How much do we tend to know after five weeks? Bill Connelly compares five-week data to full-season data to find out if we should be worried about TCU and Baylor.
30 Dec 2009
by Tom Gower and Mike Kurtz
Mike: So, the Pro Bowl team has been announced, and the NFL's plan to distract us all from what is nearly certain to be a disappointing week of football has once again succeeded with flying colors. Every TV show, radio program and Web site will be yakking about the selections for the next week or so, because it plays directly to that holy grail of useless sports yakkery: ratedness. Players you like that other people dislike are always underrated. Players you dislike that everyone else likes are clearly overrated.
Funny how that works out.
Your Scramble writers have decided to have a look at the offensive "skill" positions and take a look, using FO's advanced stats, at how the big names everyone is throwing around have really done this year. As your Scramble writers will note on more than one occasion, FO's stats are of course not perfect, but they do reveal some interesting information.
Mike: That is true, but he only just barely makes it out of small sample size theater.
Mike: The most interesting thing to me is that for quarterbacks, DYAR and DVOA roughly match up. When you're trawling FO stats for diamonds in the rough, a good way to go about it is to find a player with mediocre DYAR and limited touches, but who has excellent DVOA.
Tom: I feel like there's a narrow band, in terms of chances. Nineteen of the top 20 quarterbacks by DYAR have 433 to 573 attempts. It would be like all the running backs having 250 to 330 carries.
Mike: True. There are probably a number of reasons for this; running back committees, more injured running backs than quarterbacks, etc. So, the cream of the quarterbacks don't have the same concerns that sap their running back brethren's touches.
Tom: The guys at the bottom end of the scale are Matt Ryan and McNabb, who each have a backup in the rankings. I think the productivity curve for quarterbacks doesn't change like it does for running backs.
Mike: Por que?
Tom: Running backs tend to get worse as they carry the ball more. Quarterbacks can be just as good at attempts 21-to-25 as they are at 1-to-5.
Mike: Right. In conclusion, quarterbacks are boring.
Tom: Aside from Small Sample Size Theater backups, I don't have many complaints. I don't think Princess has really been the fifth/sixth best quarterback this year, but he's probably put up that level of production, so it's tough to argue with it.
Mike: Amusingly, the question hasn't been how he would fare late in the season, but how his offensive line would, which truly boggles the mind.
Mike: Running backs are a good illustration of my point regarding high-value, low-use players and DVOA/DYAR. You have a grab-bag of good backs who are in committees, endured injury, or came into their own later in the season who are overlooked by the counting stat and celebrated by the rate stat.
Tom: Nos. 1 and 2 are same across the board, Grant is third by DYAR but 12th by DVOA. Steven Jackson is fourth by DYAR and 14th in DVOA, which is completely remarkable considering how incredibly terrible his team is.
Mike: Indeed, but it does help to illustrate how crucial to his team he actually is, both in frequency of use and how relatively low his per-play value is.
Tom: Oh, and Pierre Thomas is fifth in DYAR and has fewer than half as many carries as Jackson.
Mike: Which explains his incredible DVOA.
Tom: I'd also like to point out that maybe the Curse of 370 is really having an effect. Chris Johnson leads the NFL in carries. With 322. I guess Portis only had 325 to lead the league in 2007, but otherwise the leader hadn't been below 350 since 1990. This year, Johnson is the only back over 305 carries.
Mike: While the Curse of 370 has had some effect on that, I think we may be overstating its effect. The rules clearly and overwhelmingly favor passing at this point. I think a large part of the effect we're seeing with running backs is simple realization that the passing game is a superior method to move the ball.
Tom: This year we've seen rushes as 43.58 percent of all plays from scrimmage. Last year that was 44.57 percent, the year before 43.52 percent.
Mike: The top quarterback this year has roughly 3.7 DYAR per attempt, whereas the top running back this year has roughly 1 DYAR per carry.
Tom: That sounds an awful lot like TMQ's "the average NFL play earns 5.2 yards, why don't teams always go for it on fourth-and-5 or less?"
Mike: Let's look at a historical comparison, say, 2001. Kurt Warner led quarterbacks in DYAR, and Priest Holmes led all running backs. This was a relatively good year for offense, and gives us a decade's perspective. Warner had roughly 2.8 DYAR per attempt that year, and Holmes had roughly 1.1 DYAR per carry.
Tom: Rushing plays accounted for 44.02 percent of plays from scrimmage in 2001.
Mike: Well, what that tells us is that the best quarterback in 2009 is getting roughly an extra yard of value per play, after adjustment for defense. Running, in contrast, has stayed roughly the same.
Tom: So, what? NFL teams should be passing 70 percent of the time?
Mike: It demonstrates that what passing they have been doing has been more effective, whereas running has been stagnant. Something has changed to make passing more effective, and it would be natural for coaches to dial up more passes in light of this.
Tom: Interestingly, 2001 to 2009 yards per attempt comparisons support passing becoming more effective; we see an increase from 5.9 net YPA to 6.2. One question I have is first-half play calling. Arguably that's more revealing about what a team wants to do, since the second half, particularly the fourth quarter, is more about the team's situation. The NFL, then, has clearly become a pass-heavy league, I think something like 28 teams pass more than they run in the first half.
Mike: Not only more pass-heavy, but simultaneously pass-effective. The real question is if the league sees this as a positive good, or if there will be some sort of readjustment to even things out some time in the future.
Tom: This year we saw 21.5 points per game, in 2001 there were 20.2 points per game, so a modest improvement. Last year, however, was 22.0 points per game, with more running. It's messy.
Mike: Given the way passing and rushing interact, I'm not sure points per game is a good measure of dominance for one or the other, even after adjusted by percentage. One of the most interesting things this year is that the running back Mr. Dependable is Correll Buckhalter, leading the league in success rate. In a system designed around successes like DVOA, you would assume that the top player by success rate would be high on the list, but Buckhalter is only slightly above-average under both DVOA and DYAR. Buckhalter seems to be succeeding, but only just, and thereby creating less value than other, less dependable backs.
Tom: And a guy who is 33rd of 46 in success rate is the top back by DYAR and second in DVOA.
Tom: Wideouts are another mess. Vincent Jackson has the top DYAR and second by DVOA, with Wes Welker second in DYAR and 13th by DVOA. Sidney Rice is third and third and Reggie Wayne fourth and 12th by DYAR and DVOA, respectively. Four of the top receivers are in a narrow band of 104 to 115 targets, then Robert Meachem at eighth by DYAR on 55 targets as part of Small Sample Size Theater. Fun item: Roddy White is 52nd in DVOA and 39th in DYAR. I'm going to go out on a limb and say that nobody on the FO staff thinks Roddy White isn't one of the top 10 or 15 wideouts.
Mike: The problem with wide receivers is that they have a much higher variance. What really interests me is that the top names aren't, past a certain point, the top producers.
Tom: You mean they're not the top-ranked by DVOA and/or DYAR. Steve Smith (The Rules and Regulations) seems to be a perennial example of this. He's either in the top five in DYAR and DVOA or 52nd in DYAR and 58th in DVOA like he is this year.
Mike: Careful, the Steve Smith Defense Force may come and defend his honor against the mean, cheap statistics.
Tom: Well, either his ability and/or level of performance fluctuates wildly from year to year, or DVOA and DYAR are inconsistent guides to player ability.
Mike: Smith seems like an interesting case because he has been getting the ball fed to him, no matter what happened. Whether he was having an off day, the quarterback was having an off day, either of them was having a good day; they were playing a good corner, or a bad corner, or whatever coverage; if he was challenged to a martial arts fast-food delivery contest during halftime, whatever.
Mike: Well, the theory is that it essentially turns him into a barometer for how well the Carolina offense is running, in a general fashion.
Tom: That's why we have offensive and passing offense team DVOA.
Tom: Carolina was 10th in team offense DVOA in 2005, when Smith was the top wide receiver in both DVOA and DYAR. I think the basic point is that we don't think Smith was either significantly better or worse, whether he's 50th in DYAR and DVOA or in the top five.
Mike: I agree, but Smith might be somewhat special because he pretty much is their passing offense, which means he gets a lot of garbage from the quarterback.
Tom: Would you say the same for Roddy White? Last year, White was third by DYAR, sixth in DVOA and is now ranked 39th and 52nd, respectively. Atlanta was fourth in team passing offense last year and is 14th this year.
Mike: Maybe what we're seeing is that individual No. 1 wide receivers are to some extent ciphers for their teams' passing offense, except more extreme?
Tom: So does that mean we should discount No. 1 wide receivers' ratings in terms of evaluating their play, and focus on other guys?
Mike: Honestly, I'm not sure what it means, but it does highlight the need for non-statistical evaluation, even when you have tools as good as DYAR and DVOA.
Tom: That makes two of us.
Mike: As you said, nobody will argue that White is non-top 10 receiver.
Tom: Tight ends are basically just oversized third wideouts.
Tom: I entered Monday night's game with both of my games (one championship and one consolation) in the balance. In the consolation, I was down 86.6-81.7 with Bernard Berrian and Robbie Gould going against Minnesota's defense. The Bears' unexpected offensive explosion and field goals meant that was a non-close win to the tune of 20 points. In the championship game, I was up by fewer than 2 points going against Minnesota's defense. With 1 point for sacks and 2 for turnovers and blocked kicks, in addition to varying point totals for points allowed, any small slip-ups were harbingers of doom.
Mike: I can almost hear "Flight of the Valkyries" in the background.
Tom: The first half was great for me, but then Cutler was sacked and when he was intercepted in the third quarter I thought I was sunk. There was only one way to win: The Bears had to score at least 35 points. I was worried sick when the Bears got to field-goal range in overtime, and Gould lined up for a game-winner. I was joyously happy when he missed it, and then, of course, the game-winning score. I started singing "Bear Down, Chicago Bears, blah blah I forget the rest of the song."
Mike: I'm confused, was this before or after you found out that the hot chick was really your sister?
Tom: In reality, it shouldn't have been close. I made a poor waiver wire choice, dropping Matt Prater and going with Garrett Hartley, but aside from that I had an almost-perfect lineup, while my opponent started the inactive Steven Jackson and had Joseph Addai's 10 points on the bench. That's 17 points I wouldn't have been able to match.
Mike: So the Bears popped your opponent like a balloon with too much air in it!
Tom: I'm claiming properly-vindicated triumph, as I outscored my opponent on the year. I will conveniently ignore that the No. 1 team outscored both of us on the year and would have beaten me this week.
Mike: Like a balloon ... and something bad happening! I lost my consolation game, so I guess I was saved the ignominy of losing in the championships. Or something. Whatever, I have my pony.
Tom: At least until the zombiepocalypse. Anyway, way to go, 1941 Orange Bowl Loser!
Mike: At least one of us won.
(Mike shuffles over to the pony, eying Tom warily.)
Tom: Damn straight. I've now won a championship every year I've played fantasy football. It's either time to retire with my streak intact or start playing in 10 leagues a year.
Mike: It's true. As with snowboarding, there is no middle ground.
At this point I would like to apologize to our dear, dear readers for last week. We promised that your Scramble writers would answer questions in the comments thread, but sadly I was ill for most of last week and just didn't have the energy to work out a dozen or so extra questions. Tom did a bang-up job solo, however, so you can't complain!
by Bill Barnwell
Congratulations to ex-Scramblers Ian Dembsky and Al Bogdan, who led their fantasy team to the championship in this year's FO Fantasy League. They took out regular season champion Elias Holman in the championship game, winning 112-82.
The problem was that no one really had a big game for Elias. His top scorer was Mason Crosby, with 14 points; no one had fewer than five points, but his lineup was full of sevens and 12s. He needed a better day from Pierre Thomas, who scored 12 points before leaving early in the Bucs game. He was hamstrung by an injury to DeAngelo Williams, and left his weekly high scorer -- Sammy Morris, at 15 -- on the bench.
Ian and Al got 21 points from each from Ryan Grant and the 49ers defense, which was the key to victory. Each of their receivers -- Chad Ochocinco, Robert Meachem, and Antonio Gates -- scored a touchdown. They only got nine from Maurice Jones-Drew, and left both Steve Smith (89) and Arian Foster on the bench, but they won comfortably.
The rest of the league's scores:
Bill 103, Will 85 (third-place game)
Aaron 91, Rob 89
Vivek 95, Sean 89
Mike 62, Pat 39
Vince 95, Doug 68
Our final standings after the playoffs:
1. Ian/Al (9-5)
2. Elias (10-4)
3. Bill (9-5)
4. Will (8-6)
5. Rob (7-7)
6. Aaron (8-6)
7. Vivek (7-7)
8. Sean (5-9)
9. Mike (4-10)
10. Pat (7-7)
11. Vince (5-9)
12. Doug (5-9)
Mike: This is my candidate for most insidious advertising campaign of the year, even over the Windows 7/Mac OS X pissing match. These commercials (on both sides) are great examples of lies, damned lies, and statistics. Especially statistics in graphical form.
Tom: How so?
Mike: You have two companies essentially dueling with vaguely-defined data sets they can manipulate every which way because in the end all they have to do is show how much of a map they have shaded in. The dispute is over "coverage." Of course, "coverage" has a lot of meanings. Does it mean you get a signal? Does it mean you get a consistent signal? What percentage uptime is "consistent?"
Tom: This isn't even coverage, it's just 3G coverage.
Mike: Right, that an entirely different but also important point. 3G is fast. It is nice. Verizon has attempted to define the entire argument about cell phones through this one aspect of phone usage, which is possibly advantageous to them but to a degree moot, since most nice phones (the ones you have if you're even caring about 3G coverage) can work fine on 2G networks, albeit more slowly. One phone has even received praise specifically for its ability to quickly shift between 2G and 3G networks.
Tom: That's probably because that's the favorite segment of the high-market non-business customers.
Mike: A lot of it is the tail wagging the dog. We see this so often with high-end electronics. As we discussed earlier, a lot of this stuff is status. Even going by Verizon's map, AT&T still covers pretty much all the places where there are more people than cows.
Tom: That's a minor exaggeration, but yes, they do cover most major metropolitan areas.
Mike: Right. But instead of the discussion being whether the service reaches you, the consumer, it is about which network as a whole is objectively "better."
Tom: If I can't get super-premium coverage in eastern Montana, I don't want your worthless phone! Forget that I haven't been in eastern Montana in 18.5 years.
Mike: You might catch Montana Fever!
Tom: Is that like how Basebug might give you Baseball Fever?
Mike: Kind of, although "Baseball Fever" was just a clever rebranding of "Rickets."
Tom: Are you familiar with Basebug?
Tom: Basebug was a Cleveland Indians mascot in the 1970s. If Basebug bit you, you got "Baseball Fever." Which, seeing as it was the Indians, may have been less preferable to rickets.
Tom: When South Park gave us the Island of Misfit Mascots, it could have been populated almost entirely by ill-conceived baseball mascots.
Mike: Baseball is interesting because most of the nicknames came from a time where horrible racism was perfectly acceptable and random words were somehow relevant or interesting. The really old ones are the most bizarre, like the Cleveland Spiders.
Tom: It somehow feels appropriate that the last NFL team to integrate, and that under government pressure, was the Washington Redskins.
Mike: Anyway, the real problem with this commercial and the "let's compare networks at a theoretical level" argument is that they are brutally effective.
Tom: They are? I feel like this is shamefully lacking in deception for a commercial.
Mike: People love maps. Maps are simple to understand, especially cell phone maps, which we all have some experience with. It's a subtle yet powerful method of manipulation, if not deception. I was actually discussing getting a T-Mobile phone the other day, and most of the responses were along the lines of "Verizon has a better network." When I responded that T-Mobile covered absolutely every area I planned on being in for the next five years, nobody budged an inch. Verizon has a better network.
Tom: I'm sorry, why are you talking to other people about this?
Mike: It's an important consideration. A high-end unlocked cell phone is over $500. It's the kind of decision you want to make absolutely sure you don't screw up.
Tom: If you say so. My normal pattern was "ask people what to do, then do whatever I was planning on doing anyway." I realized that, so I started skipping step 1.
Mike: That explains the 15 military-grade trampolines in your backyard.
Tom: I don't have a yard.
Mike: Sure you don't. (Your secret's safe with me!)
Tom: It does, however, explain the 18 books about football I own and haven't read, yet sit squarely in my line of sight.
Mike: Anyway, it's clear how effective selling the idea of a network is, based on AT&T's counter-advertising. We've seen a lot of advertising wars (most of them about beer, strangely enough), but this is one of the few where the rebuttal takes the exact same form as the critique.
Tom: Plus the lawsuit, which was frivolous and stupid and properly promptly dismissed and, I'm sure, insisted upon by a senior executive who was told just that before it was filed.
Mike: Well, of course. Nobody ever listens to their lawyer. Although I will say that isn't even close to the most crazy phone-related lawsuit.
Tom: I don't see anything in there about bad coverage maps.
Mike: No, just good, ol'-fashioned litigatin'.
KEEP CHOPPING WOOD: It's one thing to have a bad player on your team; it's quite another to have a very good, potentially great player with a costly fatal flaw. So, when Ryan Mouton screws up and fumbles a game away, you find another guy to use at punt returner. When Adrian Peterson has yet another fumble, this one in overtime and leading to the game-losing score, Brad Childress has a difficult decision to make.
MIKE MARTZ AWARD: Riddle me this, Jim Caldwell. Players with injuries are either inactive or on a play count; I get that. But healthy players are apparently perfectly fine to play for the first 40 minutes of a game, but playing the last 20 as well is absolutely right out. Truly, it must take a fine mind to parse that distinction.
COLBERT AWARD: Bill Polian doesn't care about your silly records. Bill Polian doesn't care about your silly undefeated season. Bill Polian doesn't care what you think about whether or not players should be rested. Bill Polian doesn't care to listen to all your silly opinions, or your anger. Bill Polian will run his football team the however he damn well pleases.
Kicker: One person who definitely got Polian's message regarding Week 16 and how little it meant was Adam Vinatieri. While the Colts* did put him in position to rack up a field goal, he also missed an extra point, which is Loser League death. -2 points is not a happy return.
Wide Receiver: No drama, no strange events, just a handful of mediocre to good wide receivers putting up mediocre, 1-point games. Bryant Johnson, Antwaan Randle-El, Braylon Edwards, Mike Sims-Walker, Brian Robiskie, Hank Baskett, Bobby Wade and Todd Watkins were all in on the fun.
Running Back: This was a good week for our running back Losers, as the low point was a bad (but not earth-shattering) 3 points. Brian Westbrook simply didn't do much, and Fred Jackson had a workmanlike performance that was marred by a fumble.
Quarterback: If "uninspired" is a good word to describe this week's Loser League results, then its poster boy is Daunte Culpepper. Culpepper threw 12 times and ran twice. He didn't fumble, he didn't throw an interception, but he only gained 16 yards on the ground and 53 through the air. It's like he's just going through the motions of being a Loser en route to 3 points. Your Scramble writers are quite disappointed.
Jere, Portland ME: Yes, I'm in a league that plays through the last week but, no, it's not a playoff system: our league is simply most points wins the season. While I'm in 4th place now, I am in striking distance for 3rd and maybe even 2nd as I made some ground on those ahead of me (if AD Peterson chokes tonight, I will really make headway with the 2nd place guy). While already in a "winning" position, it would prove far more valuable to move up a notch or two in the final standings. Onto the question ...
My main QB is Brady, who will likely sit early against the Texans. My backup is Vince Young and, while I love him against Seattle, the thought is that they will be giving Johnson a ton of carries towards getting him the rushing record and thus impacting Young negatively. Our waiver wire has Josh Freeman vs. ATL and Derek Anderson vs. JAX both available with great matchups against terrible pass defenses - albeit they are questionable QBs at best. Advice please?
Mike: Tom, you know more about Fish than I do, does he have a habit of chasing numbers?
Tom: In the past, I wouldn't have said yes, but he's clearly been pumping Johnson the last nine games. There was no reason to let him carry the ball in the second half of Friday's game, but he got 10 second-half carries. Or 29 times in the Dolphins game -- LenDale White could have gotten two yards into the line the second half, or the drive before the field goal in overtime. Freeman will definitely play, and has been over 200 yards four of the past five games. He put up 250 yards and two touchdowns in the first Tampa Bay-Atlanta game.
Mike: Derek Anderson is some strange thing that is definitely not an NFL quarterback. Even against a godawful pass defense, he is not a good play. If Johnson is going to get his shot at putting up gaudy numbers, then Freeman's the clear play.
Jeff Trudrung: Going into the Championship game with the following players: Aaron Rodgers, Alex Smith, Larry Fitzgerald, Donald Driver, Greg Jennings, Cedric Benson and DeAngelo Williams ... I can play 1 QB and a total of 3 WR/R. Should I sit Rodgers for Smith because of the match-ups? What about Driver, Jennings, Benson and Williams? Thanks.
Tom: Sit Rodgers for Smith not because of the matchup, but because Smith will play. Rodgers will probably not play much.
Mike: I'm also not sure how much time Benson will get, considering he's semi-hurt. The only problem is that leaves you with Driver, who won't get many quality targets without Rodgers. This is why Week 17 is a fantasy disaster.
Tom: Fitzgerald, Driver/Jennings, Hall. Hope Massaquoi or Antonio Bryant is on waivers!
Mike: I'd probably go with Jennings over Driver.
Tom: That's where I'd lean, too, though I'd strongly consider waiver wire pickups. James Jones may have a better game than Jennings.
Mike: I agree on Massaquoi, but it's unlikely to find anyone truly useful on waivers.
Mike: This is Week 17, where Matt Mauck got his lone NFL start!
Ian Dembsky: Had to take some time from my busy schedule drinking Presidente beers at a resort in the Dominican Republic to gloat that I predicted we had the best team before the season started, and lo and behold, Championship!!! It feels nice to stop by and dominate the Outsiders in fantasy football. Scramble forever!
Mike: Scramble Forever looked good at the start of the year, and I believed in them all the way.
Tom: I liked Elias' team a lot. If DeAngelo Williams had Stewart's day, I'm guessing he'd have won instead.
Mike: Possible, Elias' team worked out quite well. Scramble on, compadre! Scramble on.
The fantasy season is over! Giant lizards loose in the streets! There is always postseason fantasy. I suppose. Anyway, feel free to send triumphant e-mails describing your great fantasy successes or agonizing defeats to scramble-at-footballoutsiders.com. Your Scramble writers might even print them!
34 comments, Last at 04 Jan 2010, 5:02pm by Kevin from Philly