Writers of Pro Football Prospectus 2008

Most Recent FO Features

BradyTom00-10.jpg

» Week 4 DVOA Ratings

Five different teams from last year's DVOA top eight rank in the bottom half of the league through four weeks of 2014. What can we learn from other teams with similar starts in the past?

15 Aug 2007

Scramble for the Ball: Fantasy Mailbag 07

by Bill Barnwell

SCRAMBLE FOR THE BALL SPONSOR
Attention Football Outsiders Readers: Get a free week of winning NFL picks from Doc's Sports Handicapping Service. You will receive the same picks Doc's paying members receive from all 5 handicappers on their "Advisory Board" and there is absolutely NO OBLIGATION what so ever for this offer. Call 1-866-238-6696 or email service@docsports.com with "Free Football Outsider" in the subject line.

 

On the first day of Football Outsiders -- as the story will be told to youngsters and any interested and well-connected biographers -- the Great KUBIAK Mainframe begat two columns on an Internet previously beset by pestilence and vaguely-informed football analysis. The first piece was Aaron's story on winning to run. The second was "Scramble for the Ball," Ian Dembsky and Al Bogdan's joie de vivre as directed towards fantasy football, gambling, professional wrestling, Rod Smart and other spurious yet fascinating ventures. It's been four seasons and three writing combinations since then, the last of which was myself and Ian in 2006. With Ian bowing out this year due to time constraints, Catholic Match Girl not returning my calls after she discovered that ... other ... girl's number in my cell phone, and other FO personalities lost and confused in the woods and/or doing a Mr. Magoo impersonation, it's my duty to inform you that this year, Scramble is a one-man show. I'll be taking over the duties myself, but don't count out the possibility of any special guests appearing as the season goes along.

With all that being said, don't expect much of a change. Scramble remains FO's column dedicated to fantasy football, football betting -- just for informational purposes, of course -- the Loser League, and other goofy topics that aren't covered in any of the other articles FO runs each week.

While the real thing is still a month away, we're now smack-dab in fantasy draft season -- the most wonderful time of the year. For owners who struggled last year, it's a chance to start anew and redeem themselves; for the owners who did not, it's a chance to gloat and build upon last year's success. While you can't bring an Outsider of your choice to your fantasy draft for anything less than four figures, we've created a forum on our new Discussion Board dedicated to discussing fantasy football questions. While the board is primarily designed for discussion between users, myself and other FO writers will be making appearances on the board as time permits. For this first Scramble of the new season, we'll field some of those questions posed on the board over the last week.

SJM: I'm in a 12-team league that starts two QB (and two RB, three WR, and one TE, K, and DEF). Standard scoring for WR and RB (plus 1/2 point per reception). Scoring for QB as follows: Completions, 0.2 points; passing yards, 40 yards per point; passing touchdowns, six points; interceptions, -4 points; fumbles lost, -4 points.

So my question is how do I draft? Do I throw all the quarterbacks and running backs into one big pile and take two of each in the first four rounds? Are quarterbacks still worth more because there are fewer good ones? Should I draft a wide receiver while I still have a starting spot open for a quarterback or running back? I'm afraid your answer will be "Buy the KUBIAK spreadsheet and good luck," but I'm hoping for more.

Bill: It's an interesting scoring system. I don't have fumbles data in front of me, so I'm ignoring that at the moment, and I'm assuming your standard scoring gives a point for every 10 yards a running back or wide receiver gains. If you apply the data back to the 2006 season, the top 12 players in the league are seven running backs (Tomlinson, Jackson, Johnson, Gore, Barber, Westbrook, Parker), four quarterbacks (Manning, Brees, Bulger, Palmer), and one wide receiver (Harrison). When you expand that out to 72 players, or the first six rounds of what would have been a perfect draft, there would have been 19 quarterbacks selected, 26 running backs, 25 wide recievers, and only two tight ends.

Now, that's solely based upon points gained, without any allowance for how to draft. That's where value-based drafting comes into play, the drafting methodology from footballguys.com. While the essay linked does a fantastic job of explaining the principles and ideas behind the VBD strategy, let's quickly mock up a VBD-style strategy by basically calculating each player's fantasy points above replacement -- namely, the value a player has over what would be considered freely-available talent. In this league, let's assume that each team has one backup quarterback, running back, wide receiver, and tight end, so that the replacement value for quarterbacks, for example, will be the 37th-best quarterback in the league, who would be the first quarterback available as a free agent.

The new drafting strategy, accounting for positional scarcity, does change things. The top 12 now features seven quarterbacks and only five running backs; the first six rounds of this draft, based upon 2006 data, would see 22 quarterbacks, 20 running backs, 22 wide receivers, and eight tight ends go amongst the first 72 picks. In short, if quarterbacks are a scarcer quantity in this league than running backs (or vice versa), it's not by much. Of course, that's based upon 2006 data; while you can do the same with our 2007 projections, well, you will actually have to buy the KUBIAK spreadsheet for that.

SuupaJim: I'm in a 10-team league with one QB, two RB, three WR, one TE, one Def, one K, sux bench and no flex spot. The scoring is pretty standard with the exceptions being 25 passing yards for a point, six points per passing TD and 0.5 points per reception.

I'm planning to pick somewhere between four and seven and grabbing two running backs in the first two rounds. So with that middle pick in the third round, I'm not sure if I should take Gates or Benson. Normally I would not want to take a bench player with my third round pick, but Benson seems to be a great value pick there.

Bill: I'm not so sure about Benson being a great value there. ESPN's draft results have Benson coming off the board at No. 25, which is right where he would be if you were in that four-to-seven range come the third round. He'd be a good pick there for a team that needed a running back, but if you take a running back with your first two picks, it's not a real value. I'd be comfortable picking Gates there, who would represent a real value pick in your system -- the quick and dirty VBD method described above values Gates as the 13th best player in the league.

nojo: I'm in a keeper league (two keepers), deciding who to keep. I'm definitely keeping Frank Gore (costing me a seventh-round pick). I'm deciding between Maurice Jones-Drew and Cedric Benson (each would cost a 16th-round pick).

Scoring is six points per TD, three points per 25 yards rushing (after the first 25), and three points per 25 yards receiving (after the first 25). starters: one QB, two RB, two WR, one TE, one D, and one K.

I know that PFP 2007 has [both Jones-Drew and Benson] projected to produce almost exactly the same numbers this year. I feel like Benson might do better this year (clear-cut starter, and MJD might lose TDs to Greg Jones in addition to the touches he loses to Fred Taylor), but MJD might have a better long-term outlook.

Bill: Well, the good news about this decision is that you're extremely likely to have a stud running back regardless of who you pick. This topic was actually addressed in a recent blog post by Chase Stuart, who contributes to the blog at Doug Drinen's fantastic pro-football-reference.com Web site.

In the two entries Doug's written at the time this column is being written, he's detailed the process of creating a formula that projects for future running back performance based upon several factors. While Doug's entries are must-reads for anyone who enjoys the work we do at FO, Doug's research (at least, as of Part II) finds that Maurice Jones-Drew is projected to score 1,277 more fantasy points over the rest of his career, while Cedric Benson, based upon his rookie performance, is only projected to score 317. Doug also points out that the only two rookie running backs in NFL history who've averaged more yards per carry than Jones-Drew are Franco Harris and Clinton Portis, so Jones-Drew is in pretty rarified air. When you factor in Benson's injury problems, I would go with MJD.

Check out the Football Outsiders comics archive and Jason's wacky Gil Thorp blog.

Andrew: I'm in a 12-team standard league where the commish, for some reason -- mainly due to members being scattered all over the place time-zone wise, including Europe -- has mandated an automated draft. All we can do is alter our rankings of how we view players and let it pick them for us.

Now ... if I know these people, at least half of them won't alter their rankings at all and will use the baseline from Yahoo sports.

So, is the best strategy to rank each and every player completely, or to tweak a few I particularly want higher than they're forecast to try to manipulate it into getting those, likewise modding a few specific ones down (e.g., Larry Johnson) to make sure I don't get those?

Bill: Ah, fleecing the lazy and/or uninformed. I'm a huge fan. Last year, I graduated from university and decided to take a trip with some friends to Las Vegas, where I would employ my honed no-limit hold 'em skills to make possible a wonderfully comfortable vacation. So after a week, when my friends left and I was still making money, I stayed. I left after three and a half weeks, all of which had been profitable, mainly because I couldn't find a hotel room for under $350 a night. Stupid conventions.

The thing is, I'm a pretty mediocre poker player. I know what I'm doing against average or worse players, but I'm no match for really advanced players. How was I able to sustain this? Every night, I'd go to Caesars' Palace, where the poker room is situated directly next to the Pure nightclub. Each of the nights I was there, the poker room would be flooded with a fresh crop of drunken, unsatisfied young males, who would seek to reclaim their masculinity through poker. Unfortunately, they'd run into me, sipping Diet Cokes, and give said money away.

Why am I tooting my own horn? Because it's the same idea. When you play poker against people who have no idea how to play or are too drunk to care, you don't make bluffs or sophisticated plays -- the opposition is too naive or foolish to care or even recognize it. Instead, you play the most conservative, straight-forward game you possibly can. The same is going to be true in your draft. Don't be too concerned about pushing the guys you like up too far, because you might overvalue them with your list when you'll likely be able to grab them in later rounds while the others are drafting players off their lists that you have no interest in. Move your sleepers up slightly, but you'll be able to grab them a couple of rounds later than you might've thought otherwise while others are drafting the Larry Johnsons of the fantasy world.

Please feel free again to post your fantasy questions to the Fantasy Football Forum on our Discussion Board -- I'll be answering questions weekly in the column before and after the season begins. Next week, I'll be partaking in the traditional Scramble preseason look at Over/Unders for each team, starting with the AFC. Is there a more dastardly curse than the dreaded FOMBC? Find out next week!

Posted by: Bill Barnwell on 15 Aug 2007

24 comments, Last at 27 Aug 2007, 1:45pm by Vivek

Comments

1
by SuupaJim (not verified) :: Wed, 08/15/2007 - 9:07pm

I'm famous!

2
by Zac (not verified) :: Thu, 08/16/2007 - 12:13am

In autopick Yahoo! leagues, it actually works best if you totally change around the order. You have to know this: Yahoo! won't draft a backup (no matter how highly you have him on your list) until all your starters have been taken.

So the best way to set up your team is as follows: Put the top 24 RBs in order (from KUBIAK or whatever) then the top 24 WRs, then the top 12 QBs, then do backup [RBs WRs and QBs], then the top 12 TEs, then the top 12 K, then the top 12 DEFs. This way, your picks will be like this: RB, RB, WR, WR, QB, TE, K, DEF. If any of your starter quality players (Top24/12) are still available, it will take them as the highest backups in round 9. Notice no one gets to take backups before round 9, so that's ok. Then it will start to take the highest rated backups left (the section marked with []). You'll end up with a team with much better RB and WR depth than those that rely on the auto order.

Just make sure that the players you rank are all players that you'd be willing to have on your team.

3
by Bobman (not verified) :: Thu, 08/16/2007 - 12:49am

Next up, Jason Beattie drowns pit bulls who don't win their fights... and frames a certain QB for doing it. He's trouble, that one, just like the little devil on Tom Hulce's shoulder in Animal House.

4
by Larry (not verified) :: Thu, 08/16/2007 - 6:11am

I don't think the advice given is right for a 2 QB league. The 37th best QB is not like the 37th best RB. Pretty much only 32 QBs can be counted on to score ANY points in a given week (a couple more might in mop-up duty or because of injury, but you can't predict which ones). Backup RBs are going to get a few carries, and a number of teams get useful production out of two RBs. So, the scarcity is higher in QBs than RBs, and the VBD calculation Bill gives doesn't seem to account for this. Realistically, there will be NO available QB free agents. I don't think VBD can account for this very easily.

5
by James, London (not verified) :: Thu, 08/16/2007 - 7:27am

Bill, can't you apologise to CMG? Buy her flowers or something?

And the Cartoon is awesome. Just awesome.

6
by Stringer Bell (not verified) :: Thu, 08/16/2007 - 9:58am

do you have the DVOA numbers for each team when they play at home vs. away for 2006? for last 5 yrs?

7
by Yuri (not verified) :: Thu, 08/16/2007 - 10:43am

Am I the only one who'd vote for a "co-pilot" on the Scramble picks? Even someone clueless would do--even if we lose the 2 voices in the dialog, we should at least have 2 sets of picks in the friendly competition. Look how much more fun Bill Simmons' column has become after we see that his wife is beating him.

8
by James, London (not verified) :: Thu, 08/16/2007 - 11:12am

#7

I vote for Raiderjoe. Or Harris. Then we could have the "Raspberry Jam Lock of the Week"

9
by B (not verified) :: Thu, 08/16/2007 - 11:25am

I see that Jason has mastered the classic Gil Thorp technique of making your voice louder by covering your ears when yelling. Maybe later he can make his voice come out of an inanimate object or turn his hand into an alien claw.

10
by SuperBears (not verified) :: Thu, 08/16/2007 - 12:22pm

I am also in a 2 QB league (for the last 7 years) and I am starting to realize the importance of drafting QBs early and often. First of all, we have 12 owners so its impossible for everyone to have 3 starters. Second, an injury to a QB is just a hit to the jugular. The problem is that its only possible to have 32 QBs score every week, unless you can predict what back-up will play because of injury. So its either points or zero. Most back-up RBs will get at least a few carries. This year I plan on have 2 QBs by the first 3 rounds and 3 QBS by round 7. Yes, I realize how stupid the scoring system is.

11
by TED F!@#ING GINN!? (not verified) :: Thu, 08/16/2007 - 1:32pm

Re 8:

I vote for Raiderjoe too.

In the comments of one of the recent articles, it was asked who would be doing Scramble with Bill this year. His answer was "It's a surprise!". I had completely convinced myself that it was going to be Raiderjoe.

Man that would be so freakin' sweet.

I can see it now: "Who should I start @ QB in a 2 QB league this week?"

"Start J. McCown (very athletic qb)and J. Russell (next J. Elway). 1 - 2 punch is unstoppable against SD defense. J McCown line up at WR J. Russell throw WR option screen then go out on route, catch 40 yd TD pass, both get points for yds., 2 td pass, 2 td catch all 1 play. Win week right there."

12
by Alex (not verified) :: Thu, 08/16/2007 - 2:19pm

8,11:
I also vote for Raiderjoe!

13
by Ape Man (not verified) :: Thu, 08/16/2007 - 4:14pm

This is a good example of why I have never understood the 2-QB league. I have seen leagues where you start 2 QB's but only the one who scores better counts - that makes a little sense, I guess, but it still doesn't seem to be an improvement over the old "one QB starter per team" system.

I play in a league that has mostly standard scoring, 1 QB, 1 RB, 2 WR, 1 Flex, TE, K, DST and IDP, and I've never felt the system was an impediment to the league's competitiveness. The more basic systems have always seemed to me to be the most fun, with the least absurdity creeping in (absurdity like 50% of the teams in the league having no legit backup QB.)

APS

14
by Kenneth (not verified) :: Thu, 08/16/2007 - 6:12pm

In all the 2QB leagues I've played in, QBs have been super important. I mean, in most weeks, there are probably only 20 QBs that put a useful score. In a 1QB (12-team) league, this means you have lots of options at QB; in a 2QB league, it means you have a good chance of being deficient at the position. So I always take at least one QB early in those leagues, usually first or second round.

Quick gloating: Best 2QB season I had was in 2004. I had the last pick of the first round (so also first in second), and it was the first year of the league. None of the numbnuts ahead of me realized the importance of QBs (started 3WRs, 3RBs too), so none were taken. So I took Manning and Culpepper with my first two picks. I don't think I had to change my roster the whole season. :)

15
by The McNabb Bowl Game Anomaly (not verified) :: Thu, 08/16/2007 - 7:33pm

I'm SJM.

Re 13:
I didn't make the rules, I'm in the league because it's with a bunch of my friends. They've had the system in place for years and like it. I'm also in a more traditional league, but I thought this was an interesting question.

Re 4, 10, 14:
That's what I suspected, that QBs become paramount because of their scarcity. Bill apparently has never played in a 2 QB league. What I was really asking was this: Do I take the QBs early because I know I need them, or do I assume that the other players are still stuck in "RB first" mode? I suspect the former. I'm drafting 9th out of 12, so I should have a good idea of how things are going by my pick, but it means that I have no shot at either a top RB or Manning.

16
by The McNabb Bowl Game Anomaly (not verified) :: Thu, 08/16/2007 - 7:35pm

PS,
I wish Bill had analyzed the top 48 players rather than 72, since I was mainly curious about my strategy for the first 4 rounds.

17
by mattman (not verified) :: Thu, 08/16/2007 - 7:40pm

I have an alternate strategy for exploiting Yahoo autodrafts. As Zac mentioned above, they don't draft backups until all the starter positions are filled. This means two things. One, your RB depth is going to suck no matter what. Two, it means you are GUARANTEED two top 24 running backs even if you don't take one until the 7th round. The best way to use this glitch and to make up for the unavoidably mediocre RB depth is to target top players at every position except running back. You could rig the rankings so you'll take, say, Peyton in the 1st, Gates in the 2nd, and Chi/Bal D in the 5th, and WR in the 3rd, 4th, and 6th rounds. You'll still wind up with something serviceable like Jacobs/Cadillac at RB and decent WR, and you've got the best player at three different positions. If you're skittish about that RB duo (I would be) you can grab a RB in the 1st then target Gates and Palmer/Brees/McNabb in the 3rd or 4th. Either way, I've found the best way to manipulate Yahoo's autodraft is to take advantage of the fact that decent RBs are still there in the late middle rounds.

18
by Bill Barnwell :: Thu, 08/16/2007 - 11:21pm

It's interesting -- I still disagree with 4/10/14 because it still depends on bench slots and player scarcity is more sketchy. The 32nd-ranked RB in Week 1 last year gained nine points, which is more than the 32nd-ranked QB, but there are 20 more guys who gained between five and nine points at RB.

Essentially, what the argument is is that maybe replacement-level should be lower, which is fine. Again, it depends on how likely teams are to be carrying QBs 30-36 or so. I suspect they wouldn't because they'd rather use the slot on one of those 30-55 RBs who are more likely to get them a few points a week as opposed to a backup QB with the potential for significant performance (e.g. Romo last year).

The weirdest league I played in was actually the first I did, as a kid playing with my father and his friends, where we drafted team quarterbacks as opposed to individual ones.

19
by The McNabb Bowl Game Anomaly (not verified) :: Fri, 08/17/2007 - 1:11am

Re: 18

I can tell you that Leinart was owned last year in this league on potential while Warner was the starter. The bench is actually only 3 spots, but I expect every team will own a 3rd QB even if that means owning Harrington, Culpepper, or Quinn. In fact, Quinn is probably more valuable than guys like Harrington because if he starts, he'll probably be better. The point is that there is no available talent. When calculating value above free agents, the free agent value is virtually 0. I can also tell you that in 2006, total points tapered off more quickly for QBs than RBs according to my league's scoring system, meaning there were fewer valuable QBs than RBs. E.g. 27 RBs generated 140+ points, but only 22 QBs did. 31 RBs produced 120+ points, but only 24 QBs did. 40 RBs earned 100+ points, but only 32 QBs did. Of course this is simplistic because it doesn't account for how many appearances a player made, but the point stands that there is less depth for QBs than RBs.

I suspect that with QB replacement value set at about 0, and only looking at the top 48 picks instead of 72, there will probably be a lot more QBs than RBs.

20
by Bill Barnwell :: Fri, 08/17/2007 - 12:29pm

I am this close to starting the RaiderJoe Tribute Contest.

21
by Ryan (not verified) :: Fri, 08/17/2007 - 6:08pm

Re: Exploiting the Yahoo auto-draft....

Anyone know what would happen if you put every kicker on the do-not-draft list?

22
by mattman (not verified) :: Sat, 08/18/2007 - 7:38pm

Re:21

That's a great plan. I decided to conduct an experiment. I've joined a Yahoo public autodraft league and put every K and DEF on my Do Not Draft List. My goal will be to go Manning/WR/Gates/WR/WR in the first 5 rounds, then theoretically I will snag the next 4 RB on the board (RB #20-24) while everyone else is rounding out their starting lineups. I should be able to wind up with an RB group something like Jacobs/Barber III/Norwood/Cadillac to go with the top QB and TE and at least two upper-tier WR (something like S.Smith/Wayne or Harrison/Fitzgerald) Then once again in the later rounds, I'll be taking players while everyone else is autodrafting backup K and DEF. I'll be sure to report the results.

23
by mattman (not verified) :: Sun, 08/19/2007 - 4:37pm

Research complete: I autodrafted a team with all kickers and defenses on the Do Not Draft List. I didn't get Manning (I had him ranked #2 behind LT for this experiment), who went at #5 one pick before me. Here's the team I wound up with

QB
McNabb (rd 5)
Favre (rd 11)

RB
Portis (rd 6)
Benson (rd 7)
Jacobs (rd 8)
Norwood (rd 13)
C. Taylor (rd 14)
Tatum Bell (rd 15)

WR
S. Smith (rd 1)
Harrison (rd 2)
Chad Johnson (rd 3)
Mark Clayton (rd 9)
D.J. Hackett (rd 10)
Ben Watson (rd 12)

So the Do Not Draft list trick works: I was able to land 3rd round caliber running backs in rounds 6, 7, and 8 while everyone else was drafting TE, DEF, and K. Unfortunately, the autodraft took backups at WR, TE, and QB when there were good RB options like Cadillac, Ahman Green, and Adrian Peterson available. Still, that's a RB group that lacks starpower but is crazy deep, especially when you factor in that my first back was taken in the 6th round. Meanwhile, there's more than enough starpower at WR, TE, and QB. I have a much, much deeper bench than the rest of the teams in the league, I'll have to drop useful players to start a K and DEF each week.

So I think removing K and DEF is the way to go for Yahoo autodrafting. Even if you still want to go RB in the early rounds, you'll still be able to stock up on the best backups before everyone else gets a shot.

24
by Vivek (not verified) :: Mon, 08/27/2007 - 1:45pm

Because of my terrible 2006 GM skills, I have the top pick in two keeper league drafts.

Lynch or Peterson for 2007? I'm going with Lynch because of the presence of Taylor in MN and Peterson's injury history.

In another keeper league, Ronnie Brown is out there too. I'm leaning towards Brown in that league.

Thoughts?