But looking at a 4 game sample IS meaningful when calling Lewis a RB1, surpassing established RBs like Marshawn Lynch, and predicting dynastic (FF) studliness?
We have a 3 game sample (with Blount back), BBs historical usage of his RBs (since the offense became centered on Brady), and the # of catches, catch rate, and TD rate that are abnormally high. If you believe Lewis' talent outweighs those factors, then you view him as a RB1, if you don't, you view him as a RB2. I'm not saying which view will end up accurate, just posting information that I've thought about.
I can't speak for anyone else, but me calling Lewis an RB1 and predicting dynasty studliness has very little to do with his stat lines the last four games, (though those certainly don't hurt), and a hell of a lot to do with the fact that he is playing football on a completely different level from nearly everyone else in the league. It'd be like chastising the Todd Gurley fans for having him as the #2 dynasty RB over "just" 314 rushing yards. I mean, sure, the 314 rushing yards certainly don't hurt, but there's kind of a lot more going into it than that.
Lewis is just 25 years old at a position where most of the best producers are much older. Bob Henry- the most accurate projector in fantasy football last year- had him projected as the #5 PPR RB for the rest of the season last week. (No clue where he is this week, since the top 200 forward isn't out yet, but I'm betting he hasn't gone down.) He is dominating snaps, (if you count the snaps White got after Dion Lewis got pulled against Jax, he's gotten 75%, 85%, 65%, and 70%), on arguably the best offense in the NFL. He leads all NFL running backs in passing targets per game despite sitting out most of the second half against Jacksonville, (where his direct backup added four more targets). He leads the NFL in broken or missed tackles forced, having one more than runner-up Devonta Freeman on half as many touches. He's averaging five yards per rush despite not having a single carry go for longer than 13 yards. And he looks every bit as good on film as those facts suggest, leaping off the screen every time he touches the ball. He joined Rob Gronkowski as the only offensive players New England felt were important enough to be worth resting in the blowout against Jacksonville. The Patriots beat writer for ESPN says "they love Dion". He got a contract extension after just three games.
And, oh yeah, he's also put up a lot of fantasy points. But mostly all of that other stuff about him being an arresting young talent on an explosive offense.
And all of that is well and good, but another fact is that he is "only" getting 15 touches/game.
It's tough for a RB to put up RB1 numbers over the course of a season with that few touches. Hill, Charles, and Anderson did it last year (assuming 12 team, PPR league), only Woodhead did it in 2013, no RB did it in 2012, Sproles did it in 2011, & no RB did it in 2010. So, in the last 5 years, 5 out of 60 RBs (8.3%) managed to be a RB1 with the limited amount of touches Lewis is getting. Plus, this isn't an anomaly for BB; since 2007, he hasn't given his "pass-catching, smaller" RB much more than 100 carries.
Again, if you, or anyone else wants to believe that Lewis is going to buck both of those trends, more power to you. But don't (not directed at you, but rather the previous poster who criticized me for using a "4 game sample") say a 4 game sample isn't enough to say his carry potential looks capped, when he others are taking that same 4 game sample (with very high catch, catch rate, and TD rate #s) to support the "he's a RB1" claim.
The type of touches matters every bit as much as the quantity of touches.
Last year, running backs across the NFL averaged 0.578 fantasy points per rush attempt. They averaged 1.041 fantasy points per reception in standard scoring, and 2.041 fantasy points per reception in PPR. Therefore, receptions were anywhere from 80% to 253% more valuable, on a per-touch basis, than rushes.
Dion Lewis only averages 9 rushing attempts per game. That's a really low number. But he averages 5.75 receptions per game. If we convert those receptions into an "equivalent" number of rushes, that's like an extra 10.35 carries in standard scoring. Or, to put it another way, assuming both players produced perfectly at league average per touch, we would expect a running back averaging 9 rushes and 5.75 receptions per game to score exactly as much as a running back average 19.35 rushes per game in standard scoring.
And that's standard. I've tried to always be clear that the majority of my hype for Dion Lewis is in PPR. (I still think he's a top-12 dynasty RB in standard, but not anywhere like he is in PPR). In PPR scoring, those 5.75 receptions per game can be expected to produce as many extra fantasy points as an additional 20.30 rushes. (TWENTY POINT THREE ZERO RUSHES.) Meaning a guy averaging 9 carries and 5.75 receptions should be about as productive, all else being equal, as a guy averaging 29.30 rushes per game. (TWENTY NINE POINT THREE ZERO.)
And, again, this is over a 4-game sample where Dion Lewis was pulled for a substantial portion of one game and his backup logged an additional four receptions. So if anything, they understate the "true value" of Dion Lewis in the New England Receiving RB role.
Over the 1995-1996 seasons, 21 running backs touched the ball more times than Larry Centers, but only five scored more fantasy points in PPR. The number of touches is by far of secondary importance to the makeup of those touches. 15 touches a game is *plenty* when you're leading all running backs in targets per game.
In fact, if we want to go with historical trends, there have been 32 running backs since 2002 who averaged 6+ targets per game. 30 of them either finished the year as a top-12 RB in PPR, or were on pace to finish the year as a top-12 RB but missed time to injury. Included in those 30 are Shane Vereen, whose 44 rushes and 47 receptions in 8 games in 2013, (just over 11 touches per game), still would have been good enough to land him as the #6 PPR running back if he'd played the whole year. Darren Sproles made the cut twice, in two different seasons where he averaged fewer than 6 rushing attempts per game.
(In case anyone's curious, the two backs who failed to finish on pace for RB1 production were Michael Pittman 2003 and Jamel White 2002.)
One of three things is going to happen this year. Either Dion Lewis is going to get hurt, or Dion Lewis is going to have his role in New England change *dramatically* (to the point where he's averaging under 5 targets per game going forward)... or Dion Lewis is going to finish as a top-12 running back in PPR.
I'm not really in the "predicting injury" business, but I would note that fewer touches does represent fewer chances for a player to get hurt. I suppose the second thing is possible, but doubtful; ESPN's Boston beat writer already weighed in that Lewis was going to be getting the strong majority of the snaps going forward, and that he's the guy who will be getting all of the play when New England spreads the field and goes pass-heavy. And other than one of those two things, Lewis is a lock for RB1 production this season. And if New England keeps him around for next year, he'll be a lock for RB1 numbers in PPR next season, too.