success. "Highest-ranked players" are defined by where the player is ranked at his position in year-to-date fantasy points scored, while "most favorable matchups" are defined as how our Points Allowed page would have ranked the given defense against that particular position entering the week.

For example, as 2012's Week 11 dawned, Drew Brees was the leading fantasy point scorer among quarterbacks (195), so he'd be labeled the "No. 1 quarterback." His New Orleans Saints, meanwhile, had afforded opposing quarterbacks the most fantasy points per game (21.4), so they'd be labeled the "No. 1 quarterback matchup."

You can read the detailed analysis, position by position, below, but here is the summary of my conclusions:

• You want the best quarterbacks over the best quarterback matchups and the best tight ends over the best tight end matchups.

• You want the truly elite running backs -- the Arian Fosters, Adrian Petersons and Marshawn Lynchs -- and you want to avoid extremely poor running back matchups.

• You want the best wide receiver matchups, not the best wide receivers.

• You don't want to take our Points Allowed page too literally.

Quarterbacks: Get the best

There's no more telling fact than this: The top five quarterbacks averaged nearly five fantasy points per game greater than quarterbacks ranked eighth to 12th -- or within two spots of the cutoff for starting spots in an ESPN standard league -- from Weeks 12-17 and nearly four fantasy points per game greater than quarterbacks who faced the five most favorable matchups.

Sure enough, in terms of total fantasy points accrued, the top five quarterbacks from Weeks 1-11 of 2011 (Aaron Rodgers, Cam Newton, Brees, Tom Brady and Matthew Stafford) still comprised the top five quarterbacks from Weeks 12-17 (ranked Brees, Brady, Stafford, Newton and Rodgers during that time). In fact, four of the five actually improved in terms of average fantasy points per game the final six weeks of the 2011 season. Rodgers was the exception, averaging 26.3 points in the first 11 weeks and 24.4 in the final six, but he was still third best at his position during that time, hardly concerning his owners.

Takeaway: Don't trade Brees, Rodgers, Robert Griffin III or any quarterback who has gotten you this far, and certainly don't trade Brady despite his Weeks 14-15 matchups against the Houston Texans and San Francisco 49ers, which rank among the eight worst matchups for a quarterback (before Week 11's games).

Running backs: Get mega-studs, avoid the worst matchups

It's good to be a top three running back, a group that averaged 14.6 fantasy points per game from Weeks 12-17, but the drop was precipitous from the fifth spot and beyond as Nos. 5-14 averaged 9.6 fantasy points per game. You read that right: Top three running backs provided a five-point advantage over players occupying the back half of the RB1 and front end of the RB2 classes.

Meanwhile, divide up running back matchups into four groups -- top eight, next eight (9th-16th), next eight (17th-24th) and bottom eight -- and each of the upper three groups (top eight, 9th-16th, 17th-24th) resulted in averages nearly three points better than that of the lowest tier (bottom eight). Incidentally, those top three groups had practically identical averages to one another.

There's no more conclusive evidence that you want no part of matchups against the very best run defenses, but otherwise, the matchups don't have much influence. But why not take it a step further: In Weeks 12-17 from 2009-11, the two best defenses at suppressing running backs' fantasy points afforded only nine games of 15 or more fantasy points, but on 18 -- or twice as many -- occasions didn't allow more than eight points to an opposing running back.

Takeaway: Run, run from Matt Forte, who has one of the worst remaining schedules for a running back. Every one of his seven remaining opponents ranks among the bottom 15 in fantasy points per game allowed to opposing running backs, and three rank among the bottom eight. Consider trading Stevan Ridley, whose Weeks 13-15 and 17 matchups come against bottom-five matchups.

Wide receivers: Matchups, matchups, matchups

Wide receivers in the study demonstrated a near-polar opposite result of running backs in the matchups column. Unlike running backs, top wide receivers -- whether No. 1 overall, top three or even top five -- provided scarcely any advantage over anyone ranked within the top 15 at the position. Top five wideouts averaged 10.4 fantasy points per game, while Nos. 11-15 ranked wideouts averaged 9.1.

When it comes to matchups, however, having the most favorable means everything. Again separating matchups into four groups, the uppermost group (top eight most favorable) of matchups resulted in a 21.7 fantasy points per game average, which was approximately three points per game better than matchups against any of the next three tiers, which had nearly identical numbers to one another.

Takeaway: Load up on matchups wherever possible, meaning those Atlanta Falcons, Dallas Cowboys and Philadelphia Eagles wide receivers might be worth the extra buck on the trade market. The Falcons play four of their final six games against teams ranked among the top eight in fantasy points allowed per game to opposing wideouts, meaning Roddy White should remain a WR1 from today forward (as would Julio Jones, depending on his health). A sneaky trade target: The Carolina Panthers' Steve Smith, 37th in fantasy points at the position heading into Week 11's action, faces the Oakland Raiders and New Orleans Saints in the final two weeks of 2012.

Tight ends: Matchups, schmatchups

Unsurprisingly, skills rule the day at tight end, as anyone who has owned Antonio Gates, Dallas Clark, Jason Witten, Rob Gronkowski or Jimmy Graham over the years can attest. They have all been multiple-year studs, and sure enough, top three tight ends averaged a whopping 9.2 fantasy points per game from Weeks 12-17. The next 10 tight ends, Nos. 4-13, averaged 5.8.

Meanwhile, tight ends who faced the three most favorable matchups averaged 6.9 fantasy points per game, while those who faced the top eight averaged 6.6 -- the latter identical to the average of those who faced the eight least favorable matchups. You read that right: Tight ends that faced the most favorable matchups performed no better than those who faced the worst.

Takeaway: Well, don't trade Graham, duh. It also supports Tony Gonzalez's and Heath Miller's candidacy for staying put rather than being traded.

Regression to the mean

It's a key factor in matchups analysis, a lesson fantasy owners need to learn (in any sport) and the reason for the final of the four short conclusions earlier in the column.

The simple explanation for regression to the mean is that if a number lies at either extreme -- such as the Buffalo Bills' 25.4 fantasy points per game allowed to opposing running backs entering Week 11, more than three points per game more than any other defense has afforded -- upon subsequent measurements, such as adding the statistics accrued in their next game, the number has a greater chance of being closer to than farther from the league average.

In other words, fantasy owners looking at that 25.4 number shouldn't expect that every future Bills opponent will accrue 25.4 points or more from their running backs in a given game. More likely, future opponents will score fewer than that -- anywhere from 10 to 18 to 24 to 25 -- even if the Bills remain the most generous defense for opposing running backs the rest of the season.

The numbers from the past three seasons corroborate this theory. Combining the 2009-11 seasons, the eight worst teams averaged 20.3 fantasy points per game allowed to opposing running backs and the eight best teams averaged 13.2. Those are the totals, so to compare, using the statistics from this column's study -- the week-over-week numbers -- the eight worst teams averaged 17.6 fantasy points per game allowed to opposing running backs and the eight best teams averaged 14.9.

In other words, defenses that resided at either extreme midseason exhibited some regression to the mean. Many of the teams with top-ranked weekly positional matchups might have dropped in the rankings, being replaced by different teams the following week. As we constantly remind, defensive performance is volatile, not only from a year-to-year basis but also week-to-week.

It's for that reason I remind never to put excessive stock in the Points Allowed page, even when matchups history might look especially favorable or unfavorable. The point is that matchups matter in certain instances, such as the ones listed above, but they're not always as cut-and-dry as the numbers on the page.