Who goes No. 1? Johnson or Peterson

I prefer Adrian Peterson. I know Peterson has his own blemishes. His yards-per-carry average has declined from 5.6 to 4.8 to 4.4 in his first three NFL seasons, he lost six fumbles during the '09 regular season, and the Minnesota Vikings' offensive line may be headed into decline. And just as you can argue that Johnson's insane skein of big plays last year might not repeat (he had seven rushing touchdowns of at least 30 yards), it's also understandable to be concerned that Peterson's whopping 14 touchdowns from inside an opponent's 5 will be hard to duplicate.

When making this decision -- Johnson versus Peterson -- I think there are two primary factors we need to consider. First, is it likelier that a running back repeats a massively big-play-infused season, a la Johnson in '09, or is it likelier that a running back repeats a high-touchdown season based primarily on short-yardage scores, a la Peterson in '09? Second, there's the question of Johnson's usage in '09. He led the NFL in carries and in touches from scrimmage, and he's not a big man. Is it likely that he gets as many opportunities here in '10?

Chris Johnson's Big Plays

Those seven touchdown runs of 30-plus yards that Johnson produced last season tied Jim Brown's single-season NFL record from 1958. (Of course, Brown did it in 12 games.) There have been only 26 instances in NFL history, dating back to 1940, in which a running back has even produced as many as four such touchdown runs in a season. Three of those performances were turned in last season: by Johnson (seven), Jamaal Charles (four) and Maurice Jones-Drew (four). None of other 23 backs managed to rack up more than half the number of long-distance scores the following season. This is certainly damning evidence to those who contend Johnson will break just as many long touchdowns in '10 as he did in '09:

*-did not play in 1999 because of injury

But it would also be foolish of me to proclaim that the majority of Johnson's fantasy value last season came from long touchdowns. Certainly they helped. In ESPN's standard fantasy scoring system, Johnson received roughly 86 of his 329 fantasy points (or 26.1 percent) on those long scoring runs alone (if we add together both the yardage points and the points for the touchdowns themselves); in leagues in which bonuses are given for longer scores, that percentage obviously would be higher. Yet a much greater percentage of Johnson's NFL-best fantasy point total came from his overall yardage passing and receiving: 2,509.

Yet I still don't think we have our arms around just how repeatable (or unrepeatable) such a big-play season is for Johnson. I'll give another, more anecdotal reason for concern. For the '09 season, Johnson accrued a whopping 693 rushing yards on carries that went for 30-plus yards (both touchdowns and non-touchdowns). That's about 35 percent of his rushing yardage. I think it's fair to say that if those breakaway runs are significantly diminished, CJ2K is coming back to the fantasy pack. According to the Elias Sports Bureau, Johnson's 693 such yards were the most in the NFL since 1997, when Barry Sanders produced 701 rushing yards on carries that went for more than 30 yards. How did Sanders follow up his incredible big-play season? In '98, he produced just 335 rushing yards of that type.

In fact, the parallels between Sanders circa '97 and Johnson circa '09 go further than just the incredible amount of yards gained on big plays:

And check out what Sanders did for a follow-up:

Again, I'm admitting up front that this is a cherry-picked argument. Not only am I looking at just one season (albeit the only vaguely recent season that compares in terms of big-play-heaviness), but Sanders was 28 years old in '97, with eight NFL seasons already under his belt, and '98 was actually his final season in pro football (though he made it through all 16 games, and obviously had a whole bunch of carries). By contrast, Johnson just finished his second NFL season, and will only turn 25 in September. Still, you'll have to forgive me for leaning toward the conclusion that there are very few (if any) precedents in which we see seasons as big-play-heavy as the one Johnson produced in '09 repeat.

Adrian Peterson's Vulture Scores

Since 1940, there have been only 13 instance when a running back has scored 14 or more touchdowns from inside opponents' 5-yard line (including Peterson's 14 in 2010) for a season, and there have also been only 37 instances in which a running back has scored at least 12 such touchdowns. Clearly, Peterson's incredible short-TD work last season was no run-of-the-mill phenomenon, either. His short touchdowns accounted for roughly 87 of his 265 fantasy points (or 32.8 percent). So I ask: Is it more or less likely that this type of extraordinary production will go away in '10, compared to Johnson's long runs?

Before AP did it, here were the 13 seasons of 14-plus short scores, and those players' follow-up efforts:

On average, these running backs saw their short scores level out at eight the following season. (It's also worth nothing that if you perform this same exercise with all the rushers who've scored 12-plus short scores in a season, you get similar results.)

So just as Johnson's bailiwick -- the ridiculous ankle-breaking long run -- seems likely to be less common in '10, Peterson's '09 specialty -- the vulture score -- figures to partly vanish as well. It's well within reason to assert that Peterson's short scores will decline from a whopping 14 to a more reasonable eight.

But what would that leave AP with? Since by definition he didn't derive many of his 1,819 yards from scrimmage via the short touchdown run, I don't think we have to forecast a significant decline in his yardage in '10, the way we do for Johnson. Fine. Give Peterson 12 total rushing touchdowns this season, but keep his yardage the same. And give Johnson a standard decrease in yards and touchdowns, per our earlier discussion. If we assume similar usage for each player from their stellar '09 seasons (more on that in a moment), we're left with this speculative 2010 comparison:

The advantage is still Johnson's, but the gap is far closer than it was in '09. This is already not a no-brainer.

The Curse Of 400?

And here's where Johnson's '09 usage comes in. CJ2K had 408 touches from scrimmage last year, which ranks as the 35th-most in a single season in NFL history. Before Johnson did it last year, there had been 40 seasons of 400 touches or more by a running back. You've probably heard about the Curse of 370, and Tristan Cockcroft has written a fresh piece about it for our draft kit. But that analysis focuses only on carries; if you look at that stat, well, Johnson had 358 totes, so you might conclude, "Phew, he barely escaped and he's good to go." But did he? Based on the following analysis, I'm not so sure. I think we may also be looking at the Curse of 400: 400 overall touches from scrimmage, runs plus receptions.

Therefore, my final big remaining question is: How do such high-usage backs (400-plus total touches from scrimmage) do in their follow-up seasons?

Well, of those 40 seasons I mentioned, we throw out four, which would tilt the scales unfairly against Johnson: Eric Dickerson ('87) didn't play during a work stoppage; Emmitt Smith ('93) held out for a better contract for two games; Ricky Williams ('04) retired; and Jamal Lewis ('04) was suspended for violating the league's substance abuse policy. Of the remaining 36 seasons:

• On average, these 36 rushers had 99 fewer touches than the year before.
• On average, these 36 rushers produced 85 fewer fantasy points than the year before.
• On average, these 36 rushers played in 2.1 fewer games than the year before.

• Five RBs (13.9 percent) had more touches the next season.
• 30 RBs (83.3 percent) had fewer touches the next season.
• One RB (Curtis Martin in '99) had exactly the same number of touches the next season.

• Seven RBs (19.4 percent) scored more fantasy points the next season.
• 29 RBs (80.6 percent) scored fewer fantasy points the next season.

Now, I suppose this isn't absolutely shocking. Even if a player doesn't get hurt the year after an extreme-usage season, you can make the argument that his touches are likely to go down simply because they were such anomalies in the first place. Thus I think the final piece of the puzzle is to see whether 400 really is a "magic" number, or whether players who log between, say, 350 and 399 touches in a season suffer the same decline in their subsequent seasons. (Remember: Peterson was in that 350-to-399 range last year: he had 357.)

Before last season, there were 122 running back seasons of between 350 and 399 touches. Throw out 15 of those (11 took place in either 1981 or 1986, meaning the subsequent year had a strike; Barry Sanders ('98) and Tiki Barber ('06) retired before their next seasons; Lydell Mitchell ('77) and Walter Payton ('77) did it in a year before the schedule expanded). We're left with 107 seasons. Let's see how much those players declined, compared to the declines we just talked about for the 400 club:

I don't think there can be any doubt that, on average, players who've just completed a season with Johnson's usage pattern do worse than players who've just completed a season with Peterson's usage pattern.

Let me also say that when you look at these lists of high-usage players, there are very few guys as small as Johnson, who's generously listed at 5-foot-11 and 200 pounds. I know that many folks like to defend CJ2K by comparing him to Sanders, another guy who made tacklers miss and was relatively small. But realize this: Sanders never once had a 400-touch season. Not once. So Chris Johnson literally just finished a season in which he was used more than Sanders ever was in his entire career. That should give us pause, I think.

Conclusions