TYA after week 8

With the Vikings' season about to go completely down the tubes (but here's some comedy if you need it), I decided to crunch some numbers to while away my sorrow. Remember this set of posts back in January?

http://www.sportingnews.com/blog/JasonWinter/132522
http://www.sportingnews.com/blog/JasonWinter/132892
http://www.sportingnews.com/blog/JasonWinter/133335
http://www.sportingnews.com/blog/JasonWinter/133744

No? Then go back and read them. Or, if you just want the Cliff's Notes version, they were my attempt to create a new quarterback rating system that took every aspect of a quarterback's play -- including rushing, sacks, and fumbles -- into account. Part of it was my desire to prove to everyone that Jon Kitna was an exceedingly mediocre quarterback, despite his throwing for 4,000 yards every season, because he threw a ton of times and took too many sacks. I was also trying to give Tarvaris Jackson as much credit as I could for his rushing prowess.

But in general, I wanted to create a system that accounted for every possible stat a quarterback could accumulate, because there was no other formula out there that did so. Passer rating completely ignores rushing and sacks, which are as much the offensive line's "fault" as it is the O-line's "credit" when a QB has all day to throw and nails a receiver 50 yards downfield. If one should count for the QB, shouldn't the other count against him? At least a little bit?

After that brief introduction, I won't go too much more into the methodology -- you can read the old posts if you want to learn more. In any case, here are the TYAs for every quarterback with at least 100 pass attempts in 2008:

	Name	TYA
1	Philip Rivers	7.41
2	Chad Pennington	7.40
3	Drew Brees	6.95
4	Tony Romo	6.70
5	Jason Campbell	6.67
6	Donovan McNabb	6.43
7	Aaron Rodgers	6.42
8	Jay Cutler	6.36
9	Eli Manning	6.27
10	Kyle Orton	6.06
11	Jake Delhomme	6.04
12	Kurt Warner	6.01
13	Trent Edwards	5.92
14	Matt Ryan	5.84
15	David Garrard	5.61
16	Jeff Garcia	5.42
17	Matt Schaub	5.38
18	Kerry Collins	5.31
19	Peyton Manning	5.15
20	Dan Orlovsky	4.87
21	Matt Cassel	4.49
22	Gus Frerotte	4.42
23	Marc Bulger	4.32
24	JaMarcus Russell	4.28
25	Ben Roethlisberger	3.98
26	Brett Favre	3.89
27	Matt Hasselbeck	3.84
28	Derek Anderson	3.67
29	Joe Flacco	3.61
30	Carson Palmer	3.51
31	J.T. O'Sullivan	3.51
32	Tyler Thigpen	3.42
33	Brian Griese	3.40
34	Jon Kitna	3.24
35	Ryan Fitzpatrick	1.84

The average for these 35 quarterbacks is 5.25. (Recall that I can't figure a league average without looking at every QB's rushing numbers, which aren't available in handy tabular form.)

Phillip Rivers, who also leads the league in passer rating, just barely edges out Chad Pennington, of all people, as the top performer so far this year. Pennington's only thrown three picks and fumbled once this year, helping him beat out MVP candidate Drew Brees, who's tossed seven picks and coughed it up six times. Meanwhile, the Dan Orlovsky/Jon Kitna debate doesn't look like much of a choice here, though neither answer is good.

Ben Roethlisberger, notorious for taking too many sacks, takes a big hit when those are taken into account. Granted, he's currently the #18 quarterback in passer rating, so maybe being knocked down to #25 here isn't such a Humpty-Dumpty-esque fall. The league leader in sacks allowed, J.T. O'Sullivan (32) is helped by exceptional rushing numbers (29 carries for 146 yards), though he's still awful. Meanwhile, Kerry Collins, who has still remarkably only been sacked once in 171 dropbacks is just barely above average (5.31) but still shockingly better than Peyton Manning (5.15). And you think the Bengals miss Carson Palmer? Imagine how bad Ryan Fitzpatrick would look if he hadn't rushed for 121 yards on 21 carries?

For Vikings fans, Gus Frerotte clocks in solidly below average (4.42), right between Matt Cassel and Marc Bulger. Tarvaris Jackson, with 76 "touches" (passes + sacks + rushes) is virtually the same, at 4.37. There's really no right answer to the Vikings' quarterback situation. Hey, Kerry Collins is a free agent next year! (Yes, I'm kidding.)

A baseball post

Because I haven't done one in a while and figured this would be the last chance I get for one until baseball is completely forgotten in the rush of football season.

Adam Dunn is undeniably a very good hitter, which is something people who can't look past batting average are unable to wrap their heads around. In lieu of his recent trade to the Diamondbacks and his impending free agency, many are saying he'll have to go to the AL to DH because of his fielding, which is oft described using terms that would make your mother blush.

But here's the thing: He's not that bad. In fact, he might actually be a good left fielder.

(Though I couldn't find any picture of him actually fielding or even wearing a glove. I like to think of the image to the right as having the caption, "Don't tell me another one of you idiots think I suck because I hit .247 for my career!")

Revized Zone Rating is a fielding stat that basically measures how often a fielder makes an out on a ball hit into his "zone." Make every play and you're a 1.000 fielder. Make 80% of them and you're .800. Seems simple and accurate enough; after all, it's a fielder's job to make outs. I believe it counts errors into its measure (as "non-outs") and a shortstop throwing to first for the out counts just the same as if he caught a line drive.

So, looking at qualifying left fielders in the majors this year, Dunn ranks fourth out of 12 players with a .899 rating. I think they're a little strict at determining the number of innings you need to qualify, but that's still pretty good. Expand it to all left fielders and Dunn would at least still be in the top half, if not top 10. Click on Dunn's name to pull up his player card and you'll see his LF RZRs the last three years are .878, .826, and .899. In other words, for the past three years, he's been no worse than an average left fielder, according to RZR.

Ah, but look at his earlier years: an abysmal .569 in 2004 and .595 in 2005. That'll get you a rep right quick for being an awful fielder. But if he's improved to being in the .800s now, why do people still think he's awful?

The answer is that nobody really knows how well anyone fields because there are no day-to-day statistical updates or easy, utterly impartial way to compute fielding aptitude. You can tell if a hitter went 1-4 or that he his hitting .325 for the season or has a 3.48 ERA. Those numbers are recorded, double-checked, and easily verified. But could you tell me a hitter's batting average if it wasn't computed for you and if you didn't keep records of it, except in your head? Even if you watched him every day, could you tell me how many home runs Adam Dunn has hit this year? 20? 30? 40? You'd never know.

Fielding is the same way, and, because of that, once we have a notion of how a player is in the field, we tend to think players' fielding aptitude never changes. Unlike virtually every other aspect of baseball -- hitting, pitching, baserunning, and so on -- players are stuck with the label of "bad fielder" or "good fielder" early in their career, and it sticks forever, except occasionally in the case of a good fielder who is very obviously not good any more (like Barry Bonds late in his career). That's why a lot of the same players win Gold Gloves every year. Does the same player win the MVP -- which is largely determined by hitting prowess -- every year? If not, why should the same player be considered the best fielder every season? Look at the bottom of the AL center field RZR ratings. Bet that guy will get some Gold Glove votes this year (and maybe win it outright)?

I'll grant you that fielding stats are an inexact science, but if you look at the objective numbers pretty much anywhere for Adam Dunn (take a look at his RF9 compared to league average RF9 here), you'll find that he's, at the very least, not an embarrassment to the position and a smart team wouldn't immediately try to shoe-horn him into a DH position.

Which, of course, means the Twins would. If they wanted to sign him. Which they, in all likelihood, won't because he hits OMG .247. Oh, and he probably lacks team leadership and is poor at fundamentals. Who wants 40 HR a year from a guy who can't lay down a bunt?

Sunday statistics

We all know the Vikings heavily favored the run over the pass in 2007. With 2,634 yards on the ground and 2,745 through the air (which includes negative sack yardage), the team almost achieved something nearly unheard of in the modern-day, pass-happy NFL: more team rushing yards than passing yards. It was, in fact, a very real possibility much of the season, until the second-to-last game against Washington, in which the team threw for 212 yards while rushing for just 87.

So how rare is it these days for a team to run for more yards than it throws? And where do the 2007 Vikings rank? Here are the "bottom" 10 teams in "pass yardage minus rush yardage" over the last 10 years; in other words, a team that passes for 3,700 and runs for 1,500 would have a "pass minus rush" of 2,200:

Year	Team	PassYds	RushYds	Pass-Rush
2006	Atlanta Falcons	2371	2939	-568
2003	Baltimore Ravens	2255	2674	-419
2000	Cincinnati Bengals	1946	2314	-368
2004	Atlanta Falcons	2412	2672	-260
2005	Chicago Bears	2002	2099	-97
2001	Dallas Cowboys	2218	2184	34
2007	Minnesota Vikings	2745	2634	111
2005	Atlanta Falcons	2679	2546	133
2005	San Francisco 49ers	1898	1689	209
2004	Baltimore Ravens	2312	2063	249

There you have it. Out of 315 teams over the span from 1998 to 2007, only five (1.6%) have rushed for more yards than they have passed; the 2007 Vikings come in 7th on the list. Atlanta's presence on the top 10 list three times is obviously due to the presence of Michael Vick. Baltimore makes it twice thanks to Jamal Lewis and a typically inept passing attack. And look at the yardage totals of those 2005 49ers.

In fact, with the possible exception of Vick, the quarterbacks who led these teams reads like a horror show: Kyle Boller, Akili Smith, Kyle Orton, Quincy Carter, and Alex Smith. Poor Anthony Wright was a backup with both the 2001 Cowboys and 2003 Ravens.

And oh yeah, there's Tarvaris Jackson. For what it's worth, though, the Vikings have the highest passing-yardage total on this list of infamy, and you have to go down to the #14 team, the 2001 Pittsburgh Steelers, to find a higher number in the passing column.

For completeness's sake, here are the top 11 (you'll know why in a minute) teams in pass minus rush since 1998:

Year	Team	PassYds	RushYds	Pass-Rush
2000	St. Louis Rams	5232	1843	3389
2005	Arizona Cardinals	4437	1138	3299
2007	New England Patriots	4731	1849	2882
2007	New Orleans Saints	4314	1466	2848
2004	Indianapolis Colts	4623	1852	2771
2002	St. Louis Rams	4154	1405	2749
1999	Chicago Bears	4136	1387	2749
2006	New Orleans Saints	4503	1761	2742
2007	Green Bay Packers	4334	1597	2737
2002	Oakland Raiders	4475	1762	2713
2004	Minnesota Vikings	4516	1823	2693

Had to squeeze Daunte Culpepper's 2004 Vikings on the list! And yes, Virginia, the Bears did throw for 4,136 yards one year, though it took nearly equal contributions from three quarterbacks (the immortal Shane Matthews, Cade McNown, and Jim Miller), who threw for 1,645, 1,465, and 1,242 yards, respectively.

See, isn't math fun?

The Minnesota Twins are clutch!

Yeah, yeah, I don't believe in clutch hitting, either. But still...the Twins are 9th in the AL (out of 14 teams) in OBP, 10th in SLG% and yet they're fourth in runs scored! Fourth! Yeah, they've stolen a lot of bases (42, 4th) at a good success rate (75%), but that's still less than one SB per game. And their batting average is third in the league, which is testament largely to their "hack first and slap it to the other field" mentality (12th in walks, 13th in HR) than anything else. So, what gives?

The answers can be found here and here.

The first link takes you to the batting splits for the AL this year. Do a search for "RISP" (that's Runners in Scoring Position, FYI, which means runners on second and/or third) to go down to the "Bases Occupied" portion of the link. You'll see that the AL, as a whole, posts the following numbers (Avg/OBP/SLG):

Bases empty: .250/.314/.385
Men on: .271/.347/.412
RISP: .266/.354/.407

That's to be expected. Teams hit better with runners on base, the argument being that pitchers are working out of the stretch instead of the windup. If they were better out of the stretch, they'd use it all the time (which brings me to the question of why pitchers work out of the windup in obvious non-steal situations, like bases loaded or a very slow runner on first).

Now, look at the second link. That takes you to the Twins' batting splits for 2008, where you find the following numbers:

Bases empty: .255/.302/.366
Men on: .282/.351/.399
RISP: .311/.387/.458

Like, whoah. The OPS of the league with runners in scoring position is .761, but the Twins are hitting a robust .845. This is one case where the Twins' "swing first" philosophy might actually be paying off. Walks and strikeouts don't typically score runners from second or third, but singles and high batting averages do, although the team does have 50 unintentional walks in 555 PA with RISP.

On the other side of the coin, you have the Atlanta Braves. A friend of mine, a Braves fan, has been at a loss to understand why the team, which has outscored its opponents 258 to 204, is only three games over .500. The Braves are 2nd in batting average in the NL (out of 16), 3rd in OBP, and 5th in SLG%, but only 8th in runs scored.

Here are the NL batting splits for 2008:

Bases empty: .256/.321/.408
Men on: .265/.347/.413
RISP: .260/.354/.398

And for the Braves:

Bases empty: .289/.351/.434 (good)
Men on: .267/.351/.409
RISP: .265/.352/.393

In truth, the Braves are about even with the rest of the league when it comes to averages with men on base overall and RISP. But look at those numbers with the bases empty! The league OPS with men on is 31 points higher than with the bases empty. The Braves, on the other hand, bat 25 points lower. If they could hit like that with men on base, they'd look like the '27 Yankees. (Random point: If the Yankees are good in 19 years, will we need to distinguish between the two "'27 Yankees" teams?)

In all likelihood, the Twins will come down from their lofty RISP numbers and the Braves will either improve theirs or their hitting with the bases empty will come down (Chipper Jones notwithstanding). Some might attribute this to "great (or poor) clutch hitting," but I have a hard time believing that an entire team can be this good (or this bad) over 50+ games. A few hitters, over a short period of time? Maybe. But not entire teams for a third of the season.

What to expect from Jared Allen

It's not every day your team acquires the NFL's leader in any significant statistic the previous season. Being very good, such players usually aren't available, and when they are, they command top dollar, as Jared Allen did in his trade-and-sign deal that netted him a potential $72 million over 6 years.

That said, big-money deals for players who sometimes enjoyed career years often go south when the fans of the player's new team turn on him for not performing the same way he did just prior to the signing. That's just the law of averages at work; rarely does a player lead the league in the same category multiple years in a row, whether due to injury, age, ineffectiveness, or just luck (good in his pre-signing year or bad in his post-signing years). Will Viking fans think Allen a bust if he doesn't notch at least 15 sacks in 2008? Or 18? Is it even realistic to think that way?

Jared Allen led the NFL with 15.5 sacks in 2007. That makes him the 22nd player in 26 years to lead the league in that category since sacks became an official stat in 1982. Only Reggie White and Mark Gastineau have ever captured back-to-back sack titles. Chances are, Allen won't do it again, since his previous three years' sack totals are 9.0, 11.0, and 7.5 -- solid numbers all, but not what you're looking for from the league's most expensive defensive end.

On the other hand, Allen is young; he was 25 in 2007 and turned 26 just last month. Only 6 of those 28 sack leaders were 25 or younger, and most still had (or, in the case of Shawne Merriman and Dwight Freeney, should have) several good years ahead of them. Comparing Allen to 27-year-old Derrick Burgess or 30-year-old John Randle probably won't paint us an accurate assessment of his future. In fact, looking at sack leaders alone is probably not a wise choice in general, since Allen wouldn't be any less talented if, say, some other player had notched 16.0 sacks in 2007.

Instead, we'll look at two factors: Allen's sack total and his age. 17 players have racked up 15 or more sacks at or before the age of 25, including Allen. In the chart below, I've tracked their sacks in their "15" season (Year N), along with their sack totals in each of the next three seasons (Years N+1, N+2, and N+3):

Player	Age	Season	Year N	Year N+1	Year N+2	Year N+3
D. Thomas	23	1990	20	13.5	14.5	8
T. Harris	25	1989	19.5	7	3	17
A. Tippett	25	1984	18.5	16.5	9.5	12.5
R. White	25	1986	18	21	18	11
R. Dent	24	1984	17.5	17	11.5	12.5
S. Merriman	22	2006	17	12.5
R. Dent	25	1985	17	11.5	12.5	10.5
S. Rice	25	1999	16.5	7.5	11	15.5
P. Swilling	25	1989	16.5	11	17	10.5
D. Freeney	24	2004	16	11	5.5	3.5
J. Allen	25	2007	15.5
C. Simmons	25	1989	15.5	7.5	13	19
S. Jones	24	1986	15.5	6	7.5	6
L. Marshall	24	1985	15.5	12	8	8
B. Smith	23	1986	15	12	11	13
L. Williams	24	1986	15	8	11	14
M. Merriweather	24	1984	15	4	6	5.5
Average			16.7	11.1	10.6	11.1

Of the 16 non-Allen players who managed 15 or more sacks in a season, all but one -- the great Reggie White -- had fewer sacks the year after their big season. 6 experienced a drop-off of 0.5 to 5.0 sacks, and 9 lost more than 5.0 sacks from their totals. Not very encouraging, all things considered.

On the other hand, even the players who lost more than 5.0 sacks probably wouldn't be considered total "busts," if they'd signed free-agent deals after their big seasons. Derrick Thomas certainly wasn't a bust, by any stretch. Tim Harris had 13.5 and 19.5 sacks in back-to-back years before his plunge, but he still rebounded nicely a few years later. Richard Dent still managed double-digit sacks for 6 of his next 8 seasons. Clyde Simmons and Lee Williams were fine after one-year drop-offs. Only Mike Merriweather and Sean Jones might be considered absolute flukes (though Jones did have a few nice seasons later in his career), and Dwight Freeney still has a chance to cement himself as a great modern pass-rusher.

Even with the flukes and busts taken into account, 15-sack men averaged about 11 sacks a year over their next three seasons. Considering how much the Vikings have lacked a consistent pass rush this decade (with only Kevin Williams and Lance Johnstone notching double-digit sacks), I think most fans should be happy with 10-12 sacks each year from Allen. I know I will. There seems to be little chance of his improving on that total -- of the 46 post-Year N seasons above, only 4 resulted in a higher sack total, and two of those were by Reggie White.

Then again, playing Jon Kitna twice a season can do wonders for one's sack totals.

The third-year wideout "myth"

Wide receivers need at least three years to be good. Running backs can excel right out of the gate. Quarterbacks need at least a year or two -- and often more -- to excel.

These are all known "facts," especially to fantasy football fanatics looking for that great draft pick in August. But how much of it is true? The "third-year wide receiver" opinion has been rebuffed by many in recent years, but it still persists, and why is that? Could there be a kernel of truth in the oft-stated belief that the third year is the year wide receivers "put it all together"? And just how good are first-year backs? And when does a quarterback start to show real, or at least fantasy-caliber, skill?

To answer this, I've gone back to the ever-popular Historical Data Dominator. I then searched for rookie wide receivers from 1978 (when the NFL adopted a 16-game schedule) to 2007 with 1,000 yards receiving, second-year WRs with 1,000 yards, third-year WRs with 1,000 yards, etc. I did the same for running backs, using rushing yards. For quarterbacks, I elected to go with 2,400 yards and 16 TDs, an average of 150 yards and 1 TD per game.

Before you think 3,000 passing yards would be a better total, note that even in the pass-happy 2007, the average team threw for 3,652 yards and ran for 1,775, a ratio of just over two to one. If anything, the yardage threshold should be lower, or the rush/receive threshold should be 1,200 yards, so as to be more in line with the 2,400 passing yards. I'll get to that later.

It's not a perfect comparison, but 1,000 yards rushing/receiving is generally considered to make a player "good," while a young QB who throws for 2,400 yards and 16 TDs is also considered to be relatively competent (provided he keeps his interceptions down). Here are the number of players, from 1978 to 2007, who meet these requirements:

Year	RB	WR	QB
1	46	10	4
2	52	35	35
3	58	52	40
4	57	57	45
5	55	57	44
6	43	50	42
7	29	45	39
8	27	36	37
9	15	32	29
10	8	24	31

Without a doubt, third-year wide receivers, as a whole, significantly out-perform their second-year counterparts. However, it should be noted that the difference between third-year WRs and second-year WRs (17) is less than the difference between second-year WRs and rookie WRs (25). So, yes, those third-year guys are might excel, but don't be afraid to take a flier on a good-looking second-year wideout in your draft.

As for running backs, the common wisdom -- that first-year running backs are perfectly valid draft picks -- also seems to hold true, though, as with wide receivers, their peak years seem to be seasons three through five. There's a significant drop-off after year six though, and another precipitous drop after year eight. Again, no surprise there; running backs don't have the greatest shelf life. (LaDainian Tomlinson, it should be noted, will be entering his eighth season in 2008.)

You probably don't need to know that drafting rookie quarterbacks is a risky proposition, and this chart supports that. After that rookie year, though, you get nearly the same number of 2,400-yard, 16-TD quarterbacks every season through year 10. That's probably due in large part to teams not being willing to give rookie QBs playing time (even though it might not be a bad idea, long term) and throwing them into the fire their second or third year.

That said, 1,000 yards isn't that great. That's only 62.5 yards per game. These days, 1,200 yards is probably a better indicator of stardom, or at least a good fantasy player. Here are the number of RBs and WRs who managed 1,200 yards in seasons from 1978 to 2007:

Year	RB	WR
1	19	2
2	28	17
3	29	20
4	34	25
5	38	24
6	26	24
7	15	22
8	12	10
9	10	7
10	6	8

Those are some notably different results. For this level of production, first-year running backs are a relatively poor choice. There was only a difference of 6 (52 to 46) between first- and second-year RBs getting 1,000 yards, but the difference for 1,200 yards is 9 (28 to 19). And wide receivers? Only Randy Moss and Anquan Boldin have managed 1,200 yards as a rookie since 1978. (Bill Groman also did it in 1960.) But there's very little difference from season two on (until season seven, at least). I ran a similar study with QBs, using 3,000 yards and 24 TDs as a benchmark, but it didn't yield any notable results.

So what does it all mean? Yes, third-year wide receivers are actually a pretty good choice, but don't overlook promising second-year players, especially if you're looking for big (1,200+ yard) production. While you can get decent production from a first-year running back, you're not all that likely to find a #1-caliber guy among rookie runners (Adrian Peterson notwithstanding). And don't even think about drafting Matt Ryan, Joe Flacco, Brian Brohm, or John David Booty. Not this year, at least.

Comparing QBs across eras

Ever since I started this blog, I've challenged customary statistical measures and the notion that athletes "back in the old days" were better than they are now. (My second post, in fact, is along those lines.) Sure, Tom Brady had them fancy-schmancy 50 touchdown passes last year, but it took a real man like Bob Griese to actually quarterback his team to a perfect season -- never mind the facts that Griese threw 97 passes in 1972 (and 11 more in the Super Bowl) and that his team was #1 in rushing and #1 in points allowed that year. Clearly, Griese was the superior quarterback. Back then, guys played both ways, never got too injured to play, and rubbed dirt in their cuts -- and liked it!

So today, I decided to see if I could find a way to rate quarterbacks across generations to see if, really, Tom Brady's 50-touchdown year was "better" than Joe Montana's best season or even how it compared to old-timers like Roger Staubach and Ken Stabler -- both of whom ate nails for breakfast and could kill a man with their pinky fingers.

Directly comparing stats would be pointless. The NFL of the 70s and even early 80s, as most know, was very different, especially in the passing game. In Staubach's best year, he threw only 27 touchdown passes. Peyton Manning has had seven seasons of 27 or more TD passes. 14 of the top 20 seasons in passer rating have come since 1989. While this might mean that the quarterbacks of today are better -- despite all the rules changes that have served to open up the passing game -- it's probably better to compare quarterbacks to their peers rather than the passers of 30 years before or 30 years after their time. The NFL of 2007 is practically a different league than the one of 1977, or even 1987.

The method I'll be using is similar to ERA+, OPS+, and other measures used to compare baseball players across generations. It's basically:

(100*player stat for year X)/(league average stat for year X)

This could be used for any stat, but I'll use the familiar passer rating. What this essentially does is take a player's passer rating for a season and divide it by the league's rating for that season. A player whose rating was 90% the league's rating would have a value of 0.9. Then the result is multiplied by 100, just for ease of viewing. The above-stated player would have a PR+ (passer rating +) of 90. A league-average player would have a PR+ of 100. If the league's passer rating was 80 and a player had a passer rating of 120, his rating would be (120/80) = 1.5 * 100 = 150.

By comparing a player to his peers within a season, we should get a better idea of who truly was excellent in any given season. There are obviously other factors that go into a QB's statistical success, most notably his teammates, but that's impossible to factor out. Following is a list of all qualifying QBs since the merger I was able to find with PR+ of 150 or better, meaning they were about 50% (or more) better than the "average" QB of their day:

Player	Year	PR+
Staubach	1971	176.7
Stabler	1976	162.6
Jones	1976	161.2
Anderson	1974	155.9
Stabler	1974	154.6
Montana	1989	153.3
Staubach	1973	153.3
Griese	1971	153.3
Griese	1977	151.9
Tarkenton	1973	151.1
Staubach	1977	150.5
Brodie	1970	150.1

Well, that's something. With the exception of Montana's 1989, all the best seasons come from the 1970s. Staubach's 1971 was phenomenal. His passer rating was 104.8, compared to a league average of 59.3. By comparison, the actual best single-season passer rating, Peyton Manning's 121.1 in 2004, would have had to be 143.0 to achieve a 176.7 PR+! With the best possible rating capped at 158.3, this is virtually impossible.

Ah, but there is a caveat. Staubach threw only 211 passes in that 1971 campaign, a relatively small sample size. (This also raises the question of whether a 30-year-old guy who missed 25% of the season and threw only 325 passes is worth $60 million.) That was still good enough to rank 20th in the 26-team NFL, so he wasn't a complete fluke, but it still seems like an awfully small sample size.

Even with Staubach out of the picture, it's clear that the top of the list is dominated by 1970s quarterbacks. But here lies our second caveat: Because passer ratings were so low in the 70s, there was a lot of room for a good quarterback to excel and post a high PR+. If, for example, the average league passer rating was 110 (you know, the NFL merges with the Arena League), it would be statistically impossible for a QB to post a 150 PR+ (which would be a rating of 165) because of the cap at 158.3.

However, even taking that into account, no QB has ever approached 158.3 for a season. Even Manning's and Brady's awesome 2004 and 2007 seasons don't come close, and the 158.3 is an artificial cap rather than a natural one. For example, it's impossible for a running back to average better than 99 yards per carry, no matter how unlikely it might be. The cap on passer rating would be akin to the NFL saying no running back can go into the record books with a better than 6.0 yards-per-carry average for a season, which both Jim Brown and Barry Sanders have bested. Still, 6.0 is a practical limit, unlikely to be bested in 99.99% of seasons, as the 158.3 barrier will likely never be broken.

Still, the relatively low passer ratings of the 70s did give Staubach, Stabler, et al, more of a chance to look good relative to their peers. For the record, here are the top passer ratings, and, by extension, PR+, for each year from 1970 to 2007:

Player	Year	PR+
Staubach	1971	176.7
Stabler	1976	162.6
Anderson	1974	155.9
Montana	1989	153.3
Staubach	1973	153.3
Griese	1977	151.9
Brodie	1970	150.1
Manning	2004	149.7
Anderson	1975	149.5
Marino	1984	148.8
Young	1994	147.1
Young	1992	147
Warner	1999	145.4
Brady	2007	144.9
Montana	1987	140.6
Young	1997	139.6
Anderson	1981	139.6
Cunningham	1998	139.1
Esiason	1988	138
Young	1991	137.2
Staubach	1978	136.7
Staubach	1979	136.1
O'Brien	1985	136.1
Young	1993	135.9
Griese	2000	135
Anderson	1982	135
Kelly	1990	134.9
Kilmer	1972	133.5
Bartkowski	1983	133.5
Manning	2005	133.1
Pennington	2002	132.6
Warner	2001	132.4
McNair	2003	131.1
Harbaugh	1995	129.9
Young	1996	129.6
Kramer	1986	129.5
Manning	2006	128.7
Sipe	1980	128.2

* Daunte Culpepper's great 2004 season rates with a PR+ of 137.1, which would put it at #21 on the above list. Both he and Manning "wrecked the curve" that season; if one of them hadn't existed to drive up the league-average rating, the other would rate much higher.

* Billy Kilmer in 1972 had the lowest raw passer rating of a league leader (84.8) but is still better in PR+ than 10 other top seasons, including two of Peyton Manning's and one of Kurt Warner's.

* Both 2004 and 2007 saw the highest league average passer rating, 80.9, contributing to Manning's 2004 at #8 and Brady's 2007 at #14.

* The lowest league-average rating was 1977, which clocked in at 57.8, allowing Bob Griese to come in sixth on this list with an 87.8 rating.

* The bottom of the list is cluttered with guys who had a slightly above-average season when no other QB could manage a breakout year. Yes, Brian Sipe, Steve Bartkowski, Chad Pennington, Tommy Kramer, and Jim Harbaugh each led the league in passer rating one year. Though it should be clear from this chart that they weren't really that good -- the league around them was just mediocre. And for a guy who's only topped 20 touchdowns three times in the pass-happy early 2000s, I've always thought Steve McNair was awfully overrated.

At the least, I think this shows that maybe the best QBs of the 1970s were, if anything, a little underrated today, with the big passing numbers that everyone is throwing around. I think I'd still rather have Peyton Manning or Joe Montana as my QB than Roger Staubach or Kenny Stabler, but it's nice to see that some of those old-timers could still produce great seasons.

An even newer way to evaluate QBs

Even semi-casual fans now understand that the old standby of passer rating has its flaws. While new-age stats like AYA (adjusted yards per attempt) and even my own TYA (total yards/attempt) might be better, it's still an inexact science. And statheads with an eye on the draft all know that just because a player put up big numbers in college doesn't mean that success will translate to the pros.

Fortunately, I've come up with a new system that should not only solve these lingering issues but also provide a great way to evaluate both college and pro talent. Wondering if Matt Ryan or Brian Brohm is the best available quarterback in the draft? My new system shows that they should both have similarly excellent careers, and it can all be boiled down into one stat, which I've abbreviated FLAV, which stands for "First-Last Absolute Value."

FLAV is computed by taking the number of letters in a quarterback's first name and subtracting the number of letters in his last name. If you come up with a negative number, simply multiply by -1 to make the value positive. The lower the number, the better the quarterback will be (with a notable set of exceptions, mentioned below), with a value of 0 -- representing a QB with equal letters in each name -- being the optimal result, though QBs with FLAVs of 1 or 2 often have solid careers.

With this in mind, it's easy to see why Brohm and Ryan, both with an FLAV of 0, should be fine. And this should end the debate, once and for all, as to who the best quarterback of all time was. In my mind, it has to be Brett Favre, probably the best FLAV0R (standing for First-Last Absolute Value of 0 with a Ring) to ever play the game, though an equally solid case could be made for FLAV0R Steve Young. And is it any wonder why Michael Vick wanted to be known as Mike Vick a few years into his NFL career?

Many of today's best quarterbacks, like Tony Romo, Carson Palmer (FLAV 0), Drew Brees, Peyton Manning (FLAV 1), David Garrard, and Jeff Garcia (FLAV 2) enjoy success due to their low FLAVs. In fact, the top of the all-time passer rating list is dominated by low-FLAV players Young, Peyton Manning (FLAV 1), Kurt Warner (FLAV 2), Tom Brady (FLAV 2)...and Ben Roethlisberger? A FLAV of 11!?

That's where the F3 corollary comes in. It seems that quarterbacks with three-letter first names are immune to the FLAV rule. In fact, QBs with three-letter first names have had some of the best careers in the NFL. Just look at the large number of successful Joes -- Montana, Namath, Theismann -- with seemingly bloated FLAVs, as well as greats like Dan Marino and Len Dawson. Combine a low FLAV and a three-letter name and you're destined for stardom, not to mention dating and impregnating supermodels, like Tom Brady has done. Many teams overlook this vital part of the FLAV equation, which explains why nobody gave Eli Manning much of a chance in the playoffs and why Delaware's Joe Flacco will likely slip to the second round of the draft.

By now, though, you're probably already ready to poke a hole in this theory by pointing out one of the greatest QB flops of all time: Ryan Leaf. With a strong arm and, more importantly, a FLAV of 0, he seemed destined for greatness.

Ah, but there's a secret about Leaf that few know and that Leaf was always terrified would be discovered, which may have contributed to his high-strung antics and poor NFL career. Like our 42nd president, Leaf was not born with the same last name he carried into adulthood. In fact, it turns out that he is the distant relative of another NFL player, one whose name spells near-automatic doom for any quarterback unlucky enough to fall from his family tree.

It's true: Ryan Leaf's birth name was really Ryan Houshmandzedah. That's an FLAV of 12, folks. If only the Chargers had done their research.

Finally, this post wouldn't be complete without touching up on the Minnesota Vikings' tumultuous quarterback situation. While Kelly Holcomb (FLAV 2) looked good at times in 2007, even low-FLAV players can age ungracefully. Brooks Bollinger (FLAV 4) is clearly not the answer, and any talk of bringing in Sage Rosenfels (FLAV 4) should stop. Donovan McNabb (FLAV 1) seems a good choice, but where does that leave Tarvaris Jackson (FLAV 1)? With such a low FLAV, what is stopping him from becoming the next great, or at least above-average QB?

I think the answer lies in the name "Tarvaris" itself. Let's face it -- do you know any other people named "Tarvaris"? I don't. The name appears to be completely made up, and, while it satisfies the basics of having a low FLAV, the football gods do not take kindly to those who play fast and loose with its simple traditions. If Tarvaris had a more normal-sounding (and preferably seven-letter) first name, like "Terrell" or "Roberto," he could have been one of the greatest QBs of all time. As it is, the Vikings were clearly duped to trade up for him by his low FLAV, and he may sadly never live up to those lofty expectations.

Hey, do you suppose Milt Plum or Sammy Baugh are still available?