Tuesday, August 10, 2010

IN DEFENSE OF POINTS-PER-RECEPTION LEAGUES: As we gear up for ALOTT5MA FFL 2010 (invitations soon; need to confirm who's returning from 2009 first), a guest essay from Dan Suitor:

As the heat of summer turns into the slightly lesser heat of early fall, our country is faced with a series of important choices. We're buffeted at all sides by the opposing parties, with ideological arguments and rhetoric flying this way and that like so many scattered leaves. Our decisions in these coming months will shape the path we take, perhaps not just this year but for many to come.

So, with your future at stake, the question is: who are you taking in the first round of your fantasy football draft?

So many breathless pundits are worried about the redistribution of wealth, yet I am more concerned about the redistribution of fantasy points. While the majority of the past decade was marked by the preponderance of workhorse running backs, the “bell cow” is rapidly becoming an endangered species (interesting that it's always an animal metaphor).

In the year 2000 there were 19 running backs with 250 or more rushing attempts, and nine who tallied 300 or greater. Those figures held pretty constant over the decade; from 2000 to 2006 the NFL averaged 17.14 rushers with 250 or more carries, and 10 with 300 or more, per season. Those means have low standard deviations to boot (1.46 and 1.41, respectively).

This trend has been derailed over the past three season, and in a dramatic way. 2006 saw 17 rushers with over 250 carries and 10 with over 300. By 2007 those figures were down to 12 and 6 (decreases of 29.4% and 40%), and last season saw them tumble even further to 9 and 6 (52.5% and 60% of the 2000 to 2006 means). Even worse, the concentration of carries (in this case standard deviation) has become markedly diluted . In 2000, the standard deviation of carries among the top 50 rushers was 99.1*, and in 2006, that number was 81.6. For the past three seasons, the standard deviation among the top 50 rushers has averaged 70.1. What all those pretty numbers tell us is that the dichotomy between the workhorse and the journeyman is rapidly shrinking; on average, teams are spreading out their rushing attempts.

*This sample group included four quarterbacks: Kordell Stewart, Donovan McNabb, Daunte Culpepper, and Rich Gannon. This is notable because quarterback carries tend to stay relatively steady year-to-year, and a larger pool of running backs will push them further down the list. That said, the 2000 season may have had the greatest collection of rushing quarterbacks the league has ever seen.

Simply put: the workhorse running back is extinct. Welcome to the age of the running back by committee.

So what, then, does this mean for fantasy football? The running back has traditionally been the foundation, capstone, and window dressing of the average fantasy roster. Perhaps it is time to shift the focus of the fantasy game away from the noble halfback, and onto the stalwart flanker. Elite wide receivers have always been seen as worthwhile investments, but what if we could find a way to give the number two or three wideout the same juice that he number two or three running back has typically had.

To that end, I present to you the point-per-reception (or PPR). It's quite a phonetic concept, in that it is exactly what it sounds like: a rule that provides your fantasy players with a point for every ball they catch.

The point-per-reception (or any variant thereof, such as .5 PPR or .25 PPR) came into vogue in the 90's as a way of enhancing the value of wide receivers during time when there were lots of workhorse backs dominating lots of leagues. If you didn't have an elite rusher, like Emmitt Smith or Barry Sanders or Marshall Faulk or Terrell Davis or who-have-you, you probably weren't winning the league.

The PPR acts as a sort of equalizer. It bumps up the point totals of elite wide receivers by 70 to 100 points per year, putting them in the same echelon as the elite running backs. For example: last year the top six running backs scored 343, 267, 261, 232, 223, and 220 points respectively. The top six wide receivers scored 209, 206, 204, 192, 182, and 180 points. Demonstrably fewer, correct? Adding a point per reception, the points for the top six wide receivers go to 310, 287, 287, 255, 267, and 273 points.

Of course, running backs get a boost too. The top six running backs would've had their scores bumped up to 393, 310, 313, 310, 242, and 272, but those are the elite players. The top 50 running backs averaged 140.76 points apiece, while the top 50 wide receivers come in at 126.18. If we add a point per reception, we get averages of 172.42 for the RB's, and 191.8 for the WR's. Given that most leagues play more wide receivers than running backs, it works out pretty well.

Furthermore, it allows for more strategy in the game. It gives value to players like possession receivers and scatbacks who might not otherwise have a place in the fantasy game (see: Burleson, Nate or Sproles, Darren). Given how often running backs are in committee situations these days, we might see an even further dilution of the running back pool.

Football is a cyclical game; today's Run and Shoot is yesterday's Flexbone, today's Wildcat is yesterday's Single Wing. Someday there will be scores of running backs grabbing 300 carries a season, but that season is not going to be this season. Instead of fretting yourself into nervous fits over which platoon-mate to start in a given week, or watching helplessly as the goal-line back vultures a touchdown from your player, just imagine a peaceful world where wide receivers help your team as much as running backs, and where getting your hands on Wes Welker is as exciting as drafting Adrian Peterson.

The point-per-reception is calling you, my friend. Don't let her linger.

8 comments:

  1. If there's only 6-7 reliable every-down backs -- Johnson, Peterson, MJD, Gore, Jackson, Turner, Mendenhall, Grant (b/c Rice gets vultured) -- that seems to make a stronger case for PPR because not everyone's going to get even one of these guys.  Still, "receptions" aren't real-world useful in football -- yardage is -- so this still feels wrong. 

    Maybe the answer is to move in the traditional yardage bonus for long TDs from 40+ out to ... 20+ out?  Would that do a fairer job of recognizing WR talent?

    ReplyDelete
  2. isaac_spaceman10:45 AM

    The thing is that nobody knows who the reliable every-down backs are until midseason.  Picking at the end of the draft the last few years, I've had Chris Johnson/Thomas Jones (2010, rounds 2-3), Marion Barber/LenDale White/Chris Johnson (2009, Rounds 2, 4, 12), Bryan Westbrook (end round 1), Addai/Barber (2008, Rounds 4, 9).  The good unknown RBs are out there.  You just have to be willing to spend a few picks to find them.   

    ReplyDelete
  3. StvMg2:26 PM

    I was in a point-per-reception league last year. Hated it. Of course, it didn't help matters that I tried to capitalize on that scoring system by selecting Matt Forte with the second overall pick. Oops.

    ReplyDelete
  4. Dan Suitor11:20 PM

    I think the PPR is less about rewarding the reception as a pure accomplishment then it is about rewarding certain types of players, thus bringing the goal set of/valuation in fantasy game a little more in line with that of the real deal. Possession receivers and Third-down backs are incredibly important to actual success (see Wayne Chrebet or Mike Alstott), so to give them PPRs is to recognize their unique, yet equally praiseworthy, talents.

    Also, thank you so much for running this. The  length crept up on me, but it was incredibly fun to write.

    ReplyDelete
  5. Dan Suitor11:22 PM

    Not to mention the 2 - 4 undrafted (or last 5 rounds) guys who will end the year in the top 30 in value.

    ReplyDelete
  6. Daniel Fienberg12:13 AM

    Matt Forte screwed me in not one, not two, but THREE fantasy leagues last year, as I drafted him everywhere from third to ninth in the first round. Well, not literally everywhere. But I drafted him in three different first round positions in three different leagues. This year, I shall take a different strategy...

    -Daniel

    ReplyDelete
  7. But Chrebet is useful because he accumulates yardage -- which is covered -- while Alstott is a mix of TDs and 1st down conversion, the former of which is covered and the latter only gets catured by stats like DVOA, which no one has tried yet in fantasy.

    ReplyDelete
  8. isaac_spaceman11:43 AM

    I'm not sold on DVOA.  The year-to-year variability suggests that it may be more like a defensive baseball stat, where you really need three years of aggregated data to get a good read on true talent level.  The problem with that is that most RBs basically only have 3-5 useful years in them, so by the time you have a good read on their DVOA, they're nearing their sunset period.  The other problem with DVOA is that some of the factors it excludes (especially OL performance) are more or less fixed variables, not random.  So they're actually predictive, and therefore a lot of what you're excluding in seeking true talent shouldn't be excluded for the purpose of calculating an expected regression to the mean. 

    ReplyDelete