Monday, July 13, 2009

Does a great fantasy receiver need a #2?

Like to the running back conundrum I questioned last year, another fantasy football paradox popped into my head over the weekend as I was wondering what effect the departure of T.J. Houshmanzadeh might have on Chad Johnson in 2009. (I really need a life.) Viking fans will also recall that, for all the hubbub on what effect losing Cris Carter would have on Randy Moss, Moss went on to have his best season, yardage-wise, in 2003. I've often heard two lines of argument about drafting wide receivers:

Situation A: A #1-caliber receiver has no good #2 on his team
"He's all they've got! They have to throw to him! He'll have a great year!"

Situation B: A #1-caliber receiver has a good #2 on his team
"Opposing defenses have to cover #2, as well! #1 will have a great year!"

Yeah, I'm gonna have to look into that.

So I spent an inordinate amount of time on pro-football-reference.com looking at all receivers for the last 10 years with at least 1,200 yards receiving and looking at their "#2" receivers in that season, so see if there was any correlation between great receiving seasons and particularly good (or bad) seasons by complementary receivers. There were 95 seasons that matched this criteria -- 93 by wide receivers and two by Tony Gonzalez. The reason I limited it to 10 years was because, frankly, I would have to look up each player's "#2" on his team's page for that season, which took long enough as it was. 95 seasons is probably enough to give us a reasonable sample size, and by limiting the study to the last 10 years, I keep it firmly rooted in the "modern" NFL with its oft-explosive passing game.

I use "#2" in quotation marks because, several times, a team had more than one player with 1,200 yards receiving, meaning that one player's "#2" receiver actually racked up more yards than him. Specifically, there were 24 such pairings (twice counting 12 different sets of players) from 1999 to 2008, from Torry Holt (1,635 yards) and Isaac Bruce (1,471 yards) in 2000 to Jimmy Smith (1,213 yards) and Keenan McCardell (1,207 yards), also in 2000.

I took the 95 pairs and sorted them by the total yards for the #1 receiver. I then split the #1 receivers' seasons into three parts, of 32, 31, and 32 players. The top 32 had the best seasons, the middle 31 had the second-best, and the bottom 32 had the third best. If there is a correlation between great seasons by #1 and great or not-great seasons by #2, we should see some sort of significant difference in their corresponding #2's yardage totals. Here's what I got:













Avg. #1 Yds.

Avg. #2 Yds.
Top 32
1,495958
Mid 31
1,336

932
Bot 32
1,241950


Doesn't seem like much of a difference between the three categories. The correlation between the two sets of numbers is -0.06, which also indicates that there's virtually no connection between yardage totals for #1 and yardage totals for #2.

A few other interesting stats...

Greatest difference between #1 and #2: 1,122 yards (Steve Smith/Ricky Proehl, 2005, 1,563 to 441)

Smallest difference between #1 and #2: 4 yards (Hines Ward/Plaxico Burress, 2002, 1,329 to 1,325)

Most frequent #1-#2 pairing: Torry Holt/Isaac Bruce (2000, 2001, 2002, 2004)

Most 1,200 yards seasons, 1999-2008: Randy Moss and Marvin Harrison (6 each)

And Terrell Owens, in his five 1,200 yard seasons, has had a different #2 in each one: Jerry Rice, J.J. Stokes, Tai Streets, Brian Westbrook, and Jason Witten.

Now, this study isn't perfect. Sometimes, a #2 puts up poor numbers not because of a lack of talent, but due to some other reason, such as injury. For 2008, Anquan Boldin is a perfect example. Had he not missed four games, he almost certainly would have cracked 1,200 yards in his own right (he had 1,038), thus altering not only Larry Fitzgerald's numbers in my data but adding a new point of his own. Calvin Johnson losing Roy Williams after just five games also might have had some impact on his numbers. (Shaun McDonald's 332 yards receiving for the 2008 Lions -- fittingly -- makes him the worst "#2" in my data.) Still, it could also be argued that Fitz and CJ put up their good numbers without a solid #2 for part of the season (and no, I don't count Steve Breaston), even if you could theoretically add together the yardage numbers for several receivers and paint a more accurate picture of their "#2 receiver."

Another thought is that a #1 receiver, especially one who's having a great season, is going to get a lot of balls thrown his way (similar to the argument in situation A) and the #2, by default, isn't going to get as many passes thrown his way and, therefore, have worse numbers. There might be something to that, but I think the effect is minimal.

Still, when it comes to trying to pick a wide receiver, at least for fantasy football, I believe it comes down to not thinking too hard: Pick the best guy, period. You can take QB and best offensive philosophy (a la the current Patriots or Cardinals or the early-2000s Rams) into account, but don't overly concern yourself with his other receiving teammates, either for the positive or the negative.

6 comments:

BJ said...

It's posts like this that remind me why I love reading your blog. Good stuff.

Jason said...

Thanks!

Peter said...

I wonder if this also applies to picking fantasy QBs and putting too much stock into them gaining or losing a strong WR2.

For example, would Warner's value suffer if Boldin were traded?

Would Favre/Rosey/Jackson benefit from an effective Rice or Harvin (or combination of the two) and Berrian continuing to be a decent WR1?

Jason said...

That's tough to quantify, because you'd have to have some way to separate QB/WR value. And year to year, players can go up or down in stats for reasons other than gaining/losing a player. Even with Fitz and Boldin, Warner could very well get injured, start turning it over like a madman, etc., this year.

One thing I did discuss with a friend many fantasy football drafts ago was: "Would you rather have a good QB with bad receivers or a good WR with a bad QB?" The answer has generally been "Good QB with bad receivers." Look at what Drew Brees has done with that bunch of yokels in New Orleans, and look at what Randy Moss has done without a decent QB throwing him the ball. Not real scientific, but it's been enough for me in recent years.

Peter said...

That line of thinking (good QBs improve WRs more than good WRs improve QBs, with which I happen to agree) suggests that Cutler will have a better year than Orton this year. I've read a lot of good and interesting articles saying Orton's numbers will improve and Cutler's will decrease, and it's hard to argue with those. I'll be very curious to see what happens with those two this year.

Jason said...

Well, that's assuming Cutler's better than Orton to begin with. I think that's at least somewhat debatable. Their average and rating stats are pretty similar; Cutler's numbers look better, undoubtedly, because he threw a lot more passes, and debateably, because he had a better supporting cast (O-line and receivers). I think it'll be really interesting to see who actually does better this year.