SITE NEWS: We are moving all of our site and company news into a single blog for Sports-Reference.com. We'll tag all PFR content, so you can quickly and easily find the content you want.

Also, our existing PFR blog rss feed will be redirected to the new site's feed.

Pro-Football-Reference.com » Sports Reference

For more from Chase and Jason, check out their work at Football Perspective and The Big Lead.

Archive for September, 2006

Get ‘em while they’re hot

Posted by Doug on September 30, 2006

As I've mentioned before, this site pays its bills through a page sponsorship system that lets you put a message on a particular page of your choosing.

When new pages become available, like Reggie Bush's or Scott Linehan's, they are priced through a Dutch auction, which means that the price starts out really high and then drops by a little each day until someone sponsors it (or until it hits some pre-set minimum). So if you're interested in supporting this site by sponsoring the page of a rookie or a rookie coach, keep an eye on them. The 2006 team pages will also be available soon. Here's the list of rookies:

Brad Childress
Gary Kubiak
Scott Linehan
Eric Mangini
Rod Marinelli
Mike McCarthy
Sean Payton

Marques Colston
Greg Jennings
Laurence Maroney
Reggie Bush
Joseph Addai
Mike Bell
Maurice Jones-Drew
Owen Daniels
DeAngelo Williams
Michael Robinson
Jerious Norwood
Vince Young
Chad Jackson
Vernon Davis
Leon Washington
Charlie Whitehurst
Wali Lundy
Santonio Holmes
Anthony Fasano
Joe Klopfenstein
Jerome Harrison
Hank Baskett
Dave Thomas
Lawrence Vickers
LenDale White
Mike Jennings
Brad Smith
Demetrius Williams
Derek Hagan
Bruce Gradkowski
Tony Scheffler
Leonard Pope
Brian Calhoun
Sinorice Moss
Jeff Webb
Kevin Dudley
Quinton Ganther
John Madsen
Jason Rader
Montell Owens
Quinn Sypniewski
Travis Wilson
Brandon Williams
Kevin Youngblood
Sam Hurd
Ferri Diamond
Brandon Marshall
Ahmad Hall
Adam Jennings
Jason Avant
Jamaica Rector
Ethan Kilmer
Boone Stutz
J.D. Runnels
Marcedes Lewis
Ruvell Martin
Devale Ellis

Comments Off | Posted in P-F-R News

Curses

Posted by Doug on September 29, 2006

A couple of posts ago, a commenter named Jacob requested an analysis of just how unlikely all these unlikely Madden cover tradgedies have been. So I did some research. As usual, Wikipedia has a pretty good summary.

It hurt my head a little to track down the details because of the naming conventions of the Madden games. For instance, Shaun Alexander's outstanding play in 2005 landed him a spot on the cover of Madden 2007, which was released in 2006. So, if a player is on the cover of Madden X, then he kicked butt in Year X-2, but should be struck by misfortune in Year X-1. Given this, I think it might be possible that the curse is capable of going backwards in time to rewrite history. Spooky stuff.

Anyhow, for the sake of definiteness, I will list players by their curse year (X-1).


  • Shaun Alexander, 2006 - he was having an unimpressive year through three games, and it now looks like he's going to miss some games with a broken foot.
  • Donovan McNabb, 2005 - he missed 7 games. His numbers were good when he played, but he was cursed with lots of off-field headaches.
  • Ray Lewis, 2004 - he missed one game, but otherwise had a fairly typical year. Numbers down a little, but they do not now look out of place in his career stat table.
  • Michael Vick, 2003 - he was hurt in the preseason and missed most of the year.
  • Marshall Faulk, 2002 - he missed two games (as he had in each of the previous two seasons). His numbers did decline dramatically.
  • Daunte Culpepper, 2001 - he missed five games to injury. When he played, his numbers were down from the previous year, but he was still on pace for almost 3800 passing yards and 28 combined touchdowns.
  • Eddie George, 2000 - he played the full 16 games and had arguably the best season of his career.
  • Barry Sanders, 1999 - he retired unexpectedly.

I am quite sure I am not the first blogger to attempt to debunk The Madden Curse. There's not much to it, really. It's a combination of selective memory, regression to the mean, and random chance.

The selective memory comes into play after the curse is in the public consciousness, or it can be applied in hindsight. Consider this blurb from the previously-cited Wikipedia article on the Madden Curse. On Eddie George:

Although he had the best year of his career, rushing for 1,509 yards, catching 50 passes for 453 yards and scoring 16 total touchdowns, he was cursed by bobbling a pass in the playoffs. The pass was then intercepted, ironically, by Baltimore Ravens linebacker Ray Lewis, who returned the ball for a touchdown.

This is almost too absurd to comment on, but people do cite this when talking about the curse, which proves that you can find anything if you look hard enough. More evidence is found in the Ray Lewis writeup:

It was also Lewis' first season without a single interception, after posting a career-high 6 the previous year.

Ray Lewis is a very versatile linebacker, but when it comes right down to it, the reason everyone fears him is because of his ability to intercept the ball. I mean, the guy has averaged two interceptions per year --- two! --- during his career. For him to go a whole year without one certainly requires explanation. The Marshall Faulk comment says this:

He never broke through the 1,000 yards rushing mark for the rest of his career.

Now that's legitimate Curse material, but things like this are only brought into evidence when they are bad. Daunte Culpepper came back a few years after The Curse to have an unbelievably great season. Where was the curse then? Ask a Curse-o-phile, and he'll look at you funny. The Curse had worn off by then, of course, what are you nuts? But, if Culpepper had not had that great season later on, I guarantee you that Wikipedia article would say, "Culpepper never again returned to form." (I hate to keep picking on the wikipedia article, but given the nature of how it came to be, I think it is relatively safe to assume it does reflect the mentality of those who believe in the Curse) Try this: ask a Curse proponent right now whether Donovan McNabb is still cursed, or whether the curse has expired. He's having a pretty amazing year as of now, so McNabb Curse talk is limited to just the season afterward. But if this year turns sour, it will surely be added to the list of evidence for the Curse.

The point is, lots of thing happen during an NFL season, and even more happen during an NFL player's career. If you pick out just the bad ones and ignore the good ones, it's easy to contruct an impressive list of misfortunes.

And then there is regression to the mean. In order to get on the cover of Madden, you have to turn in one of the best performances of the year. In almost all cases, that requires a bit of luck. The next year, you're still good, but there is no reason to expect the luck to be there. The league's best performers in any category will always decline, as a group. Just for fun, let's check out the LEAGUE'S LEADING RUSHER CURSE. Again, the year listed is the curse year, the year after leading the league in rushing:


  • Shaun Alexander, 2006 - detailed above. He is doubly cursed.
  • Curtis Martin, 2005 - his rushing yardage dropped by almost a thousand, and he lost four games to injury. The Jets went from a playoff team to 4-12.
  • Jamal Lewis, 2004 - his rushing yardage dropped by more than a thousand yards. And he spent six months in the clink. Further, "the Baltimore Ravens also failed to make the playoffs that season (2004), after winning their division the year before." Is that last one pretty cheap? Of course it is, but it came directly from the wikipedia Madden Curse writeup for teammate Ray Lewis who, by any measure, had a better season than Jamal.
  • Ricky Williams, 2003 - his numbers decline drastically and he never returned to form. The following season started his string of retirements and suspensions.
  • Priest Holmes, 2002 - he missed two games due to injury.
  • Edgerrin James, 2001 - he wrecked his knee and missed 10 games. This was the only year of James' Colt career that they had a losing record and the only year they did not make the playoffs.
  • Edgerrin James, 2000 - Uh, the next year, he wrecked his knee and missed 10 games. Also, he probably bobbled a pass.
  • Terrell Davis, 1999 - he blew out his knee in the season's fourth game, and was not having a good season prior to that. The Broncos went in the crapper after winning two straight Super Bowls. Davis never again had anything remotely resembling a good season.

Or what about the LEAGUE LEADER IN RECEIVING TOUCHDOWNS CURSE?

  • Steve Smith, 2006 - missed the season's first two games with an injury. No TDs so far this year.
  • Muhsin Muhammad, 2005 - missed a game due to injury and also saw his numbers essentially cut in half.
  • Randy Moss, 2004 - lost three games to injury and had the only sub-1000-yard season of his career. Still has not returned to form.
  • Terrell Owens, 2003 - his team missed the playoffs for the first time in three years, his numbers declined, he missed a game due to injury, and he was horribly mistreated by his coaching staff. The events of this season set into motion a chain of events that essentially made TO the most hated man in the NFL.

OK, enough pomposity. Let's return to the original question, which is: exactly how unlikely is this string of occurrences? We need to examine two separate issues: (1) decreased performance, and (2) injuries.

First the performance. As I alluded to above, the top players in any category in any year will have a tendency to decline. But how much decline is reasonable, and how much is the result of the Curse? I looked at all top 3 (by fantasy points) quarterbacks and running backs since 1988. The quarterbacks, on average, declined by 3.3 fantasy points per game. The running backs declined by 2.0 points per game. Here are the declines for each of the cursed players (not counting Barry Sanders and Ray Lewis).


Cursed player Cursed Yr FantPtDiff
=======================================
Eddie George 2000 2.4
Daunte Culpepper 2001 -2.8
Marshall Faulk 2002 -9.4
Donovan McNabb 2003 -2.0
Michael Vick 2004 -7.3
Shaun Alexander 2006 -11.8

That's an average decline of about 5 points, when we should be expecting an average decline of 2 or 3 points. Note also that, even though Culpepper and McNabb declined, they declined by less than typical top-producing quarterbacks usually decline. Looked at this way, three of the six players did worse than reasonable expectations, and three did better. Although admittedly, the ones that did worse did a lot worse than the ones who did better did better. Still, compare that with the Leading Rusher Curse


Cursed player Cursed Yr FantPtDiff
=======================================
Shaun Alexander 2006 -11.8
Curtis Martin 2005 -7.8
Jamal Lewis 2004 -6.6
Ricky Williams 2003 -5.7
Priest Holmes 2002 9.3
Edgerrin James 2001 -3.9
Edgerrin James 2000 1.4
Terrell Davis 1999 -13.6

The average decline was about 4 points, when a decline of 2 points would be expected. But six of the eight did worse than reasonable expectations. This seems at least as bad as the Madden Curse.

Now let's consider the injury half of the equation. We can get the relevant data from yesterday's post, and by running the analagous numbers for quarterbacks.


  • Eddie George played 16 games.
  • Culpepper missed five games. My estimate is that an "average" quarterback has about a 20--25% chance of missing five or more games in a season.
  • Faulk missed two games, which we saw yesterday is just about average for a running back. I'd estimate a 30--40% chance of missing at least two games.
  • Vick missed 11 games, which I estimate has probability 10--15%
  • McNabb missed 7 games: about a 20% chance of that.
  • Alexander's fate remains to be determined. The reports I'm hearing right now suggest that he'll miss two to four games. If that's what it ends up being, it is again quite unremarkable for a running back.

Just as a very rough intuitive comparision, I'd suggest that this string of injuries is about as odd as Jeff Wilkins going through a 1-for-6 stretch in which he attempted field goals of 46, 34, 42, 29, 32, and 40 yards. That would be pretty rare, but I wouldn't assume Wilkins was cursed if it happened.

And again, the Leading Rusher Curse appears to be nearly as bad. Madden Cursed players missed an average of 4.7 games, while Leading-Rusher-Cursed backs missed an average of 4.4 games the next year (that's counting Jamal's jail time. I counted Alexander as missing 3 games in each case.)

Curses happen.

11 Comments | Posted in Voodoo and witchcraft

Injury rates for running backs

Posted by Doug on September 28, 2006

Shaun Alexander's injury prompted a thread over at the footballguys messageboard about running back injury rates. This article, which I wrote over five years ago, was referenced.

Whenever something I wrote over five years ago is referenced, I wince. Why? Because I was an idiot five years ago, that's why. It's always embarrassing to realize how much of an idiot you were five years ago, but I have convinced myself that it's better than not thinking you were an idiot five years ago. With rare exceptions, if you look at something you wrote five years ago and do not now see lots and lots of ways to improve it, that's a bad thing. It means your writing is not improving. That's what I tell myself, anyway.

In that old article, I looked at all running backs who played in a certain number of games in Year N, and then measured how many games they played in Year N+1. In order to be able to plausibly assume that all games not played were due to injury, I only included players who met a certain performance benchmark in Year N. That benchmark was 6 fantasy points per game, which I now realize was too low. It's quite possible that some of the games missed in Year N+1 were due to benchings rather than injury.

So I re-ran the study with a more relevant data set. I looked at all top 10 (in terms of fantasy points) running backs who played 16 games in Year N, with Year N ranging from 1988 to 2004. It would be pretty rare for a running back to be flat-out benched so quickly after a top 10 year, so it's relatively safe to assume that any missed games the following year were due to injury. There may be a very rare benching or a suspension or two in the data, but I doubt it is messing with the numbers too much. Since I required 16 games in Year N, we are looking at backs who presumably had no pre-existing conditions entering the season. So the question we are answering here is: if a running back was productive last year and healthy entering training camp, how many games do we expect him to miss due to injury this year?

Answer: 2.4 games. Here is the distribution:


G PCT
==========
16 48.8
15 15.0
14 7.1
13 6.3
12 5.5
11 2.4
10 2.4
9 0.8
8 1.6
7 0.8
6 2.4
5 1.6
4 1.6
3 0.0
2 1.6
1 0.0
0 2.4

That means that 48.8% of the previously-healthy-and-productive runners played 16 games the next year, 15.0% played 15, and so on. The interesting thing is to compare this distribution to that of the running backs who were in the fantasy top 10 and did not play 16 games in Year N.


G PCT
==========
16 50.0
15 12.5
14 10.0
13 5.0
12 7.5
11 5.0
10 5.0
9 0.0
8 0.0
7 0.0
6 0.0
5 2.5
4 0.0
3 0.0
2 2.5
1 0.0
0 0.0

That's an average of about 2 missed games, and a slightly greater percentage of backs playing the full 16. So what we're seeing here is a lack of evidence that past injury history is an indicator of future injuries.

Now, there were 127 backs in the first group and only 40 in the second group, so we're into small sample size territory. Let's move the cutoff to top 20 instead of top 10. This is the next-year's-games distribution for all fantasy top 20 running backs (1988--2004) who played less than 16 games:


G PCT
==========
16 42.1
15 12.1
14 8.4
13 6.5
12 5.6
11 5.6
10 4.7
9 0.9
8 1.9
7 0.9
6 1.9
5 2.8
4 0.9
3 0.9
2 1.9
1 0.9
0 1.9

That's an average of 3 missed games on the dot. So maybe previously-injured running backs do have a higher injury rate, but maybe not. And even if so, it doesn't appear to be much higher.

7 Comments | Posted in General

Willis McGahee vs. New York Jets

Posted by Chase Stuart on September 27, 2006

My brother and I talk (a lot) before every Jets game, but the conversation this weekend was much less agreeable than normal. Here's the basic rundown.

Him: Well, the Jets have to shut down McGahee. He just kills the Jets every time they face him.
Me: Well...I'm not so sure I'd go that far. I'd only be especially concerned about McGahee in light of him killing the Jets in recent games, if a player's past performance against team X is a greater predictor of his future performance against team X, than his past performance against the rest of the league. And I don't believe that to be the case.

Him: (Rolling his eyes, because I talk like this too often). Listen. McGahee has rushed for over 100 yards in each of the last three games against the Jets. I don't care what he does against the rest of the league, he kills the Jets. He owns the Jets, so they have to put forth all their effort to stop him.
Me: Three is a pretty small sample size, you know. McGahee hasn't been very good against the rest of the league, he's only averaging 3.6 YPC, and he wasn't very good last year or this year. Let's not make him into some unstoppable back.

Him: It's irrelevant what he does against everyone else. For whatever reason, he owns the Jets. They need to stop him.
Me: You have to remember, that splits (do) happen by chance. If he's not very good against the rest of the league normally, I'm not going to be overly concerned about a sample size of three games. (Note: I didn't actually provide a link during our conversation.)

Him: Well in those three games he obliterated the Jets, so I think that's cause for concern.
Me: I just don't think he's that good. I wouldn't sweat it. Do you have any explanation for why he's done so well against the Jets? Do you have an explanation for why he's done relatively poorly against the rest of the NFL?

Him: I'm not breaking down the game tape, but it's enough for me to know that he crushes the Jets, we watch it every time. I don't watch the other teams, so I don't know why he doesn't do as well against them. What's your explanation?

Me: He's been fortuitous to have some big games against the Jets. It happens.
Him: Are you saying he's been lucky to dominate the Jets?

Me: I just think he's been fortuitous.
Him: Ok, let's say he rushes for over 100 yards this week, and then again when the Jets and Bills play. Now he's done it five straight times. Is he "fortuitous" then too?

Me: It depends. How many rushing yards does he have this year?
Him: He rushes for 1130 yards, but 230 of them come against the Jets.

Me: Then yes, I'd say he's fortuitous.
Him: Ok, that's all I needed to hear. You're nuts.

Outside of letting you know that my whole family is crazy, I bring this up to highlight a classic debate between objective and subjective analysis. And of course, we know what happened: McGahee set a career high rushing for 150 yards this weekend. In my brother's mind he feels vindicated; I'm far from ready to admit I was wrong. I'd argue that McGahee had only a 25% chance or so of having a big day, and the fact that he did said unlikely event doesn't change much in my eyes. During our dinner conversation I didn't have the chance to conduct the proper statistical analysis, but now I do.

So here's what I did. From 1996-2005, a running back rushed for over 100 yards against the same opponent in three consecutive games on sixty-nine different occasions. Below is the complete list, with the player's name along with the opponent he rushed against, and the year and week that he did it for the third time. For some players, like Shaun Alexander against the Rams, you'll see a player and team combination listed multiple times. Alexander broke the century mark in five straight games against the St. Louis, and is listed three times because he had a trio of three straight 100-yard rushing games.


Player Year Wk Opp
Shaun Alexander 2005 9 crd
Shaun Alexander 2004 10 ram
Shaun Alexander 2005 5 ram
Shaun Alexander 2005 10 ram
Shaun Alexander 2005 14 sfo
Tiki Barber 2004 12 phi
Tiki Barber 2005 11 phi
Tiki Barber 2005 14 phi
Michael Bennett 2002 14 gnb
Jerome Bettis 1997 8 cin
Jerome Bettis 1997 12 cin
Jerome Bettis 2001 9 cle
Jerome Bettis 1998 8 kan
Jerome Bettis 1997 6 rav
Jerome Bettis 1997 11 rav
Jerome Bettis 2004 12 was
Domanick Davis 2005 14 oti
Stephen Davis 2000 17 crd
Stephen Davis 2001 13 crd
Stephen Davis 2001 17 crd
Stephen Davis 2002 1 crd
Terrell Davis 1998 9 cin
Terrell Davis 1998 11 kan
Terrell Davis 1997 2 sea
Terrell Davis 1997 10 sea
Terrell Davis 1998 6 sea
Terrell Davis 1998 17 sea
Terrell Davis 2001 13 sea
Marshall Faulk 2003 13 min
Eddie George 2000 16 cle
Ahman Green 2001 11 det
Priest Holmes 2002 15 den
Priest Holmes 2002 12 sea
Edgerrin James 2003 12 buf
Edgerrin James 2005 10 htx
Edgerrin James 2001 7 kan
Edgerrin James 2000 16 mia
Edgerrin James 2002 2 mia
Rudi Johnson 2005 14 cle
Thomas Jones 2004 17 gnb
Fred Lane 1998 9 nor
Jamal Lewis 2002 10 cin
Jamal Lewis 2002 13 cin
Jamal Lewis 2003 7 cin
Jamal Lewis 2003 14 cin
Jamal Lewis 2004 3 cin
Jamal Lewis 2002 16 cle
Jamal Lewis 2003 2 cle
Jamal Lewis 2003 16 cle
Curtis Martin 1999 6 clt
Deuce McAllister 2004 12 atl
Deuce McAllister 2004 16 atl
Deuce McAllister 2003 8 car
Willis McGahee 2005 17 nyj
Clinton Portis 2005 1 chi
Clinton Portis 2003 14 kan
Clinton Portis 2003 2 sdg
Clinton Portis 2003 11 sdg
Barry Sanders 1997 14 chi
Barry Sanders 1997 12 min
Barry Sanders 1997 16 min
Emmitt Smith 1998 16 phi
Emmitt Smith 1999 5 phi
Emmitt Smith 2001 12 was
Emmitt Smith 2002 13 was
Robert Smith 2000 12 car
Fred Taylor 2002 14 cle
LaDainian Tomlinson 2004 15 cle
Ricky Williams 2003 16 buf

During those trio of games, the runners averaged 132 rushing yards per game. Of those 69 RBs, 51 played a fourth consecutive game during this time-span (from 1996-2005). How did those runners do? Drum roll...

Twenty-five of the 51 rushed for over 100 yards. The average rushing output was 96.5 rushing yards.

At first glance, I wasn't sure what to make of this. On one hand I think it might support my brother's contention: most players don't rush for over 100 yards most weeks, so for half the RBs to do so is pretty strong evidence that there is some truth to a player having his way with a particular team. On the other hand, half the time the RB didn't hit the century mark, and the average and median production levels were also in the two digits.

I finally figured out what (I think) is the correct way to tell if this is "good" or "bad" production from our RBs. For all 51 RBs, I then looked at how many rushing yards per game they averaged in the year they played that fourth consecutive game. The 51 RBs averaged 91.2 rushing yards per game in that year. To sum, of the 51 RBs that have rushed for over 100 yards against the same team in three consecutive games, they averaged 130.0 yards per game during those weeks. When they played that opponent a fourth time, the RBs averaged 96.5 YPG, and during that season the RBs averaged 91.2 YPG. Those RBs also hit the 100 yard mark in 43% of the games they played that season, versus breaking triple digits in 49% of the games against that particular opponent.

So it looks like you might expect a RB that has dominated a team in the past to average about five more yards per game against that particular opponent than against the rest of the league. I'm not sure if this means my brother won or I won, although in the end we're both just happy the Jets won.

In 2006, a few of those streaks were on the line.

  • Shaun Alexander rushed for 89 yards against the Cardinals in the Seahawks opener, a Seattle win.
  • Tiki Barber rushed for over 100 yards against the Eagles in five straight games, but the Birds shut Barber down to the tune of only 51 rushing yards in Philadelphia's home opener, in a Giants win.
  • Rudi Johnson continued his dominance against the Browns in the Bengals' home opener, by rushing for 145 yards and leading Cincinnati to victory.
  • Willis McGahee rushed for 150 yards against the Jets in the Bills' home opener, but the Jets held on to win, 28-20.

One stat you might not have caught this week: the Jets became the first team in NFL history to win a road game while allowing both a 300 yard passer and a 150 yard rusher. (As you might remember, New York needed to make some history the last time the Jets won in Buffalo, when Chad Morton became the first player to ever return a kickoff for a touchdown in regulation and then in overtime.) You might be wondering how that happened, and while I'm sure you'll hear the typical poetic reasoning (the Jets are well-coached, never give up, play with heart, etc.), here's the real formula:

  • Return a fumble for a TD, force two other turnovers and commit none
  • Score 21 points on three red-zone trips
  • Hold the opponent to 13 points on four red-zone trips

8 Comments | Posted in General

Westbrook is good and the Texans are bad

Posted by Doug on September 26, 2006

I had a couple of requests via email yesterday.

One was to provide some historical context for just how bad the Texans' defense has been. They've given up a lot of yards, and they've given up the most points in the league. But if you want to wade into a worst-ever brawl, you'd better be packing more than 98 points allowed in a three-game stretch. Since 1978, NFL teams have played 11,813 three-game stretches (some of which overlap with each other), and 361 have been worse than the Texans' last three games. So they're in the 97th percentile for badness. But as of now they do not seem to be anything special. They're just regular old worst-in-the-league bad.

In case you were curious, the worst defensive three-game stretch of the last 25 years belongs to the Titans, who in weeks 13--15 of 2004 gave up 140 points: 51 to the Colts, 49 to the Chiefs, and 40 to the Raiders. On the other end of the list, the 2000 Titans, the 2000 Steelers, and the 1985 Bears had three-game stretches where they allowed 3 points.

The other request was for an analysis of where Brian Westbrook's Sunday performance ranks historically in terms of fantasy points per touch. If you set the cutoff at 10 touches, then it was the best performance since 2000. Westbrook is no stranger to this list:


Name YR WK RSH YD REC YD TD
===================================================
Brian Westbrook 2006 3 8 117 4 47 3
T.J. Duckett 2004 14 12 65 0 0 4
Charlie Garner 2002 5 8 94 4 83 2
Larry Johnson 2005 1 9 110 1 11 2
Clinton Portis 2003 14 22 218 2 36 5
Marshall Faulk 2000 5 7 55 6 116 2
Jerome Bettis 2005 17 10 41 0 0 3
Ryan Moats 2005 14 11 114 0 0 2
Marshall Faulk 2001 12 12 70 6 128 3
Kevin Faulk 2002 10 5 21 7 109 2
Brian Westbrook 2003 11 9 48 5 60 3
Brian Westbrook 2003 15 8 59 3 45 2
Michael Pittman 2004 9 15 128 2 30 3
Shaun Alexander 2002 4 24 139 3 92 5
Tatum Bell 2005 5 12 127 1 5 2
Randy Jordan 2000 16 11 55 1 55 2
Frank Gore 2005 16 10 68 0 0 2
Tiki Barber 2004 1 9 125 5 75 1
Clinton Portis 2002 15 21 130 3 75 4
Antowain Smith 2000 17 17 147 1 4 3
Corey Dillon 2000 8 22 278 0 0 2
Tiki Barber 2000 1 13 144 3 25 2
Derrick Blaylock 2004 12 8 57 3 21 2
Garrison Hearst 2001 11 12 106 1 6 2
Chris Warren 2000 8 10 64 1 11 2
Antowain Smith 2005 6 12 88 0 0 2
Shaun Alexander 2005 3 22 140 0 0 4
Brian Westbrook 2005 3 13 68 6 140 2

If you lower the cutoff to five touches, there are a few performances that can top Westbrook's. Who can forget Jerome Bettis's 5 carries for 1 yard and 3 TDs in week 1 of 2004?

10 Comments | Posted in Fantasy, General

Slow starters

Posted by Doug on September 25, 2006

Three weeks of football are in the books (well, almost), and nobody seems to have told your first-round pick that the regular season has started. Is it time to cut bait on your slow starters, or can you sit tight and hope they'll get it turned around?

The historical data, as usual, reveals a lot of guys who you should have given up on, and a lot of others who flipped the switch started producing right away. In 2000, Tony Gonzalez had a miserable first three games, compiling less than 100 total yards and no touchdowns. He heated up in week four and never cooled down, finishing the season as the league's top-scoring fantasy tight end.

Marshall Faulk's 2002 paints a different picture. After having had two of the best fantasy seasons in history the previous two years, Faulk started the season with three underwhelming (for him) games. He never stopped the slide, ending the season even less productively than the first three weeks had been.

But are there more Gonzos or more Faulks? Only when we collect all these stories do the trends start to show themselves. I searched the archives for all players since 2000 who (a) finished the previous season in the top 10 in fantasy points per game, (b) played in all of the first three weeks, and (c) had a substantially lower points per game average in the first three weeks than in the previous year (five points per game for QBs, RBs, and WRs, three points for TEs). Then I asked: was their points per game average for the remainder of the season closer to last year's average, or was it closer to their average for the first three disappointing weeks?

The results: 30 of the 40 players rest-of-season averages were closer to the previous year's. So the good news is that most of these guys got their acts together. And a lot of the ones who didn't were obvious: Mike Anderson in 2001 (he lost his job to Clinton Portis) and Muhsin Muhammad (who was Orton-ized) in 2005, for example.

The bad news is that most of them didn't get their acts all the way back together. Only 16 of the 40 had a rest-of-season average that was as good as their previous season's average. Bill James once said (something like): if a guy hits .300 one year, and .250 the next, I'm going to project him to hit .275 in the following year. It's essentially the same story here.

Here is the data:


--Prev Yr- --First 3- -- Rest ---
Player YR RNK PPG RNK PPG RNK PPG
==============================================================
Jon Kitna 2000 8 16.6 15 9.6 13 14.7
Jeff Garcia 2001 2 24.6 8 18.7 1 22.5
Mark Brunell 2001 7 17.7 17 12.3 7 18.3
Jeff Garcia 2002 2 21.8 16 13.1 4 19.6
Drew Bledsoe 2003 5 19.9 13 14.0 16 11.4
Trent Green 2003 8 19.0 14 14.0 5 20.3
Peyton Manning 2003 4 20.4 16 12.9 2 21.8
Trent Green 2004 6 19.1 14 13.4 3 22.2
Trent Green 2005 5 20.6 16 12.0 4 17.8
Peyton Manning 2005 2 26.1 17 11.9 1 19.9

Emmitt Smith 2000 6 15.3 18 7.7 13 12.2
James Stewart 2000 8 13.0 19 6.0 6 15.0
Stephen Davis 2001 9 15.3 15 7.6 8 13.1
Mike Anderson 2001 5 18.3 22 5.4 23 6.2
Marshall Faulk 2002 1 24.3 5 17.0 9 14.4
Shaun Alexander 2002 4 16.4 17 8.7 5 18.9
Curtis Martin 2002 5 15.5 30 4.1 12 13.5
LaDainian Tomlinson 2003 3 19.2 9 13.2 1 23.4
Deuce McAllister 2003 5 18.0 14 11.2 7 17.7
Charlie Garner 2003 9 16.0 15 10.1 21 8.0
LaDainian Tomlinson 2004 3 21.5 7 16.3 1 19.9
Fred Taylor 2004 10 14.8 15 8.9 8 13.5
Michael Pittman 2005 9 14.7 38 2.3 27 6.1

Terrell Owens 2001 1 16.0 15 8.3 1 16.5
Randy Moss 2001 2 14.6 23 6.8 3 12.9
Terrell Owens 2002 2 15.0 22 9.3 1 17.6
Terrell Owens 2003 1 15.8 15 10.2 8 11.1
Marvin Harrison 2003 2 14.9 33 6.5 3 14.0
Peerless Price 2003 7 11.1 50 3.4 26 7.1
Muhsin Muhammad 2005 1 14.9 23 8.6 36 6.1
Marvin Harrison 2005 5 12.6 27 7.3 3 13.7
Reggie Wayne 2005 8 12.0 41 5.5 15 9.1

Tony Gonzalez 2000 1 10.1 7 3.1 1 13.7
Rickey Dudley 2000 3 6.8 9 1.8 4 4.9
Tony Gonzalez 2003 2 7.5 9 2.6 1 11.1
Bubba Franks 2003 4 5.7 12 2.0 10 4.2
Billy Miller 2003 6 5.0 17 1.3 11 4.1
Alge Crumpler 2005 4 8.1 10 4.7 2 8.0
Tony Gonzalez 2005 2 10.5 12 4.1 4 6.9
Jermaine Wiggins 2005 6 6.8 13 3.7 12 4.0

2 Comments | Posted in Fantasy

Game similarity scores

Posted by Doug on September 22, 2006

Back in this post, I took a quick ad hoc look at teams that started the season similarly to this season's Atlanta Falcons. As promised, I decided to do a more systematic analysis of all 32 teams. Here it is.

STEP 1: for every pair of team games since 1978, build a similarity score according to the following formula:

  1. In traditional Bill James fasion, we start with 1000

  2. if one team won and the other lost, subtract 1000.

  3. if one team was on the road and the other was at home, subract 200.

  4. from whatever is left, subtract (Team1Score - Team2Score)^2 + (Team1OppScore - Team2OppScore)^2 + (Team1Margin - Team2Margin)^2

Yes, I know there is some redundancy there, and the formula might be amenable to some algebraic tinkering that would make it look prettier. I basically just messed around until I came up with something that looked generally OK. Why squared? No good reason. Why 200 for home/road? Because it seemed sort of OK. That's about the level of rigor I was using.

STEP 2: For each team in the 2006 NFL, compare them with every other team since 1978 and construct a first-two-weeks similarity score with that team, according to the formula: Similarity = Week1Similarity + Week2Similarity. Why compare week 1 to week 1 and week 2 to week 2, instead of just comparing the first two weeks as a whole? I'm not sure I can defend that.

STEP 3: For each team in the 2006 NFL, look at the twenty most comparable teams and compute their average eventual record, weighted according to how strong the similarity is.

Here are the top twenty comparables to the 2006 Atlanta Falcons:


TM Year Week 1 Week 2 SIM Record
===============================================
atl 2006 R 20- 6 H 14- 3
phi 1981 R 24-10 H 13- 3 1966 10- 6-0
ind 2005 R 24- 7 H 10- 3 1942 14- 2-0
rai 1983 R 20-10 H 20- 6 1914 12- 4-0
tam 2005 R 24-13 H 19- 3 1876 11- 5-0
mia 1998 R 24-15 H 13- 7 1836 10- 6-0
pit 1998 R 20-13 H 17-12 1776 7- 9-0
det 2000 R 14-10 H 15-10 1762 9- 7-0
ram 1978 R 16-14 H 10- 0 1750 12- 4-0
jax 2001 H 21- 3 H 13- 6 1748 6-10-0
kan 1992 R 24-10 H 26- 7 1744 10- 6-0
ram 1986 R 16-10 H 16-13 1736 10- 6-0
atl 1998 R 19-14 H 17-12 1728 14- 2-0
sea 2004 R 21- 7 R 10- 6 1724 9- 7-0
tam 1980 R 17-12 H 10- 9 1722 5-10-1
pit 1978 R 28-17 H 21-10 1708 14- 2-0
buf 1980 H 17- 7 H 20-10 1688 11- 5-0
rai 1987 R 20- 0 H 27- 7 1662 5-10-0
cle 1978 H 24- 7 H 13-10 1660 8- 8-0
jax 2004 R 13-10 H 7- 6 1656 9- 7-0
mia 1979 R 9- 7 H 19-10 1656 10- 6-0

The weighted (and scaled to 16 games) average of those teams' win totals was 9.9 wins, and we might view that as a reasonable over/under for Atlanta's 2006 win total. According to this method, here are the projected win totals for each team this season:


TM Proj Wins
==============
ind 10.1
atl 9.9
chi 9.8
cin 9.8
nor 9.8
bal 9.8
jax 9.6
sea 9.6
sdg 9.6
nwe 9.6
min 9.1
ari 8.9
dal 8.5
pit 8.5
den 8.1
buf 8.0
phi 7.9
nyg 7.7
stl 7.6
nyj 7.5
sfo 7.0
mia 6.4
hou 6.3
kan 6.2
was 5.9
det 5.8
ten 5.6
car 5.6
gnb 5.6
cle 5.4
tam 5.3
oak 5.3

The main takeaway lesson from this exercise is the same as in this post: don't get too excited or too upset about early season results. The Bears have been destroying teams, and this method projects them to win only 7.8 of their remaining 14 games. The Raiders have looked as bad as an NFL team can look, and the method says they'll win 5.3 games.

In some sense, this exercise is just a whole lot of work to get (I'm assuming something very close to) the same results you'd get by running a simple regression of wins versus first-two-weeks record and scoring margin. But I like this method better, because it's not a black box.

You say the Bears should expect to win X games this year. Your friend calls BS: haven't you seen how dominant they've looked? If regression is what you've got, it's tough to give a decent counterargument unless he understands regression. But this method lays the reasoning right out there in a crystal clear way: the 1986 Falcons won their first two games by a combined 41 points and they ended up winning 7 games. The 1994 Seahawks won their first two by scores of 28-7 and 38-9, and they finished at 6-10. That is, of course, the same kind of information that your regression was taking into account, but it's just so much more transparent here.

Also, while I doubt it's actually happening here, this method is theoretically capable of picking out subtle combinations of things that regression wouldn't tell you, because you wouldn't think to ask it. For example, I don't know if this means anything or not, but it's intriguing that the Colts, who have scored a lot of points and also given up a lot, project better than the Chargers, Ravens, and Bears, who have scored a lot and given up almost none.

For those who want to investigate, here are those four teams' comparable lists:


TM Year Week 1 Week 2 SIM Record
===============================================
ind 2006 R 26-21 H 43-24
stl 1985 R 27-24 H 41-27 1948 5-11-0
nyj 1987 R 31-28 H 43-24 1922 6- 9-0
nor 2002 R 26-20 H 35-20 1902 9- 7-0
den 1993 R 26-20 H 34-17 1864 9- 7-0
atl 2004 R 21-19 H 34-17 1828 11- 5-0
den 1998 H 27-21 H 42-23 1796 14- 2-0
jax 1997 R 28-27 H 40-13 1750 11- 5-0
ram 1989 R 31-21 H 31-17 1732 11- 5-0
sfo 1984 R 30-27 H 37-31 1690 15- 1-0
kan 2003 H 27-14 H 41-20 1662 13- 3-0
sfo 1995 R 24-22 H 41-10 1642 11- 5-0
rai 1982 R 23-17 R 38-14 1624 8- 1-0
sea 1985 R 28-24 R 49-35 1604 8- 8-0
det 2004 R 20-16 H 28-16 1600 6-10-0
sea 1988 R 21-14 H 31-10 1578 9- 7-0
was 1978 R 16-14 H 35-30 1546 8- 8-0
dal 1983 R 31-30 R 34-17 1544 12- 4-0
pit 1992 R 29-24 H 27-10 1526 11- 5-0
mia 1990 R 27-24 H 30- 7 1512 12- 4-0
nwe 1999 R 30-28 H 31-28 1510 8- 8-0


TM Year Week 1 Week 2 SIM Record
===============================================
sdg 2006 R 27- 0 H 40- 7
gnb 1996 R 34- 3 H 39-13 1840 13- 3-0
sea 1994 R 28- 7 R 38- 9 1690 6-10-0
gnb 2001 H 28- 6 H 37- 0 1664 12- 4-0
cin 2005 R 27-13 H 37- 8 1636 11- 5-0
atl 1986 R 31-10 H 33-13 1594 7- 8-1
rai 1987 R 20- 0 H 27- 7 1564 5-10-0
buf 1981 H 31- 0 R 35- 3 1526 10- 6-0
phi 1980 H 27- 6 R 42- 7 1520 12- 4-0
den 2003 R 30-10 R 37-13 1516 10- 6-0
sdg 2002 R 34- 6 H 24- 3 1498 8- 8-0
tam 1992 H 23- 7 H 31- 3 1492 5-11-0
sfo 1996 H 27-11 H 34- 0 1472 12- 4-0
sea 1998 R 38- 0 H 33-14 1464 8- 8-0
mia 1981 R 20- 7 H 30-10 1428 11- 4-1
kan 1992 R 24-10 H 26- 7 1330 10- 6-0
sdg 1979 R 33-16 H 30-10 1330 12- 4-0
sea 2003 H 27-10 R 38- 0 1322 10- 6-0
buf 2003 H 31- 0 R 38-17 1320 6-10-0
mia 1996 H 24-10 R 38-10 1284 8- 8-0
mia 1984 R 35-17 H 28- 7 1278 14- 2-0


TM Year Week 1 Week 2 SIM Record
===============================================
bal 2006 R 27- 0 H 28- 6
rai 1987 R 20- 0 H 27- 7 1896 5-10-0
sdg 2002 R 34- 6 H 24- 3 1888 8- 8-0
atl 1986 R 31-10 H 33-13 1770 7- 8-1
gnb 1996 R 34- 3 H 39-13 1740 13- 3-0
kan 1992 R 24-10 H 26- 7 1708 10- 6-0
mia 1981 R 20- 7 H 30-10 1682 11- 4-1
sea 1998 R 38- 0 H 33-14 1660 8- 8-0
sdg 1979 R 33-16 H 30-10 1584 12- 4-0
sdg 1996 H 29- 7 H 27-14 1576 8- 8-0
dal 1981 R 26-10 H 30-17 1572 12- 4-0
mia 1984 R 35-17 H 28- 7 1564 14- 2-0
tam 1992 H 23- 7 H 31- 3 1560 5-11-0
sea 1994 R 28- 7 R 38- 9 1556 6-10-0
ram 1988 R 34- 7 H 17-10 1540 10- 6-0
sea 1984 H 33- 0 H 31-17 1534 12- 4-0
cin 2005 R 27-13 H 37- 8 1528 11- 5-0
den 2003 R 30-10 R 37-13 1508 10- 6-0
rai 1984 R 24-14 H 28- 7 1504 11- 5-0
pit 2005 H 34- 7 R 27- 7 1496 11- 5-0
dal 1995 R 35- 0 H 31-21 1494 12- 4-0


TM Year Week 1 Week 2 SIM Record
===============================================
chi 2006 R 26- 0 H 34- 7
gnb 1996 R 34- 3 H 39-13 1840 13- 3-0
rai 1987 R 20- 0 H 27- 7 1830 5-10-0
atl 1986 R 31-10 H 33-13 1764 7- 8-1
sdg 2002 R 34- 6 H 24- 3 1744 8- 8-0
sea 1994 R 28- 7 R 38- 9 1698 6-10-0
cin 2005 R 27-13 H 37- 8 1672 11- 5-0
mia 1981 R 20- 7 H 30-10 1672 11- 4-1
kan 1992 R 24-10 H 26- 7 1624 10- 6-0
tam 1992 H 23- 7 H 31- 3 1616 5-11-0
sea 1998 R 38- 0 H 33-14 1598 8- 8-0
den 2003 R 30-10 R 37-13 1594 10- 6-0
gnb 2001 H 28- 6 H 37- 0 1586 12- 4-0
sdg 1979 R 33-16 H 30-10 1540 12- 4-0
buf 1981 H 31- 0 R 35- 3 1508 10- 6-0
mia 1984 R 35-17 H 28- 7 1494 14- 2-0
dal 1981 R 26-10 H 30-17 1488 12- 4-0
sfo 1996 H 27-11 H 34- 0 1480 12- 4-0
rai 1984 R 24-14 H 28- 7 1472 11- 5-0
sdg 1996 H 29- 7 H 27-14 1432 8- 8-0
sea 1984 H 33- 0 H 31-17 1424 12- 4-0

The main objection I'd anticipate to this method is that it doesn't account for strength of schedule. The Chargers have only played the Raiders and Titans, who are terrible. The Falcons have played the Panthers and Bucs, who are, well, we don't really know what they are. As JKL pointed out in the comments to a previous post, it's very difficult to put a strength of schedule number on any of these teams right now (except for the teams that have played the Patriots, who have Tom Brady and are therefore known to be great). What do we use? Last year's record? This year's record? Some sort of power ranking scheme applied to the first two weeks? I'm not sure. And the uncertainty is enough to make me want not to waste a lot of effort trying to account for it.

9 Comments | Posted in General, History, Statgeekery

Mile High Hopes

Posted by Chase Stuart on September 21, 2006

On Tuesday, I analyzed the Jets' inept running game by comparing it to other similarly ugly starts. Based on how those teams finished their respective seasons, there wasn't much cause for optimism in New York.

But what about in Denver? Jake Plummer has been a very good QB since joining the Broncos. Plummer's ranked among the top 10 in "value added" in each of the past three years (value added defined as adjusted yards per attempt multiplied by attempts). That ignores his strong rushing skills, so by almost any objective measure, Jake Plummer's been one of the very best quarterbacks in the game since 2003.

That's what makes Plummer's terrible start so surprising. He's currently averaging 2.34 adjusted yards per attempt, well below his impressive averages of 7.09, 6.64 and 6.68 in recent seasons. Seeing Plummer's AY/A in the twos will bring to mind visions of 1999, when Plummer had one of the most damaging seasons to a franchise in history. But I find it hard to imagine he's regressed so much, especially since the talent around him is still very good.

So I'm left with this question: how often is it that a QB has a really bad start to the season but then turns it around and performs well? If it happens often, then Broncos' fans have little to worry about. If it rarely happens, Jay Cutler may be named the starter before October.

I looked at all quarterbacks since 1995 that attempted at least 25 passes in their first two games of the season (not necessarily weeks 1 and 2) and would eventually throw over 300 passes in the season.

Key:
EAtt = Early number of attempts (attempts in games 1 and 2)
EAY/A = Early Adjusted Yards per Attempt
FAY/A = Final (or season ending) AY/A
Value = AY/A * Attempts
FAtt = Final (or season ending) number of attempts
Ratio = (FAY/A^2) / (AY/A) (This isn't that important, but was the metric used to sort the list and create a cutoff point.)


Player Year EAtt EAY/A FAY/A Value FAtt Ratio
Jake Delhomme 2003 43 1.58 5.99 189.68 449 22.68
Steve McNair 2000 49 2.24 6.20 207.89 396 17.15
Jake Plummer 2003 38 2.74 6.68 335.95 302 16.30
Charlie Batch 1998 63 2.94 6.66 262.35 303 15.11
Donovan McNabb 2003 82 2.43 6.03 220.26 478 14.97
Trent Green 2004 66 4.11 7.37 721.45 556 13.22
Steve McNair 2001 51 3.76 7.01 585.04 431 13.04
Brett Favre 2003 69 2.68 5.80 108.66 472 12.53
Tom Brady 2003 72 3.17 6.28 376.50 527 12.46
Kordell Stewart 2001 59 3.36 6.23 256.89 442 11.57
Jake Plummer 2005 85 4.40 7.09 592.78 456 11.41
Peyton Manning 2005 64 5.48 7.90 956.13 453 11.37
Peyton Manning 2003 64 4.75 7.26 956.41 566 11.08
Jeff George 1995 78 4.45 6.98 662.28 557 10.95
Tony Banks 1999 48 3.40 6.08 143.18 320 10.89
Jake Delhomme 2005 57 4.23 6.76 424.27 435 10.81
Jim Harbaugh 1997 67 3.88 6.41 215.63 309 10.58
Chris Chandler 1995 44 3.57 6.12 118.32 356 10.51
Peyton Manning 2004 62 8.15 9.25 1580.54 497 10.50
Randall Cunningham 1998 41 6.85 8.46 1131.46 425 10.43
Chris Chandler 1998 59 7.61 8.76 969.29 327 10.08
Mark Brunell 1997 73 5.19 7.23 662.18 435 10.07
Brett Favre 1995 88 5.47 7.38 906.99 570 9.97
Steve McNair 2003 62 6.32 7.85 913.44 400 9.75
Chris Chandler 1997 42 5.90 7.54 624.21 342 9.62
Rich Gannon 2002 92 5.57 7.28 947.79 618 9.52
Scott Mitchell 1995 79 5.24 7.06 741.70 583 9.52
Trent Green 2003 53 5.51 7.15 827.77 523 9.28
Jeff Blake 1996 73 3.75 5.89 191.13 549 9.25

There are a few promising examples up there, but remember that table covers every QB for the past eleven seasons. If Plummer ends the season with a 7.00 AY/A, he would have a ratio of 20.94, which would rank second best in the past 11 years. A season ending adjusted yards per attempt of 6.00 would still give him a ratio of 15.4, and would land him squarely in the top five. So a turn-around from horrible to good happens once every few years; a turnaround from horrible to very good happens roughly once a decade.

Jake Delhomme started off 2003 poorly, but maybe not as badly as that AY/A indicates. He was 21/43 for 218 yards, 3 TDs and 4 INTs. Plummer's currently 29/56 for 311 yards, 0 TDs and 4 INTs. But that was Delhomme's first year starting and he was learning a new system, excuses Plummer can't rely on.

Steve McNair might be a better example. He had been on the Titans for awhile and had started for a couple of years, before having a miserable first two weeks in 2000: 28/49, 245 yards, 0 TD/3 INT. That line looks a bit like Plummer's actually; and McNair ended up having a (as of then) career year passing in 2000.

And look who is number three on the list: good ol' Jake. Plummer's thrown 10 INTs in his last four week one games, so there's a valid argument that he's just a slow starter. The 2003 Jake sure turned things around, and led the Broncos to a 9-2 record (including a 2-0 start, despite his subpar performances).

A rookie Charlie Batch had serious struggles in the beginning, but ended up being one of the better QBs in the NFL that year. Donovan McNabb struggled in 2003 -- and Rush Limbaugh had some choice comments on that -- but he turned it on in the second half and led the Eagles to the NFC Championship game. He was completing just 45% of his passes and had thrown 0 TDs against 4 INTs, so he looked just as terrible as Plummer.

Lots of quarterbacks start the season poorly. Occasionally, a very good QB having a very good season will be among that group. Plummer's been good enough for long enough that if I was Mike Shanahan, I'd give him the benefit of the doubt for two more games (especially considering his other slow starts). His next two games are against the Patriots and Ravens, two teams with historically tough defenses. New England's pass D isn't what it used to be, but the Ravens defense looks menacing this year. If Plummer can't play well this Sunday Night, he'll probably be forced to split reps with Jay Cutler during the week four bye. He might start week five, but there's a good chance he'd play poorly enough to be benched. It might be wisest for Denver to then insert Jay Cutler for games against Oakland and Cleveland, to prepare him for the rest of the season.

One word of caution: it's still worth noting that we're dealing with small sample sizes. If one of Plummer's interceptions had instead been a sack, his AY/A would be 3.20 instead of 2.34, and his future would look a bit more promising. But Plummer's AY/A was over 7.00 last year, which is the sign of a very good passer. If Plummer is able to reach that level again this year, it would truly be a historic turn-around. If he can't, Jay Cutler better be ready.

13 Comments | Posted in General

Rerun: The Dungy Index

Posted by Doug on September 20, 2006

It's a busy week, so I'm resorting to re-runs. This post was written for sabernomics in January of 2006.

I don’t know why this didn’t occur to me last year, but the methodology I developed in this post to rate quarterbacks on their postseason success makes as much sense — probably more sense — to apply to coaches rather than quarterbacks.

Since the methodology was described and discussed in the previous post, there’s not much to do here but post the list. So here goes.


expected actual Marginal
Coach record record wins
===========================================
Bill Belichick 7- 6 11- 2 +4.4
Joe Gibbs 13-10 17- 6 +3.7
Chuck Noll 12-12 16- 8 +3.9
John Fox 3- 4 5- 2 +2.3
Jimmy Johnson 6- 7 9- 4 +2.8
Brian Billick 3- 4 5- 2 +2.2
*Tom Landry 16-15 19-12 +3.1
Bill Walsh 8- 6 10- 4 +2.1
Jerry Burns 2- 4 3- 3 +1.3
Bill Parcells 9- 9 11- 7 +2.0
Raymond Berry 2- 3 3- 2 +1.3
Bum Phillips 3- 4 4- 3 +1.4
Tom Flores 7- 4 8- 3 +1.5
Barry Switzer 4- 3 5- 2 +0.9
Don McCafferty 3- 2 4- 1 +0.7
*John Madden 7- 7 8- 6 +0.8
John Robinson 3- 7 4- 6 +0.7
Dan Reeves 10-10 11- 9 +0.8
Mike Holmgren 10- 9 11- 8 +0.8
Mike Shanahan 7- 6 8- 5 +0.7
Herman Edwards 1- 4 2- 3 +0.5
Jon Gruden 5- 3 5- 3 +0.4
Jerry Glanville 3- 4 3- 4 +0.4
Ray Malavasi 3- 3 3- 3 +0.4
Marv Levy 11- 8 11- 8 +0.3
Jeff Fisher 5- 4 5- 4 +0.4
Sam Wyche 3- 2 3- 2 +0.2
Andy Reid 7- 5 7- 5 +0.0
Bill Cowher 11- 9 11- 9 -0.2
George Seifert 10- 5 10- 5 -0.2
Dick Nolan 2- 3 2- 3 -0.0
Tom Coughlin 4- 5 4- 5 -0.3
Ted Marchibroda 2- 4 2- 4 -0.2
Dave Wannstedt 2- 3 2- 3 -0.1
*Bud Grant 9- 9 8-10 -0.7
Bobby Ross 4- 4 3- 5 -0.6
Chuck Knox 8-10 7-11 -1.0
Art Shell 2- 3 2- 3 -0.4
Steve Mariucci 4- 3 3- 4 -0.6
Don Coryell 4- 5 3- 6 -0.8
Tony Dungy 6- 7 5- 8 -1.1
Dick Vermeil 7- 4 6- 5 -1.0
Red Miller 2- 3 2- 3 -0.5
*Don Shula 19-12 17-14 -2.1
*George Allen 3- 4 2- 5 -1.0
Mike Martz 4- 3 3- 4 -1.0
Jim Fassel 3- 2 2- 3 -1.0
Mike Ditka 7- 5 6- 6 -1.5
Mike Sherman 3- 3 2- 4 -1.2
Dennis Green 6- 6 4- 8 -2.0
Wayne Fontes 2- 3 1- 4 -1.5
Jack Pardee 3- 3 1- 5 -1.8
Marty Schottenheimer 9- 8 5-12 -3.7
Jim Mora 3- 3 0- 6 -3.3

[Technical note: this list includes all coaches who coached in at least five postseason games since 1970, and only includes games coached since 1970. The coaches marked with an asterisk coached in one or more postseason games prior to that, but those games are not counted.]

Remember, this measures a coach’s (team's) postseason performance relative to his (team's) postseason expectations. So if a coach, say Marty Schottenheimer, scores poorly here, it’s not clear whether he deserves blame for underachieving in the postseason or credit for overachieving during the regular season. [In a future post], I’ll take a stab at ranking the coaches based on their regular season performance only.

8 Comments | Posted in General, History

Running in neutral

Posted by Chase Stuart on September 19, 2006

After two weeks, the Jets have yet to establish any sort of running game. New York running backs have 52 rushes for only 120 yards, a meager 2.30 yards per carry average. The competition hasn't been particularly tough, either. The Patriots allowed 99 yards on 23 carries to Buffalo RBs, while the Titans let LaDainian Tomlinson and Michael Turner rush for 209 yards on 32 rushes.

The stats reveal the story: the Jets can't run the ball with any sort of consistency. In a key situation in the third quarter, New York ran had a 3rd and 1 at midfield, and ran twice for -1 yards. Derrick Blaylock's running style was much better suited for Kansas City's zone blocking offense, and he's been ineffective whenever he's not running in space. Kevan Barlow hasn't been much better, and is taken down by the first defender on nearly all his carries. The Jets offensive line, while looking good in pass protection, does not have the power or the experience to be an effective rushing offense.

That's my subjective opinion, of course, after watching both games. Things look bleak for the Jets, but probably no bleaker than they did for say, the 1996 Bengals. After two weeks, the Queen City's runners had just 55 yards on 34 carries, an ugly 1.62 YPC average. Cincinnati ended the season ranked 19th in rushing yards by RBs, and 19th in YPC, despite that terrible beginning. So what's so special about the '96 Bengals? That team averaged the fewest YPC by running backs after two weeks. The following table shows the ten worst teams from 1995-2005:


Team Year Car Yd YPC
cin 1996 34 55 1.62
atl 1997 50 84 1.68
crd 1999 54 110 2.04
ram 1996 44 91 2.07
det 1997 27 56 2.07
kan 2000 54 117 2.17
mia 2004 37 81 2.19
buf 2000 55 123 2.24
gnb 1998 49 113 2.31
tam 1996 49 115 2.35

So how did those teams end the season? Here's each team's season ending ranks in running back rushes, rushing yards and yards per carry.


Team Year Car Rk YD Rk YPC Rk
cin 1996 16 19 19
atl 1997 21 21 22
crd 1999 28 30 30
ram 1996 23 26 20
det 1997 13 1 1
kan 2000 21 23 19
mia 2004 31 32 31
buf 2000 13 19 27
gnb 1998 16 26 27
tam 1996 8 21 30
Average 19 22 23

Obviously, the 1997 Lions stand out, as Detroit ranked 1st in rushing yards despite ranking 13th in carries, and having a horrible start to the season. Barry Sanders had 25 carries for 53 yards after two weeks; Sanders rushed for over 100 rushing yards in each of the remaining fourteen games that season. The last week of the season -- against the Jets -- Sanders topped the 2,000 yard mark for the season.

Here's a quick rundown of the other nine teams.

1996 Cincinnati Bengals: Top overall draft pick Ki-Jana Carter was the main RB through two weeks, but was ineffective. Garrison Heart ended the season with 225-847-3.8, and Eric Bieniemy added in 269 yards on 56 carries.

1997 Atlanta Falcons: A year before the Dirty Bird took the country by storm, Jamal Anderson rushed for 1002 yards on 290 carries (3.5 YPC). Anderson struggles through two weeks continued for the first half of the season (124-323-2.60), but he turned things on in the second half of the season, rushing for 679 yards on 166 carries (4.09 YPC). The next season, Anderson would rush for 2,122 yards and 16 TDs in 19 regular and post-season games.

1999 Arizona Cardinals: Former Jet Adrian Murrell had 112 yards on 44 carries through two games, after rushing for 1,000 the three previous seasons. Murrell hit the wall that season, and the Cardinals had no rushing game. Three other Arizona RBs had ten carries for negative two yards, and it didn't get much better over the next four months. The highlight: Michael Pittman rushed for 133 yards on 23 carries in week ten.

1996 St. Louis Rams: After trading Jerome Bettis to Pittsburgh, the Rams handed the reins (and the ball) to first round pick Lawrence Phillips. Phillips' first two games were probably excused because of his inexperience, but things wouldn't get much better as the season progressed. Phillips did rush for 100 yards in two games, both against the Falcons. Phillips is not to be confused with DeShaun Foster. The Rams started off the season 4-10, but that didn't stop St. Louis from blowing out Atlanta 59-16; Harold Green also went over 100 yards in that game, and the Rams runners had 35 carries for 268 yards.

2000 Kansas City Chiefs: A year earlier, Chiefs QBs handed off to Chiefs RBs 497 times, the second most in the last eleven years (2004 Steelers). All-Pro Fullback Tony Richardson led the team with 147-697-4.7, after rushing 27 times for 86 yards the first two weeks. In weeks 15 and 16, Richardson carried 40 times for 245 times, but he wasn't utilized much in the middle of the season. Kansas City wasted 124 carries on Donnell Bennett, Mike Cloud and Frank Moreau, which netted only 287 yards.

2004 Miami Dolphins: This one's fresh in our memories, as Ricky Williams retired in August and left the Dolphins high and dry at the running back position. Miami has rebounded nicely after drafting Ronnie Brown, but the ugly start to 2004 was a sign of things to come. Lamar Gordon rushed for 54 yards on 31 carries after two weeks, and was placed on IR shortly after. Sammy Morris and Travis Minor were respectable for the next 14 games, but that didn't stop Miami from ranking last in rushing yards and second to last in yards per rush.

2000 Buffalo Bills: The 2000 Bills started off 2-0 despite their top RB rushing for 42 yards and 31 yards in those games. Buffalo would use several backs with minimal success, which looks to be the same formula for the 2005 Jets. Shawn Bryson, Sammy Morris (yes, of the famed 2004 Dolphins), and Antowain Smith would form one of the least intimidating running back-by-committee groups ever. Buffalo went 8-8 despite not having a 100-yard rushing performance until week 17. The Bills were actually 7-4 through 11 games, despite Shawn Bryson's 72 rushing yards being the season-to-date season high.

1998 Green Bay Packers: In 1997, Dorsey Levens ranked in the top five in yards from scrimmage and total TDs. Levens didn't play well in the first two games, and was injured in week two. Darrick Holmes rushed for 274 yards in games against the Eagles and Giants late in the season, and Levens would return and perform admirably down the stretch. Travis Jervey and Raymont "Ultraback" Harris were ineffective in the middle of the season, and Green Bay's rushing attack was well below average. But Brett Favre's MVP season assured that the Pack would return to the playoffs, where they'd lose to Terrell Owens and the 49ers thanks to "The Catch II."

1996 Tampa Bay Buccanneers: Errict Rhett had back-to-back 1,000 yard campaigns to start his career, but his hold out before the 1996 season left Tampa Bay without a runner. The Bucs would draft Warrick Dunn in 1997 and Rhett's career fizzled out shortly afterwards. For a year though, Tampa Bay had a pitiful ground attack. Reggie Brooks led the way with 20 yards in week 1, and Mike Alstott topped all Bucs with 19 rushing yards in week 2. Tampa Bay would end up running quite a bit, mostly because Trent Dilfer wasn't any good. But the Bucs started 0-5, and gave 288 carries to Brooks and Rhett. Those two runners had just 907 yards, for a 3.15 YPC average.

So what does this mean for the 2006 Jets?

The Jets don't have Barry Sanders or Brett Favre, so the post-season seems like a long-shot for New York right now. Most of those terrible running teams didn't recover, and gave 100 carries to multiple ineffective running backs. That seems like the most likely path for the Jets too, as Blaylock and Barlow should continue to split time. Sometimes, the statistics will reveal something you wouldn't figure out just by watching the games. For anyone that's seen the Jets play this year, this isn't one of those times.

6 Comments | Posted in General

Michael Vick and the Falcons

Posted by Doug on September 18, 2006

Michael Vick, who I'm a big fan of, had 127 yards rushing this week, which is the tops for a quarterback since 2003 and is third-most in the last decade or so. Here are the top rushing games by quarterbacks since 1995.


Player YR WK RSH YD TD
=========================================
Michael Vick 2002 13 10 173 2
Michael Vick 2003 14 14 141 1
Donovan McNabb 2000 13 11 125 1
Michael Vick 2004 8 12 115 0
Michael Vick 2004 2 12 109 0
Aaron Brooks 2000 15 11 108 0
Donovan McNabb 2002 8 8 107 1
Michael Vick 2004 11 15 104 0
Kordell Stewart 1998 6 7 103 0
Donovan McNabb 2002 5 12 100 1
Steve McNair 1998 10 8 95 1
Michael Vick 2002 8 10 91 2
Michael Vick 2002 7 6 91 1
. . .
Steve Bono 1995 5 5 74 1

For you youngsters who don't remember that Steve Bono game, this link is required viewing. Buddy Ryan and youtube: the ultimate combination.

Speaking of Vick, he led his team to another victory while throwing only 15 passes. Last week a reader, prompted by Philip Rivers' performance against the Raiders, asked how teams fare when they throw very few passes. As was conjectured, teams almost always win when they throw 15 or fewer passes. This list is since 2000.


TM YD WK OPP Passing Result
=======================================
sdg 2004 15 cle 4- 6- 85 W, 21- 0
dal 2000 15 was 5- 8- 89 W, 32-13
hou 2005 7 clt 6- 9- 48 L, 20-38
hou 2002 14 pit 3-10- 33 W, 24- 6
pit 2005 1 oti 9-11-218 W, 34- 7
sdg 2006 1 rai 8-11-108 W, 27- 0
oak 2003 11 min 9-13-195 W, 28-18
chi 2005 10 sfo 8-13- 67 W, 17- 9
jax 2002 6 oti 7-13- 84 L, 14-23
sfo 2005 10 chi 1-13- 28 L, 9-17
oak 2002 17 kan 7-14- 79 W, 24- 0
cin 2000 8 den 2-14- 34 W, 31-21
nyj 2003 11 clt 11-14-219 L, 31-38
pit 2005 7 cin 9-14- 93 W, 27-13
ten 2004 1 mia 9-14- 73 W, 17- 7
jax 2002 14 cle 10-14- 73 L, 20-21
dal 2001 12 was 7-14-130 W, 20-14
jax 2002 4 nyj 10-15-173 W, 28- 3
tam 2000 15 mia 11-15-147 W, 16-13
dal 2000 8 crd 9-15-154 W, 48- 7
tam 2000 14 dal 9-15- 65 W, 27- 7
buf 2003 14 nyj 9-15- 72 W, 17- 6
pit 2005 15 min 10-15-149 W, 18- 3
was 2001 16 nor 9-15- 94 W, 40-10
atl 2003 16 tam 8-15-119 W, 30-28
gnb 2003 12 sfo 10-15-138 W, 20-10

And it provides another example from the correlation-is-not-causation file. If the Texans had just quit throwing the ball in the second quarter of yesterday's game, they surely would have won.

Finally, let's put the Falcons impressive start into historical perspective. Atlanta is one of four teams to start the season with two wins and to hold their opponent to seven or fewer ponts in both games. The others are San Diego, Baltimore, and Chicago. Between 1978 and 2005, only 14 teams started the season this way:


TM YR Week 1 Week 2 Record
=========================================
pit 2005 ten, 34-7 hou, 27-7 11- 5- 0
ind 2005 bal, 24-7 jax, 10-3 14- 2- 0
sea 2004 nor, 21-7 tam, 10-6 9- 7- 0
ind 2003 cle, 9-6 ten, 33-7 12- 4- 0
sdg 2002 cin, 34-6 hou, 24-3 8- 8- 0
car 2002 bal, 10-7 det, 31-7 7- 9- 0
gnb 2001 det, 28-6 was, 37-0 12- 4- 0
jax 2001 pit, 21-3 ten, 13-6 6-10- 0
nwe 1997 sdg, 41-7 ind, 31-6 10- 6- 0
tam 1992 pho, 23-7 gnb, 31-3 5-11- 0
rai 1987 gnb, 20-0 det, 27-7 5-10- 0
nwe 1986 ind, 33-3 nyj, 20-6 11- 5- 0
buf 1981 nyj, 31-0 bal, 35-3 10- 6- 0
phi 1980 den, 27-6 min, 42-7 12- 4- 0

Despite generally high offense over the past few years, these kinds of starts have happened more in the past two years than they did in the entire decade of the 80s. That's a bit strange.

Also, this gives me an idea. What I ought to do is cook up a similarity scoring system for team games. Using that, I could identify the ten (or however many) teams throughout history who have started the season most like this year's Falcons (or whatever team) in terms of points scored, points allowed, and competition faced. Seeing how those teams ended up doing would be interesting. I'll try and get that done sometime later in the week.

10 Comments | Posted in General

Looking back, looking forward: Wide Receivers

Posted by Chase Stuart on September 15, 2006

With week one in the books, lots of football fans have already put 2005 in their rearview mirror. On the other hand, you've probably been bombarded with writers and sports anchors telling you "to remember it's only week one, and player X will be fine just like last season." That brings us to what's on my mind today: just what exactly did happen last year?

Reggie Wayne and T.J. Houshmanzadeh had a lot in common in 2005. Both played on playoff teams in the AFC, caught passes from a superstar QB, and played alongside a superstar wide receiver. Houshmanzadeh missed two complete games, while Wayne was rested for most of the last two weeks of the year. The two receivers had similar end of season stats, too.


Rec Yards Y/R TD Pts
Reggie Wayne 83 1055 12.7 5 218.5
T.J. Houshmanzadeh 78 956 12.3 7 227.8

On the surface, those numbers look pretty similar. But it turns out Houshmanzadeh had a much more impressive season than Wayne. Why? Houshmanzadeh produced better numbers despite playing a significantly harder schedule than Wayne.

In order to rank the receivers and the schedules, I needed a statistic. I used the following formula (no doubt familiar to fantasy football fans) to come up with "Pts" for each receiver in the NFL: (Receiving Yards)/10 + (Receiving Yards) * 6 + Receptions. (Note: I'm not terribly concerned with what's the "best" statistic to use here, like I was when I tried to rank the quarterbacks. This scoring system was just convenient for me and in my current database. It's fine for serving our current purpose.)

So which teams allowed the most points to opposing WRs?


Pts Rec RecYD TDs Ratio
San Francisco 49ers 664 231 3238 18 1.36
St. Louis Rams 576 192 2616 19 1.18
New England Patriots 567 174 2744 20 1.16
New York Giants 564 193 2773 15 1.16
Denver Broncos 555 208 2662 12 1.14
Houston Texans 555 181 2655 18 1.14
Tennessee Titans 553 168 2307 24 1.14
San Diego Chargers 539 206 2549 13 1.11
Seattle Seahawks 530 195 2447 14 1.09
Arizona Cardinals 522 194 2378 15 1.07
Miami Dolphins 515 178 2453 15 1.06
Cincinnati Bengals 505 183 2508 11 1.04
Kansas City Chiefs 501 170 2361 14 1.03
Buffalo Bills 496 178 2464 11 1.02
Carolina Panthers 488 185 2352 11 1.00
Minnesota Vikings 482 176 2197 14 0.99
Atlanta Falcons 476 188 2245 10 0.98
Washington Redskins 469 168 2270 12 0.96
Baltimore Ravens 464 176 2306 9 0.95
Jacksonville Jaguars 463 157 2202 14 0.95
Philadelphia Eagles 455 154 2150 14 0.93
Cleveland Browns 451 162 2064 13 0.93
Indianapolis Colts 449 170 2126 11 0.92
Detroit Lions 445 160 2100 12 0.91
Pittsburgh Steelers 438 171 2228 7 0.90
Dallas Cowboys 434 144 2147 12 0.89
Oakland Raiders 427 148 2033 12 0.88
New Orleans Saints 420 121 1936 17 0.86
Chicago Bears 419 160 2014 9 0.86
New York Jets 404 151 1733 12 0.83
Tampa Bay Buccaneers 398 139 2058 8 0.82
Green Bay Packers 369 122 1677 12 0.76
Average 487 172 2312 13

That last column on the right is what we're really concerned about. The Green Bay Packers allowed 24% less points to opposing WRs than the rest of the NFL. Only two wide receivers -- Braylon Edwards and Az-Zahir Hakim -- had over 100 receiving yards against Green Bay last year. On the flip side, the 49ers defense was horrible, a whopping 36% worse than the league average. Ten times a receiver went over the century mark, with Torry Holt, Anquan Boldin and Larry Fitzgerald doing it two times each.

If you follow this line of thinking, you'll conclude that catching 180 yards against the 49ers would be equivalent to 100 against the Packers. So I divided every receiver's point total from every game by the ratio of his opponent's stingiest relative to the league average. So when Joey Galloway had 5/53/2 against the Packers, those 22.3 points get converted to 29.5 points (22.3/0.757); when Galloway put up 8/149/1 against the 'Niners, those 28.9 points get converted to 21.2 points (28.9/1.363). This makes sense because a receiver should be rewarded for producing against a tough opponent and his performance should be diminished for dominating an inferior one.

I ranked every receiver in the NFL by how many points they scored in 2005. I've listed the top 75 receivers, and provided both their points totals and their corresponding ranks. Then I adjusted each player's point total by opponent, and listed the receiver's adjusted points and adjusted rank totals. The players are sorted not by either ranking, but by those "most hurt by a tough" schedule to "most helped by an easy schedule. (For you math guys out there, the list is sorted from lowest to highest after dividing the difference in points totals (actual points minus adjusted points) by the natural logarithm of the individual receiver's actual points.)


Name Pts Rk APts ARk
Steve Smith 340 1 363 1
Muhsin Muhammad 163 33 179 28
Travis Taylor 135 43 149 39
T.J. Houshmandzadeh 228 14 244 11
Roy Williams 162 35 176 30
Roddy White 93 71 105 61
Chad Johnson 298 3 312 2
Joey Galloway 272 7 286 4
Donald Driver 239 10 252 10
Marcus Robinson 113 53 124 48
Antonio Bryant 194 25 205 21
Derrick Mason 211 19 222 15
Michael Jenkins 105 62 114 55
Deion Branch 208 22 218 16
Braylon Edwards 101 66 109 57
David Givens 146 41 155 35
Ricky Proehl 92 72 100 67
Chris Henry 109 57 116 52
Antonio Chatman 131 45 139 44
Troy Brown 98 68 104 62
Az-Zahir Hakim 95 70 100 68
Dennis Northcutt 101 64 105 59
Brian Finneran 123 49 126 47
Ashley Lelie 133 44 136 45
Rod Smith 232 13 235 12
Josh Reed 89 74 90 73
Mark Clayton 112 54 114 54
Keenan McCardell 216 17 217 17
Donte Stallworth 207 23 208 20
Eric Moulds 187 28 187 26
Justin McCareins 127 46 127 46
Eric Parker 153 38 153 36
Hines Ward 234 12 233 14
Joe Horn 120 50 119 50
Darrell Jackson 105 61 104 63
Antwaan Randle El 104 63 103 64
Jabar Gaffney 118 51 116 53
Greg Lewis 111 55 109 58
Lee Evans 168 32 165 34
Eddie Kennison 213 18 210 19
Chris Chambers 269 8 266 7
Dante Hall 97 69 94 70
Andre Johnson 145 42 141 42
Laveranues Coles 188 27 183 27
Matt Jones 114 52 110 56
Amani Toomer 170 31 166 33
Brandon Lloyd 151 39 147 41
Samie Parker 107 59 103 65
Marty Booker 126 47 121 49
Jerry Porter 199 24 194 25
Arnaz Battle 87 75 82 75
Corey Bradford 107 58 102 66
Bryant Johnson 89 73 84 74
Plaxico Burress 239 11 233 13
Reggie Brown 125 48 119 51
Doug Gabriel 111 56 105 60
Santana Moss 286 6 279 5
Terry Glenn 223 15 216 18
Kevin Curtis 183 30 174 31
Jimmy Smith 208 21 200 23
Shaun McDonald 99 67 92 72
Drew Bennett 156 37 148 40
Terrell Owens 160 36 150 38
Randy Moss 209 20 198 24
Brandon Stokley 101 65 92 71
Ernest Wilford 151 40 140 43
Bobby Engram 163 34 151 37
Isaac Bruce 107 60 95 69
Marvin Harrison 269 9 254 9
Keyshawn Johnson 191 26 177 29
Joe Jurevicius 184 29 170 32
Reggie Wayne 219 16 202 22
Larry Fitzgerald 308 2 287 3
Anquan Boldin 289 5 267 6
Torry Holt 289 4 262 8

So what's this all mean? For starters, Steve Smith is really, really good. Sure he ranked first on the unadjusted list, but he also was hurt more than any other WR by a tough schedule (in part because of just how many total points he scored, but mostly because of a difficult schedule). Seven of Smith's games were against the toughest five teams in the league, and that doesn't even include Smith's dominant performance at Soldier Field in the playoffs. We saw this weekend just how much the Panthers missed Steve Smith, as Carolina didn't get a first down until the fourth quarter.

Two of my favorite receivers in the league, Anquan Boldin and Torry Holt, are the two players that get downgraded the most by this system. One word of caution: part of the reason Holt's schedule looks so easy is because his opponents (for the most part) had to face Holt, Anquan Boldin and Larry Fitzgerald last year. But in general, the Rams' and Cardinals' receivers had very easy schedules, which likely inflated their numbers. When you consider that the Rams lost Mike Martz and the Cardinals ranked 1st in pass attempts last season, there are lots of reasons to expect those three star receivers to have weaker years in 2005.

Another pair -- Chad Johnson and T.J. Houshmanzadeh -- excelled in '05 despite playing eight games against top 11 defenses (and Houshmanzadeh missed two games against the Jaguars and Titans, which explains why his schedule was slightly harder than Johnson's was). Of course, this formula does nothing to eliminate the huge advantage both Bengals receivers receive by getting to catch passes from Carson Palmer.

Muhsin Muhammad, Travis Taylor, Roy Williams, Roddy White and Joey Galloway all looked pretty good in this system. Doug Drinen's written a lot about how Galloway's constantly underappreciated, and this system concurs. Those players might be expected to have better years in 2006, with presumably lighter schedules. On the flip side, Joe Jurevicious and Keyshawn Johnson switched teams after having very easy schedules last year, so you'd probably expect a downgrade in their 2006 production.

That's the big value in this system. A player's ability is more consistent from year to year than a player's opposing schedule, even if he doesn't change teams. Remember, pre-season predictions about strength of schedule are usually meaningless. So if I was going to try and predict how a player was going to do this year, I'm confident that I'd be able to increase my accuracy by basing my projections on these adjusted 2005 numbers, rather than focusing on the raw data.

One last thing -- I've used this system with great success at the quarterback position. Things are a bit dicier at wide receiver, and I'm not exactly sure how this relates to Doug's article on defending number one and number two receivers. My system could probably be improved a good bit by combining these two things. I'll put it on my to do list.

3 Comments | Posted in Fantasy, General

More on spreading it around in the passing game

Posted by Doug on September 14, 2006

Last week in this post, I constructed a measurement of the extent to which a given team spreads the ball around to different receivers. It was based on seasonal yardage totals, and was certainly susceptible to pollution from a number of factors. For example, the 2005 Eagles ranked very high on the spread-it-around index, but that's largely due to the fact that Reggie Brown and Greg Lewis had similar yardage totals to Terrell Owens for the season. Had Owens stuck around, they would probably have ranked very low. This index begs to be interpreted as evidence of the size of the talent gap among a team's receivers or evidence of the coaching staff's philosophy. While it probably is evidence of one or both of those in many cases, there are definitely exceptions.

In the comments to that post, the always-perceptive monkeytime suggested a different way to measure spread-it-around-ness: average number of different pass receivers per game. Seems like a good idea, so I decided to run it. As usual, I'll just post the lists and let you pick through them to see if you can find anything interesting.

To make sure we're clear on what this means, look at the first line. It says that the 2005 Packers, in an "average" game, had 7.75 different players catch a pass. The MIN column says that they had no fewer than 6 players catch a pass in every game (and no, that 1 in the 49ers MIN column is not a typo). You can figure out what the MAX column is.

A caveat: this data comes from my database, which only includes offensive stats for QBs, RBs, WRs, and TEs. So tackle eligibles and passes to Mike Vrabel are not included. Once again, the genius of Tom Brady eludes quantification.


MIN AVG MAX
==========================
gnb 2005 6 7.75 10
ten 2005 5 7.75 9
ari 2005 5 7.56 9
min 2005 5 7.56 10
stl 2005 6 7.44 9
det 2005 6 7.38 9
cin 2005 4 7.31 9
nwe 2005 5 7.25 9
den 2005 6 7.12 9
tam 2005 5 7.00 9
jax 2005 5 6.88 9
phi 2005 5 6.81 11
buf 2005 4 6.69 9
kan 2005 4 6.62 9
bal 2005 5 6.56 9
sea 2005 3 6.56 10
cle 2005 5 6.44 8
nor 2005 4 6.44 9
mia 2005 4 6.44 9
nyj 2005 4 6.31 9
was 2005 5 6.31 9
sfo 2005 1 6.12 9
dal 2005 4 6.12 8
sdg 2005 5 6.06 8
atl 2005 4 6.00 7
oak 2005 4 6.00 8
pit 2005 2 6.00 9
ind 2005 5 5.88 8
car 2005 5 5.88 8
chi 2005 3 5.81 8
hou 2005 3 5.81 7
nyg 2005 3 5.31 7

oak 2004 7 8.56 11
gnb 2004 7 8.12 10
tam 2004 6 7.62 10
jax 2004 5 7.62 11
nyj 2004 3 7.25 9
det 2004 5 7.25 9
sfo 2004 5 7.19 9
ten 2004 5 7.19 10
min 2004 5 7.12 9
sea 2004 5 7.06 9
stl 2004 4 7.06 10
cin 2004 5 7.00 9
phi 2004 5 7.00 10
den 2004 5 7.00 9
bal 2004 5 6.94 10
kan 2004 6 6.88 8
nwe 2004 4 6.81 10
mia 2004 5 6.69 10
chi 2004 5 6.69 8
ari 2004 4 6.56 8
car 2004 4 6.56 11
buf 2004 5 6.38 9
dal 2004 5 6.19 8
cle 2004 4 6.19 8
nor 2004 4 6.12 8
ind 2004 5 6.12 8
hou 2004 3 6.00 8
was 2004 4 5.94 8
atl 2004 4 5.69 8
sdg 2004 3 5.56 8
nyg 2004 4 5.56 8
pit 2004 3 5.50 7

gnb 2003 6 7.94 12
phi 2003 6 7.62 10
det 2003 5 7.62 11
jax 2003 5 7.44 10
tam 2003 5 7.38 10
cle 2003 5 7.25 9
cin 2003 6 7.19 9
nwe 2003 5 7.12 10
ten 2003 5 7.00 8
was 2003 3 7.00 10
kan 2003 5 6.88 8
sfo 2003 5 6.81 9
nyj 2003 5 6.75 8
pit 2003 4 6.62 8
oak 2003 5 6.62 9
min 2003 5 6.56 9
car 2003 4 6.44 9
stl 2003 5 6.44 8
ari 2003 4 6.44 9
dal 2003 4 6.38 9
chi 2003 5 6.38 9
buf 2003 5 6.38 8
nyg 2003 5 6.31 8
ind 2003 4 6.31 8
nor 2003 5 6.25 9
sea 2003 5 6.25 8
den 2003 4 6.06 8
bal 2003 4 5.88 8
mia 2003 5 5.88 8
hou 2003 4 5.88 8
atl 2003 4 5.81 7
sdg 2003 4 5.62 8

nwe 2002 6 7.94 10
cin 2002 5 7.81 10
gnb 2002 6 7.56 10
tam 2002 6 7.25 9
phi 2002 4 7.19 10
stl 2002 4 7.19 9
nyj 2002 6 7.00 8
chi 2002 5 7.00 9
ten 2002 4 7.00 10
det 2002 6 7.00 9
dal 2002 4 7.00 10
was 2002 4 6.94 10
cle 2002 4 6.94 9
den 2002 4 6.88 9
sea 2002 5 6.81 9
oak 2002 4 6.81 9
mia 2002 5 6.69 9
nyg 2002 4 6.62 8
buf 2002 5 6.56 8
car 2002 3 6.56 10
sfo 2002 5 6.44 9
atl 2002 4 6.44 8
pit 2002 5 6.44 8
min 2002 5 6.44 7
ind 2002 5 6.31 8
kan 2002 5 6.31 8
ari 2002 4 6.25 8
nor 2002 4 6.25 8
bal 2002 5 6.12 8
sdg 2002 4 6.06 8
hou 2002 1 5.69 7
jax 2002 3 5.50 7

cin 2001 5 7.56 9
gnb 2001 5 7.50 9
sfo 2001 4 7.31 9
oak 2001 4 7.25 9
atl 2001 5 7.06 9
det 2001 5 7.06 9
bal 2001 4 7.00 9
min 2001 5 6.88 9
pit 2001 4 6.88 10
car 2001 5 6.88 8
cle 2001 5 6.75 8
den 2001 3 6.69 10
nwe 2001 5 6.62 9
tam 2001 5 6.62 8
buf 2001 5 6.56 9
ten 2001 4 6.56 10
mia 2001 4 6.56 8
stl 2001 5 6.56 8
was 2001 4 6.50 9
chi 2001 4 6.50 9
ari 2001 4 6.44 9
nyg 2001 5 6.44 9
sea 2001 4 6.38 8
kan 2001 5 6.38 8
phi 2001 4 6.31 8
sdg 2001 4 6.25 8
ind 2001 5 5.94 8
jax 2001 4 5.88 7
nor 2001 5 5.88 8
dal 2001 2 5.81 9
nyj 2001 4 5.69 7

phi 2000 5 7.56 10
det 2000 5 7.25 10
ari 2000 5 7.19 10
gnb 2000 4 7.12 9
oak 2000 4 7.00 9
was 2000 5 6.81 9
sea 2000 5 6.75 9
pit 2000 6 6.62 8
stl 2000 5 6.62 9
cle 2000 4 6.62 10
chi 2000 4 6.50 9
nyg 2000 5 6.44 8
min 2000 5 6.44 8
nwe 2000 4 6.38 9
sfo 2000 5 6.38 8
car 2000 5 6.38 7
mia 2000 5 6.31 10
den 2000 3 6.31 8
sdg 2000 5 6.19 9
buf 2000 5 6.19 8
ten 2000 5 6.19 9
cin 2000 2 6.12 9
bal 2000 4 6.12 9
nyj 2000 5 6.00 7
kan 2000 4 6.00 8
ind 2000 5 5.94 7
atl 2000 4 5.94 8
nor 2000 4 5.81 8
tam 2000 3 5.62 8
dal 2000 3 5.56 9
jax 2000 3 5.50 7

15 Comments | Posted in General

More week one thoughts

Posted by Doug on September 13, 2006

Just a few quick bits of trivia today...

If you're disappointed with the fantasy production you got from your wide receivers this week, you're not alone. Week one's top 10 receivers:


Donte Stallworth 20.1
Laveranues Coles 15.3
Plaxico Burress 14.0
Terrell Owens 14.0
Michael Jenkins 13.7
Larry Fitzgerald 13.3
Eric Moulds 12.8
Jerricho Cotchery 12.5
Anquan Boldin 12.2
Antonio Bryant 11.4

As a group they averaged 13.9 points, which is the lowest figure for the top 10 receivers in any week since (at least) 2000.

Of course, it wasn't just the wide receivers, and it wasn't just fantasy football. Teams simply weren't putting real points on the real scoreboard. In terms of average points scored, this was the 36th-lowest-scoring week (of 570 total weeks of football) since the merger. It was the lowest-scoring week since 2000. Since 1990, only 12 weeks have been lower-scoring.

The last time we saw as many as three shutouts in a week was 1991. The last time we saw more than three was 1983.

Another oddity about the just-completed slate of games is that the road team won 11 of the 16 games, which is 68.75%. In only 19 of the 570 weeks since the merger have road teams won a higher percentage of the weekend's games than they did last weekend. Random trivia I learned running that query: in week one of 1983, road teams won 12 of 14 games, which was the best week ever for road teams. Week 13 of the same season is the only week in which no road teams won.

3 Comments | Posted in General

The Tootsie Pop Effect

Posted by Jason Wood on September 12, 2006

Admit it, as a kid you thought there was an explicit answer to the owl's query about how many licks it took to get to the center of a Tootsie Pop, didn't you?

These days our curiosities may revolve around other things, but are no less nebulous at times.

Today my curiosity turns to how likely a wide receiver is to enjoy 1,000+ yards receiving for the first time in any given season. As an Eagles fan, my expectations for the newly-acquired Donte Stallworth were minimal. After all, he failed to break the 1,000-yard mark in his first four seasons. But his Week One output (6 receptions, 141 yards, 1 TD) has my homerism in full bloom and I now have visions of Fred Barnett dancing in my head. So I asked myself, how often is a receiver likely to break the 1,000-yard mark in Year 5 or later after not doing so in Years 1-4?

Among players debuting in 1990 or later, 75 wide receivers have recorded at least one 1,000-yard season. Of those, only 20 had their first 1,000-yard campaign in Year 5 or later:

LastName FirstName 1st 1K Season # 1K Seasons
Kennison Eddie 9 2
Brown Troy 9 1
Jackson Willie 8 1
McCaffrey Ed 8 3
Ismail Qadry 7 2
Ismail Raghib 6 2
Scott Darnay 6 1
Stokley Brandon 6 1
Martin Tony 6 4
McDuffie O.J. 6 1
Jackson Michael 6 1
Horn Joe 5 4
Mason Derrick 5 5
McCardell Keenan 5 5
Graham Jeff 5 1
Westbrook Michael 5 1
Chambers Chris 5 1
Moore Rob 5 3
Thigpen Yancey 5 2
Mathis Terance 5 4

Of those 20, nine (45%) had one and only one 1,000-yard season while seven (35%) have recored three or more 1K seasons. More germane to my Stallworth curiosity, nine (45%) of the 20 came in Year 5. It appears that WRs who hit the 1K mark in Year 5 have a higher rate of excellence in future years:

  • Average # of 1K seasons for WRs who have their first 1K season in Year 5 = 2.89
  • Average # of 1K seasons for WRs who have their first 1K season in Year 6+ = 1.73

20 instances in 15 years is, in my mind, neither something we can label as common or exceedingly rare. The fact is, it happens, but it's less likely a receiver will break into the elite 1,000-yard category and stay there if it takes him until his 5th season.

  • Average # of 1K seasons for WRs who have their first 1K season in Year 5+ = 2.25
  • Average # of 1K seasons for WRs who have their first 1K season in Years 1-4 = 2.82

I'm not sure whether Donte Stallworth will have a career like Derrick Mason's, but I'll just be happy if he doesn't end up with one like Michael Westbrook's.

6 Comments | Posted in Fantasy, History, Statgeekery

Week one success

Posted by Doug on September 11, 2006

Here is a quick statistical look at the relevance of week one.

I looked at all week one games during the last ten years and ran a regression of the margin of victory (or loss) in week one versus the rest-of-year win total for the team. Here is the formula:


RestOfYearWins =~ 7.5 + .035 * Week1Margin

The week one margin was indeed highly significant in the official statistical sense. That means that it definitely tells us something. But the coefficient is small, which means that, for practical purposes, it doesn't tell us much. The formula essentially projects everyone to win half its remaining games, give or take just a little.

If you just woke up from a multi-year coma, read this post (obviously the first thing you'd want to do), and then looked at the week's scores, you would expect that the Rams will win 8.8 games in 2006 --- the one they've already won, plus 7.5+.035*8 more --- and the Broncos will win 7.2.

I also tried including strength of opponent (based on last year's strength) into the regression, but it did not turn out to be significant.

15 Comments | Posted in General, Statgeekery

A last look at QBs

Posted by Chase Stuart on September 8, 2006

The past couple of days, we've looked at the best and worst quarterbacks of all time. Today, we're going to tie up some loose ends.

A new number one

Some of the commenters to yesterday's post wondered what would happen if we included data from the '60s in this study. I decided to do that, and went all the way back to 1956 -- Johnny Unitas' rookie year. The data from before 1960 might not be 100% accurate, so I included an asterisk next to any player who played in the 1950s.

Once you include the 1960s data, Fran Tarkenton shoots to number one. Johnny Unitas comes in at number five, and Sonny Jurgensen joins the list at number seven. Len Dawson and Bart Starr also are in the top fifteen, which means five of the top twelve QBs played a significant part of their career in the 1960s.


Player Name Value Career Attempts
Fran Tarkenton 7134 6467
Steve Young 7103 4149
Dan Marino 6752 8358
Joe Montana 6634 5391
Johnny Unitas* 6145 5186
Roger Staubach 5387 2958
Sonny Jurgensen* 5152 4262
Ken Anderson 5135 4475
Dan Fouts 5017 5604
Peyton Manning 4927 4333
Len Dawson* 4437 3741
Bart Starr* 4354 3149
Trent Green 3788 3329
Kurt Warner 3487 2340
John Elway 3155 7250
Roman Gabriel 2931 4498
Warren Moon 2908 6823
Jim Kelly 2885 4779
Bob Griese 2874 3429
Craig Morton 2704 3786
Brett Favre 2672 7612
Daunte Culpepper 2619 2609
Boomer Esiason 2592 5205
Mark Brunell 2429 4334
Earl Morrall* 2418 2689
Phil Simms 2368 4647
Neil Lomax 2364 3153
Bert Jones 2295 2551
Steve McNair 1973 3871
Rich Gannon 1939 4206
Don Meredith 1907 2308
Daryle Lamonica 1866 2601
Frank Ryan* 1808 2133
Dave Krieg 1689 5311
Y.A. Tittle* 1663 2339
Tom Brady 1614 2548
Matt Hasselbeck 1603 2205
Randall Cunningham 1593 4289
Steve Beuerlein 1583 3328
Marc Bulger 1552 1518
Ken O'Brien 1524 3602
John Brodie* 1514 4491
Bernie Kosar 1501 3365
Neil O'Donnell 1499 3229
Jeff George 1456 3967
Terry Bradshaw 1453 3901
Troy Aikman 1439 4715
Joe Theismann 1429 3602
Bob Berry 1319 1173
Joe Namath 1296 3762

The last name on that list might be surprising. Lots of modern football analysts like to call Joe Namath the most overrated player in the HOF, and point to his high number of INTs as evidence that he was much more style than substance. I even thought Namath might fall below the average in my scoring system, but there he is -- not far from three other Super Bowl winners: Terry Bradshaw, Troy Aikman and Joe Theismann.

I won't reprint the single-season leaders, but here are the players from the 1960s that now join the top 50 list:


Name Year Tm Att AY/A Value Added
Johnny Unitas 1964 bal 305 9.00 1156
Milt Plum 1960 cle 250 9.13 1116
Bart Starr 1966 gnb 251 9.01 1015
Johnny Unitas 1963 bal 410 7.66 956
Sonny Jurgensen 1967 was 508 6.57 937
George Blanda 1961 hou 362 7.46 936
Sonny Jurgensen 1961 phi 416 7.12 936
Len Dawson 1966 kan 284 8.23 926
Y.A. Tittle 1963 nyg 367 7.83 919
Johnny Unitas 1967 bal 436 6.67 849

One season that's not on the list but deserves mention: Greg Cook, 1969. Cook had the fourth best season of any QB with fewer than 200 attempts. He averaged 7.66 AY/A and added 506 points of value. Good numbers, but what makes it so interesting? Cook was a rookie. Cook threw just three more passes in his NFL career. He suffered a shoulder injury before his second season started, which basically ended his career. But for one year, Cook looked like the next great QB.

The worst QBs

You didn't think I'd just look at the best QBs of the 1960s, did you? George Blanda had the third and sixth worst seasons of all time. In other words, Joey Harrington has company. Here are the players from before the merger that would join the list of the worst 50 seasons of all time.


Name Year Tm Att AY/A Value Added
George Blanda 1962 hou 418 2.85 -941
George Blanda 1965 hou 442 3.15 -878
Frank Tripucka 1961 den 344 2.46 -832
George Herring 1961 den 211 1.04 -808
Tobin Rote 1959 det 162 0.35 -776
Dan Darragh 1968 buf 215 1.47 -760
Dick Wood 1966 mia 230 1.75 -740
Zeke Bratkowski 1960 chi 175 0.95 -650
Al Dorow 1961 nyj 438 3.40 -644
Norm Snead 1968 phi 291 2.82 -638
Jack Kemp 1969 buf 344 3.26 -630
Mike Taliaferro 1968 nwe 176 1.44 -628
Cotton Davidson 1962 rai 321 3.15 -624
John Hadl 1962 sdg 260 2.70 -623

Did you catch that the 1961 Broncos have two QBs on that list? And on the original list, the 1998 Chargers also had two QBs on there? Let's take a look at undoubtedly two of the worst passing duos of all time.


Player Year TM | G | Comp Att PCT YD Y/A TD INT
Craig Whelihan 1998 sdg | 11 | 149 320 46.6 1803 5.6 8 19
Ryan Leaf 1998 sdg | 10 | 111 245 45.3 1289 5.3 2 15
Frank Tripucka 1961 den | 14 | 167 344 48.5 1690 4.9 10 21
George Herring 1961 den | 14 | 93 211 44.1 1160 5.5 5 22

The 1961 Broncos "win" out: they "added" -1640 points of value, while the Chargers combo had only -1612.

On the career list, George Blanda again factors in pretty prominently. Here are the worst 50 QBs of all-time, updated to include QBs from 1955 and later.


Player Name Value Career Attempts
Rick Mirer -2512 2043
Mike Phipps -2447 1799
Joey Harrington -2374 1802
Dan Pastorini -2272 3055
Mark Malone -2146 1648
Mike Taliaferro -2083 966
Frank Tripucka -1868 1277
George Blanda* -1825 3114
Jack Trudeau -1780 1644
King Hill* -1777 881
Trent Dilfer -1719 2952
Dick Wood -1715 1194
Kordell Stewart -1676 2358
Bobby Douglass -1597 1178
Jack Concannon -1591 1110
Zeke Bratkowski* -1586 1354
Ryan Leaf -1560 655
Steve Walsh -1548 1317
Danny Kanell -1540 956
Scott Brunner -1474 1046
John McCormick -1474 555
Pete Beathard -1462 1282
Jon Kitna -1451 2837
Kent Nix -1446 652
Randy Wright -1435 1119
Mike Pagel -1415 1509
Kim McQuilken -1413 272
Randy Johnson -1409 1286
Rick Norton -1350 382
Gary Huff -1343 788
Billy Joe Tolliver -1314 1707
Kyle Boller -1270 981
Kerry Collins -1226 5082
Jack Kemp* -1222 3073
Dave Brown -1218 1634
Babe Parilli* -1216 2987
Craig Whelihan -1212 557
Richard Todd -1194 2967
Al Dorow* -1191 1057
George Izo -1133 317
Marc Wilson -1117 2081
Todd Blackledge -1114 881
Jake Plummer -1098 4033
Vinny Testaverde -1083 6526
Joe Kapp -1083 918
Dan Darragh -1078 296
Heath Shuler -1049 593
Jack Thompson -1043 845
Karl Sweetan -1041 590
Chris Weinke -1039 591
Mike Tomczak -1030 2337

Value added to a franchise?

Rick Mirer's the worst QB of all time, by my system's ranking. But he played on several teams throughout his career. I was curious to see which QBs hurt which franchises the most. Lions fans, this isn't pretty.


det Joey Harrington -2374
ari Jake Plummer -2323
den Frank Tripucka -1868
hou George Blanda -1794
tam Vinny Testaverde -1725
cle Mike Phipps -1702
pit Mark Malone -1695
sea Rick Mirer -1654
tam Trent Dilfer -1541
det Milt Plum -1510

There you have it. Joey Harrington has hurt the Detroit Lions more than any other QB has ever hurt any other team. The Lions won just 19 games during the Joey Harrington era, which spanned four miserable seasons.

And what about looking at things the other way? What QB has helped his franchise more than any other? Meet the new number one, same as the old number one.


sfo Steve Young 7520
mia Dan Marino 6752
bal Johnny Unitas 6308
sfo Joe Montana 6135
dal Roger Staubach 5387
cin Ken Anderson 5135
sdg Dan Fouts 5017
ind Peyton Manning 4927
kan Len Dawson 4657
min Fran Tarkenton 4604

Peyton Manning will almost certainly move into fifth place on this list by the end of 2006. For what it's worth, Fran Tarkenton's the first QB to come up twice on the list. His work with the Vikings was the 10th best of all-time; the value he added to the Giants ranks 23rd best. Sonny Jurgensen (12th - Washington; 43rd - Philadelphia) and Craig Morton (41st - Denver, 52nd - Dallas) also had great success with a couple of teams.

Eliminating the negatives

There's lots of debate about how good Troy Aikman really was. But his failure to look great in my system is a direct result of three awful seasons, which most Cowboys fans probably don't care about anymore.


1989 -673
1990 -448
1991 348
1992 438
1993 789
1994 193
1995 647
1996 84
1997 -25
1998 400
1999 104
2000 -416
Total 1439

Personally, I think Aikman deserves to be downgraded because of his terrible 1989, 1990 and 2000 seasons. But I was curious how things would look if I made all negative seasons equal to zero. How would that change the career rankings list?

Troy Aikman moves up to 26th. Steve Young tops Tarkenton again. And Brett Favre moves from QB15 to ... QB15.


Name AdjVA Att
Steve Young 7697 4149
Fran Tarkenton 7546 6467
Dan Marino 7087 8358
Johnny Unitas* 6821 5186
Joe Montana 6648 5391
Roger Staubach 5669 2958
Dan Fouts 5652 5604
Ken Anderson 5565 4475
Peyton Manning 5520 4333
Sonny Jurgensen* 5449 4262
Bart Starr* 4900 3149
Len Dawson* 4852 3741
Boomer Esiason 4323 5205
John Elway 4177 7250
Brett Favre 4110 7612
Trent Green 3876 3329
Kurt Warner 3841 2340
Warren Moon 3670 6823
Craig Morton 3394 3786
Bob Griese 3332 3429
Roman Gabriel 3320 4498
John Hadl 3111 4687
Earl Morrall* 3108 2689
John Brodie* 3082 4491
Phil Simms 3055 4647
Troy Aikman 3001 4715
Daunte Culpepper 2947 2609
Mark Brunell 2935 4334
Bert Jones 2930 2551
Jim Kelly 2895 4779
Terry Bradshaw 2877 3901
Rich Gannon 2801 4206
Milt Plum* 2773 2419
Ken Stabler 2686 3793
Neil Lomax 2681 3153
Chris Chandler 2653 4005
Y.A. Tittle* 2624 2339
Dave Krieg 2618 5311
Daryle Lamonica 2563 2601
Jim Hart 2549 5076
Joe Namath 2522 3762
Jeff George 2501 3967
Don Meredith 2448 2308
Randall Cunningham 2440 4289
Steve McNair 2410 3871
Vinny Testaverde 2292 6526
Joe Theismann 2285 3602
Jim Everett 2241 4923
Bernie Kosar 2217 3365
Steve Beuerlein 2170 3328

Biggest movers

Who made the biggest jumps in one year? Here are the 27 QBs who saw an increase in value added of over 900 points from one year to the next. Note: Rookie QBs were not considered.


Name GoodYr Tm VA Increase in value added
Kurt Warner 1999 stl 1367 1391
Peyton Manning 1999 ind 717 1310
Randall Cunningham 1998 min 1131 1253
Tobin Rote 1963 sdg 421 1197
Ken Anderson 1981 cin 953 1178
Boomer Esiason 1993 nyj 442 1132
Jake Plummer 2003 den 336 1129
Drew Brees 2004 sdg 686 1125
Dan Marino 1984 mia 1566 1095
Scott Mitchell 1995 det 742 1093
Roger Staubach 1971 dal 900 1093
Ken O'Brien 1985 nyj 1056 1073
Carson Palmer 2005 cin 671 1026
John Brodie 1961 sfo 809 1008
Vinny Testaverde 1998 nyj 792 999
Drew Bledsoe 1996 nwe 228 994
Tommy Kramer 1986 min 710 991
Mark Rypien 1991 was 952 991
George Blanda 1961 hou 936 965
Boomer Esiason 1988 cin 1071 964
Craig Morton 1977 den 614 941
Bernie Kosar 1991 cle 450 930
Boomer Esiason 1985 cin 769 929
John Elway 1993 den 750 915
John Hadl 1973 ram 548 913
Jeff Garcia 2000 sfo 952 905
Jack Trudeau 1987 ind 66 902

And then the 28 QBs that saw a decrease of 900 or more points in value added from one year to the next.


Name BadYr Tm VA Decrease in value added
George Blanda 1962 hou -941 -1877
Y.A. Tittle 1964 nyg -555 -1475
Daunte Culpepper 2005 min -166 -1452
Kurt Warner 2002 stl -298 -1414
Rudy Bukich 1966 chi -523 -1356
Dan Marino 1985 mia 329 -1236
Steve McNair 2004 ten -287 -1200
Norm Snead 1968 phi -638 -1197
Steve Young 1999 sfo -177 -1172
Brian Griese 2001 den -346 -1136
Tom Flores 1967 buf -402 -1130
Randall Cunningham 1999 min 23 -1108
Brett Favre 2005 gnb -736 -1081
Kerry Collins 1997 car -886 -1058
Rich Gannon 2003 oak -98 -1046
Jake Plummer 2002 ari -794 -1031
Johnny Unitas 1968 bal -181 -1030
Brian Sipe 1981 cle -206 -1029
Bob Lee 1974 atl -525 -995
Roger Staubach 1972 dal -89 -989
Elvis Grbac 2001 bal -265 -973
Ron Jaworski 1981 phi -117 -951
Mark Rypien 1992 was 7 -946
Erik Kramer 1996 chi -290 -945
Babe Parilli 1963 nwe -401 -919
John Brodie 1971 sfo -4 -915
Roman Gabriel 1974 phi -110 -906
Steve Young 1995 sfo 316 -905

George Blanda followed up the 34th best season of all-time with the 3,424th best season of all time -- or stated another way, the 3rd worst season of all time. Wow. Here's his stat line from those two years:

Year  TM |   G |  Comp   Att   PCT    YD   Y/A  TD INT 
1961 hou |  14 |   187   362  51.7  3330   9.2  36  22 
1962 hou |  14 |   197   418  47.1  2810   6.7  27  42 

Blanda went from averaging a TD every 10 throws to an INT every 10 throws -- I'm guessing 1961 was his contract season. The 1961 Houston Oilers are actually the answer to my favorite trivia question: What is the only team to ever have the number 1 and number 2 fantasy WRs in the same season? That year, Bill Groman had 1,175 receiving yards and 18 total TDs, while Charlie Hennigan led all receivers with 1,746 yards and twelve scores.

How the might have fallen, II

What about the biggest changers over a two year period? I wasn't surprised to see who topped this list. Mark Rypien had 28 TDs, 11 INTs and averaged 8.5 unadjusted yards per attempt in 1991; two years later, Rypien had a 4/10 TD/INT ratio and averaged a paltry 4.7 Y/A. Here are the 25 quarterbacks who dropped the most from year N to year N+2.


Name BadYr Tm VA Decrease in value added
Mark Rypien 1993 was -679 -1632
Joe Theismann 1985 was -545 -1466
Vinny Testaverde 2000 nyj -534 -1326
Y.A. Tittle 1964 nyg -555 -1297
Milt Plum 1963 det -591 -1265
Fran Tarkenton 1978 min -403 -1170
Kurt Warner 2002 stl -298 -1156
Joe Montana 1992 sfo 30 -1148
Kurt Warner 2003 stl -32 -1147
Milt Plum 1962 det -29 -1145
Neil Lomax 1986 stl -181 -1144
Boomer Esiason 1990 cin -64 -1135
Randall Cunningham 2000 dal 19 -1112
Bert Jones 1978 bal 168 -1102
Steve Young 1999 sfo -177 -1094
Brian Sipe 1982 cle -268 -1091
George Blanda 1963 hou -136 -1072
Chris Chandler 2000 atl -84 -1053
Steve Deberg 1992 tam -129 -1033
Daunte Culpepper 2002 min -162 -1017
Jake Plummer 1999 ari -1025 -1013
Jon Kitna 2001 cin -936 -1004
Johnny Unitas 1966 bal 159 -997
Jay Schroeder 1992 rai -304 -989
Bobby Layne 1959 pit -284 -986

Notice that the two biggest drop-offs over a two-year period were both Redskins? Well the two biggest improvements also came from the same team.


Name GoodYr Tm VA Increase in value added
Daunte Culpepper 2004 min 1286 1448
Randall Cunningham 1998 min 1131 1422
Peyton Manning 2000 ind 825 1418
Brett Favre 1995 gnb 907 1412
Archie Manning 1978 nor 658 1390
Peyton Manning 2004 ind 1581 1362
Bert Jones 1976 bal 1270 1340
Jake Plummer 2001 ari 237 1262
Jon Kitna 2003 cin 281 1218
Bert Jones 1975 bal 690 1147
Elvis Grbac 2000 kan 707 1145
Jim Zorn 1978 sea 456 1116
Steve Young 1992 sfo 1184 1099
Drew Bledsoe 1997 nwe 330 1097
Jake Plummer 2004 den 297 1090
John Brodie* 1961 sfo 809 1078
Doug Williams 1981 tam 526 1041
Troy Aikman 1991 dal 348 1021
Greg Landry 1971 det 635 1006
Chris Chandler 1998 atl 969 979
Johnny Unitas 1964 bal 1156 977
Jim Everett 1989 ram 825 976
Drew Brees 2004 sdg 686 975
Boomer Esiason 1986 cin 812 972
Sonny Jurgensen 1967 was 937 969

Age cut-offs
(Note: If you read this before 11:30 EST, you probably didn't see this. It was a late edit when I was thinking about things this morning.)

I was curious to see how the numbers looked after I played with the ages of the QBs. Which QB had compiled the most impressive career by the age of 25?


Name Value Att
Dan Marino 3035 2050
Ken Anderson 1742 1089
Fran Tarkenton 1516 1541
Bert Jones 1433 1065
Bernie Kosar 1425 1427
Boomer Esiason 1421 1002
Johnny Unitas 1398 762
Joe Namath 1308 1682
Neil Lomax 1238 1355
Peyton Manning 1215 2226
Ben Roethlisberger 1105 563
Ken O'Brien 1039 691
Pat Haden 934 765
Roman Gabriel 804 698
Daunte Culpepper 792 1391
Brian Griese 725 791
Joe Montana 582 784
Dave Krieg 518 435
Charlie Batch 513 573
Greg Cook 506 197
Jim McMahon 499 648
Marc Bulger 466 214
Tony Eason 434 526
Jay Schroeder 432 750
Sam Wyche 403 220

That's a very interesting list. Players like Namath and Kosar have a good bit of mystique about them, probably not commensurate with their career production. That could probably be explained by how good they looked early on. After seeing how good Charlie Batch and Tony Eason looked early on, maybe I should have relaxed my optimism on Greg Cook's young career. And while they don't have much of a legacy, players like Neil Lomax and Boomer Esiason were legitimate star QBs before they turned twenty-five.

How about the QBs that just looked miserable early in their careers? Any of them turn it around? Troy Aikman just missed the cut-off: he was the 28th worst QB by the age of 25.


Name Value Att
Akili Smith -810 420
Gary Marangi -817 283
Jeff George -841 1125
Danny Kanell -844 653
Vinny Testaverde -873 631
David Woodley -880 961
Rick Norton -880 229
Steve Deberg -882 880
Kent Nix -935 451
Tim Couch -943 1511
Bobby Douglass -952 601
Gary Huff -957 614
Drew Bledsoe -1015 2901
Dan Pastorini -1059 1106
Dan Darragh -1078 296
Trent Dilfer -1179 1365
Terry Bradshaw -1188 1079
Jake Plummer -1200 1224
Kerry Collins -1215 1177
Randy Johnson -1231 832
Rick Mirer -1241 1258
Kyle Boller -1270 981
Kim McQuilken -1386 261
Ryan Leaf -1560 655
Joey Harrington -1795 983

A few notable players on there. Kerry Collins and Jake Plummer turned things around a little bit, although both still have their doubters. Vinny Testaverde, Jeff George and Trent Dilfer had inconsistent success. Drew Bledsoe had a pretty nice career, and of course Terry Bradshaw won four Super Bowls. Later in his career, Randy Johnson would go on to win five Cy Youngs. (Ok, so it wasn't that that Randy Johnson.)

What about the old-timers? I gushed about Roger Staubach yesterday. Was he the best QB in NFL history from age 30 on?


Name Value Att
Steve Young 6733 3324
Roger Staubach 4578 2618
Sonny Jurgensen 3588 3155
Trent Green 3567 2819
Len Dawson 3509 2680
Fran Tarkenton 3401 3445
Joe Montana 3364 2820
Johnny Unitas 3334 2870
John Elway 3235 4180
Phil Simms 3043 3651
Bart Starr 3027 1603
Dan Fouts 2868 3010
Dan Marino 2788 4177
Warren Moon 2758 5996
Rich Gannon 2688 3078
Earl Morrall 1962 1528
Terry Bradshaw 1903 1882
Jim Kelly 1870 3037
Ken Anderson 1793 2029
Bob Griese 1665 1415
Chris Chandler 1608 2763
Billy Kilmer 1574 2388
Joe Theismann 1520 2834
John Brodie 1425 2987
Jeff George 1384 1255

Steve Young destroys him on this list. Neither Young nor Staubach played much in their 20s, and it's hard to tell if they were so great in their thirties because of that, or because they were just excellent quarterbacks. One thing's for sure: you can't blame either of them. Staubach had an obligation to join the Navy; Young's predicament was even grimmer -- he was the quarterback behind Joe Montana. Trent Green will have his doubters because he's never really played in a big market, but he's put up some great numbers late in his career.

What about the QBs that didn't exactly age like fine wine?


Name Value Att
Steve Dils -493 342
Steve Bono -497 1339
Mike Pagel -522 204
Gary Cuozzo -530 326
Stan Gelbaugh -552 273
Jim Zorn -561 381
Jay Schroeder -572 1007
Jack Trudeau -585 451
Jeff Kemp -601 295
Boomer Esiason -609 2518
Bubby Brister -615 851
Frank Reich -650 762
Steve Walsh -695 126
Mike Phipps -745 482
Joe Kapp -773 704
Shane Matthews -777 547
Joe Namath -782 1157
Mark Rypien -910 1204
Richard Todd -925 862
Doug Pederson -936 514
Scott Mitchell -948 330
Dan Pastorini -1098 611
Vince Ferragamo -1106 393
George Blanda -1838 3045
Joe Ferguson -2247 2406

Not a very noteworthy bunch. You wonder how Joe Ferguson was able to hold on for so long, though. He was miserable at age 32, very bad at age 33, yet the Bills gave him the starting reins at age 34 -- when he was terrible once again.

What about the QBs that should have retired a bit earlier? Here are the 15 worst QBs from age 35 and up.


Name Value Att
Frank Reich -339 471
Dan Fouts -345 794
Chris Chandler -368 1111
Brett Favre -390 1147
Jeff Hostetler -403 546
Bobby Hebert -430 533
Erik Kramer -436 141
Joe Ferguson -470 353
Jim Harbaugh -549 929
Johnny Unitas -557 1089
John Hadl -814 490
Steve Grogan -863 493
Jim Hart -1017 1175
Ken Stabler -1473 1312
George Blanda -2568 2294

There's a certain amoutn of prestige associated with just being able to start NFL games at the QB position at 35, 36 or 37 years old. But those guys -- especially Blanda - did their teams no favors. Blanda threw 163 INTs after the age of 34.

Let's end things today on an upbeat note. Here are the top fifteen QBs that were wise not to hang up it once they turned 35.


Name Value Att
Roger Staubach 2588 1235
Steve Young 2262 1273
Rich Gannon 2021 1933
John Elway 1893 1866
Sonny Jurgensen 1438 1178
Phil Simms 1367 1394
Randall Cunningham 1184 839
Len Dawson 1133 1344
Fran Tarkenton 1111 1667
Craig Morton 992 1340
Steve Deberg 916 1613
Warren Moon 869 3798
Trent Green 801 507
Charley Johnson 768 732
Billy Kilmer 743 1033

You'll notice Moon on there, which makes me wonder what his career could have been like had he never played in the CFL. Then you see a couple of quarterbacks like Rich Gannon and Randall Cunningham, who left the NFL entirely before coming back and having some dominant seasons. One last notable that missed the cut-off: Doug Flutie, 611 value added points.

31 Comments | Posted in General, History

The Best QB of all time?

Posted by Chase Stuart on September 7, 2006

If you haven't done so already, I'd suggest reading yesterday's post discussing the worst quarterbacks of all time. Today, well examine the best QBs in the NFL over the past 36 years using the same system.

Before I write about The Best QB Ever, I want to address some of the main responses from yesterday, since they'll be relevant to any discussion about my system. There were lots of great responses, but I'll try and divide them up into the three groups.

  • Any system that says Trent Dilfer was worse than Ryan Leaf is crazy.

There's certainly some merit to this. Dilfer won a Super Bowl, while Leaf is rightly considered one of the biggest busts in sports history. But remember that this system totally ignores post-season numbers, so you have to forget about Dilfer's playoff performances (or at least accept that this is the system we'll be using).

Now that I've had a day to think about it, I think I can better explain what yesterday's list was really saying. Let's continue to use Leaf and Dilfer as examples.

In the table, "Att" represents pass attempts, "Yds" shows passing yards, and LgAvg provides the league average Adjusted Yards per Attempt (AY/A) for that season.

Ryan Leaf: -1560 career value added


Year Tm Att Yds TD INT AY/A LgAvg Value Added
1998 sdg 245 1289 2 15 2.59 5.79 -786
2000 sdg 322 1883 11 18 3.67 5.68 -646
2001 dal 88 494 1 3 4.19 5.65 -128

Trent Dilfer: -1719 career value added


Year Tm Att Yds TD INT AY/A LgAvg Value Added
1994 tam 82 433 1 6 2.11 5.74 -298
1995 tam 415 2774 4 18 4.83 5.79 -399
1996 tam 482 2859 12 19 4.41 5.54 -548
1997 tam 386 2555 21 11 5.88 5.71 66
1998 tam 429 2729 21 15 5.28 5.79 -222
1999 tam 244 1619 11 11 5.06 5.63 -141
2000 bal 225 1502 12 11 5.01 5.68 -151
2001 sea 122 1014 7 4 7.41 5.65 215
2002 sea 168 1182 4 6 5.67 5.75 -13
2003 sea 8 31 1 1 -0.50 5.57 -49
2004 sea 58 333 1 3 3.59 6.07 -144
2005 cle 333 2321 11 12 5.68 5.79 -36

When the numbers show that Leaf was "better" than Dilfer, here's what I think that really means: If you had to take the career (regular season) numbers of either Ryan Leaf or Trent Dilfer for your favorite team, you'd choose Leaf.

It shouldn't surprise anyone that Leaf's two worst seasons were worse than anything Dilfer ever did. Similarly, Dilfer had the best five seasons. But what would you rather have to deal with for your favorite team: Two all-time abysmal years and one bad year over the next three seasons? Or twelve years of QB play that goes like this: two really bad years, two bad years, three well below average years, four average years and one good year? It's pretty close, but I'd probably choose the former. And I think that's exactly what the numbers are saying.

It's always going to be tough to compare the careers of players with different career lengths. Tony Banks (-558 value added) and Alex Smith (-565 value added) rank next to each other on the all-time list. This feels about right. It's tough to say whether I'd rather watch h one absolutely abysmal season (Alex Smith, 2005) or two really bad years, six roughly average or nondescript seasons (three of which when Banks had 102, 2 and 25 attempts) and one good year (1999, +143 value added). So when you think of the system that way, it does a pretty good job. And I think it will do very well when we consider the best QBs of all time, because we're going to want to know who would we most want to have on our favorite team.

  • The league average is too high of a baseline.

This one has some merit, too. It was very easy and relatively non-controversial to use league average data. Using a replacement level baseline might be better, but it's not clear exactly what that would be. Maybe the 24th best QB, or the 32nd...or some other number. I'm open to suggestions, but I don't really have a problem using such a "demanding" baseline as league average. This will rightly penalize players who hang on too long or were miserable for a few years before turning things around.

  • Adjusted Yards per Attempt isn't a good stat, doesn't correlate well to wins, or is biased against players in the West Coast Offense.

This criticism isn't a good one. I'll put this on my to-do list for the blog, but I know I've heard that Yards/Attempt For and Yards/Attempt Allowed correlates very, very well with team winning percentage. I'm confident that Adjusted Yards per Attempt is an upgrade on yards per attempt, so I strongly believe that this is the right metric.

As for favoring one system or another, I'd also disagree with that. Completion percentage favors QBs in the WCO. Yards per completion favors QBs in vertical offenses. Yards per attempt is the best of both worlds, and it shouldn't really favor either.

Adjusted yards per attempt probably hurts the players in vertical offenses more, because of the high penalty for interceptions. But INTs are very costly to a team (check out the winning percentages of teams that lose the turnover battle), so I'm ok with this. In short, a good QB will have a good adjusted Y/A, regardless of system. For what it's worth, I think quarterback Rating cuts way too much the other way; it disproportionately rewards QBs with very high completion percentages.

The Best QB of All Time

Let's get to the good stuff. Here are the top single seasons by a QB since 1970.


Name Year Tm Att AY/A Value Added
Peyton Manning 2004 ind 497 9.25 1581
Dan Marino 1984 mia 564 8.51 1566
Kurt Warner 1999 stl 499 8.37 1367
Daunte Culpepper 2004 min 548 8.42 1286
Bert Jones 1976 bal 343 8.57 1270
Steve Young 1994 sfo 461 8.39 1222
Steve Young 1992 sfo 402 8.46 1184
Joe Montana 1989 sfo 386 8.86 1178
Randall Cunningham 1998 min 425 8.46 1131
Ken Anderson 1975 cin 377 7.65 1116
Kurt Warner 2001 stl 546 7.69 1115
Boomer Esiason 1988 cin 388 8.30 1071
Ken O'Brien 1985 nyj 488 7.74 1056
Warren Moon 1990 hou 584 7.59 1028
Steve Young 1993 sfo 462 7.78 1010
Dan Fouts 1981 sdg 609 7.17 1009
Steve Young 1998 sfo 517 7.72 994
Joe Montana 1984 sfo 432 8.01 984
Donovan McNabb 2004 phi 469 8.16 978
Chris Chandler 1998 atl 327 8.76 969
Roger Staubach 1979 dal 461 7.29 965
Neil Lomax 1984 stl 560 7.45 964
Peyton Manning 2003 ind 566 7.26 956
Peyton Manning 2005 ind 453 7.90 956
Ken Anderson 1981 cin 479 7.50 953
Mark Rypien 1991 was 421 7.95 952
Jeff Garcia 2000 sfo 561 7.38 952
Rich Gannon 2002 oak 618 7.28 948
Joe Theismann 1983 was 459 7.64 921
Steve Young 1997 sfo 356 8.28 916
Ken Anderson 1974 cin 328 7.31 915
Steve McNair 2003 ten 400 7.85 913
John Brodie 1970 sfo 378 7.22 911
Brett Favre 1995 gnb 570 7.38 907
Steve Deberg 1990 kan 444 7.87 904
Steve Beuerlein 1999 car 571 7.22 904
Roger Staubach 1971 dal 211 8.78 900
Kurt Warner 2000 ram 347 8.15 858
Daunte Culpepper 2000 min 474 7.48 855
Roger Staubach 1977 dal 361 6.63 839
Ron Jaworski 1980 phi 451 7.23 834
Trent Green 2003 kan 523 7.15 828
Jeff George 1997 oak 521 7.30 827
Ken Stabler 1976 rai 291 7.70 826
Jim Everett 1989 ram 518 7.40 825
Peyton Manning 2000 ind 571 7.12 825
Brian Sipe 1980 cle 554 6.86 823
Boomer Esiason 1986 cin 469 7.32 812
Trent Green 2005 kan 507 7.36 801
Roman Gabriel 1973 phi 460 6.32 797

Here were my general reactions after scanning the list.

  • Manning and Marino's record breaking seasons should top any list. I was worried this system might penalize them a bit, at least relative to one where passing TDs are more highly valued, but they still stand out a bit from the pack. There aren't any surprises at the top (which is probably a good thing ); Warner and Culpepper had two phenomenal seasons that should be in any discussion like this. Bert Jones' season is barely remembered, but it was undoubtedly one of the best of all time. I'm glad to see Chris Chandler's 1998 season on the list, because I always thought it was one of the best seasons that no one ever talked about.
  • Nine of the top 33 seasons were by 49ers QBs, eight of them during the WCO years. I'm very confident that Adjusted Yards per Attempt doesn't penalize WCO QBs. While this might make you think it unfairly benefits them, I don't think I'd go that far either. So many NFL teams copy the WCO because it's successful. The 49ers had a great team for about 20 years, and the West Coast Offense was a big reason why they had so much success.
  • Five years from Steve Young, four (and counting) from Peyton Manning, and three each from Kurt Warner, Ken Anderson, and Roger Staubach on the list. Amazingly enough, only one year from Dan Marino and ... none from Tom Brady. Marino's second best season, 1986, ranks 88th. Brady's best year, last season, ranks 85th.
  • Four QBs reside on both the top and bottom "50 seasons since 1970" lists: Brett Favre, Peyton Manning, Mark Rypien and Boomer Esiason. Vinny Testaverde just missed the cut: his 1998 season was the 51st best of all time.

That sets the stage for what I promised yesterday: the list of the greatest QBs of all time. All we've got to do is sum up every season of each player's career, and away we go.


Player Name Value Career Attempts
Steve Young 7103 4149
Dan Marino 6752 8358
Joe Montana 6634 5391
Roger Staubach 5286 2911
Ken Anderson 5135 4475
Dan Fouts 5017 5604
Peyton Manning 4927 4333
Trent Green 3788 3329
Kurt Warner 3487 2340
Fran Tarkenton 3401 3445
John Elway 3155 7250
Bob Griese 3116 2491
Warren Moon 2908 6823
Jim Kelly 2885 4779
Brett Favre 2672 7612
Daunte Culpepper 2619 2609
Boomer Esiason 2592 5205
Mark Brunell 2429 4334
Phil Simms 2368 4647
Neil Lomax 2364 3153
Bert Jones 2295 2551
Craig Morton 2096 3201
Steve McNair 1973 3871
Rich Gannon 1939 4206
Greg Landry 1845 2092
Dave Krieg 1689 5311
Tom Brady 1614 2548
Matt Hasselbeck 1603 2205
Randall Cunningham 1593 4289
Steve Beuerlein 1583 3328
Marc Bulger 1552 1518
Ken O'Brien 1524 3602
Bernie Kosar 1501 3365
Neil O'Donnell 1499 3229
Jim Hart 1482 4183
Jeff George 1456 3967
Terry Bradshaw 1453 3901
Billy Kilmer 1440 2028
Troy Aikman 1439 4715
Joe Theismann 1429 3602
Jeff Garcia 1248 2785
Brad Johnson 1207 3797
Danny White 1203 2950
Len Dawson 1133 1344
Ben Roethlisberger 1105 563
Roman Gabriel 1045 2267
Sonny Jurgensen 1041 736
Jim Everett 1010 4923
Chad Pennington 1008 1174
Steve Bartkowski 970 3456

  • Steve Young tops this list, as he probably should. I'm not sure whether it's extra impressive that he did it despite having less than half as many career attempts as Marino; but I'm positive it's impressive when you think of where he stood when he left Tampa Bay: -417. While lots of Young's success was due to Jerry Rice, a favorite of the PFR blog last week, consider that this list totally ignores rushing statistics. Young ran for 4,239 yards and 43 TDs. Dan Marino ran for 87 yards. Joe Montana had 1676 rushing yards and 20 TDs. Despite being one of the best running QBs of all time, Steve Young might have been the best quarterback in NFL history judged solely on his passing statistics.
  • Seeing Roger Staubach and Bob Griese so high on this list was a big surprise. Not because they weren't great QBs, but I figured their low number of pass attempts would hurt them. It didn't, and I'd consider that a very good thing. It sure makes me wonder what sort of numbers Staubach could have ended up with if he hadn't missed the prime of his career because of his Navy obligations. Staubach's first 2,000 yard season (in his second season as a starter) came at the age of 31. Had he never entered the Navy, he might have ended up as the greatest player that ever played. In 1979 he had the 21st greatest passing season on this list, and he did it at 37 years old. And yeah, he could run a bit too.
  • It's amazing that Ken Anderson can't get into the Hall of Fame. He retired as the sixth-all time leading passer, and passed for 300 yards and scored three TDs in the Super Bowl. He's won an MVP, holds the single-season completion percentage record, made four Pro Bowls and led the NFL in QB rating four times. I wonder if it would have helped him had the 1982 strike never happened; Anderson had another magnificent season, but it was only over nine games. Mark Moseley, yes the place-kicker Mark Moseley, won the NFL MVP that year. Over a sixteen game season, Anderson could have won the award, and back to back titles would have probably been enough to send him to Canton. As for my list, he very likely would have passed Staubach and ranked fourth all-time, and could have been the only quarterback besides Steve Young with four seasons of over 900 value added points.
  • Peyton Manning's going to end his career as the best QB of all-time. Maybe not in the eyes of everyone (at least not unless or until he wins a Super Bowl), but he should rank first on this list. Manning's past three seasons have added 3,493 value points to the Colts. That means you would have preferred Peyton Manning quarterbacking your favorite team for the last three years, than selecting all but seven other NFL quarterbacks for their entire careers. Wow. Steve Young's best three years added 3,416 points of value, but they were not consecutive.
  • I'm not one of those "Retire, Brett" people... but Favre's last seven seasons have added "only" 367 value points. That's roughly 54 points a year. To put that in perspective, in 2005 Jay Fiedler added 42 points of value and Jim Sorgi added 70 points. But to Favre's credit, last year was his first below average season since 2000.
  • It's a bit surprising to see Terry Bradshaw -- one of two QBs to ever win four Super Bowls -- so low on that list. But during his first five years, he accumulated -1,361 points; during his last nine years, he added 2,814 points of value. The other guy to win four Super Bowls ranked third on the list in part because he was never very bad. While Young, Marino and Staubach and Ken Anderson all had years of worse than -100, Montana's worst season was -13 as a rookie.
  • Dan Fouts is often cited as an example that it takes awhile for a QB to become very good. After all, he had more INTs than TDs each of the first five seasons of his career. But guess what? He added 341 points of value during those years, and 630 points if you exclude his rookie season. He was better than average during his second year in the league, and stayed that way for twelve seasons. His last two years hurt him a bit, though. Without them, he'd be ranked ahead of Roger the Dodger.

There's lots of interesting stuff on the career list, but we're going to look at things at a slightly different angle tomorrow.

65 Comments | Posted in General, History

The worst QB of all time?

Posted by Chase Stuart on September 5, 2006

If you ask a group of football fans to name the worst QB in NFL history, you’ll probably hear Ryan Leaf’s name a lot. Leaf once went 1/15 for four yards and two INTs in a game in 1998, and it's hard to argue that many have looked worse than Leaf. But for the amount of time football fans spend arguing over the best QB ever (and I’ll add to that tomorrow), we rarely discuss the worst. So today, we’ll try and get to the bottom of that question.

I want to be clear about a few things at the start. If the Patriots signed Paul Tagliabue tomorrow, and made him a QB, he'd undoubtedly be the worst QB ever. But he'd never throw a pass, so there'd be no proof of how bad he really was. That's going to be the problem for most of the really awful QBs; they simply not didn't stick around long enough to make a difference. So when I say "worst QB of all time", what I really mean is "since 1970, the QB that, according to a straight statistical analysis, most disadvantaged the team(s) he played for throughout his career."

To rank the QBs, I'm going to use adjusted yards per attempt. According to Doug Drinen,

[Adjusted yards per attempt is] defined as (passing yards + 10*(TD passes) - 45*(interceptions thrown)) / (pass attempts). It was devised (and the reasoning behind it explained) in a book called The Hidden Game of Football, by Carroll, Palmer, and Thorn.

The top QBs in AY/A last year were Ben Roethlisberger and Peyton Manning. Adjusted yards per attempt isn't perfect, but if I could only use one statistic -- and this study is one of the few times in life where that hypothetical is a reality -- it's the one I'd choose. But knowing a QB's AY/A isn't enough, because I want to analyze every QB since 1970. Here's a look at the league average AY/A since the AFL/NFL merger.

2005	5.79
2004 6.07
2003 5.57
2002 5.75
2001 5.65
2000 5.68
1999 5.63
1998 5.79
1997 5.71
1996 5.54
1995 5.79
1994 5.74
1993 5.59
1992 5.51
1991 5.69
1990 5.83
1989 5.81
1988 5.54
1987 5.73
1986 5.59
1985 5.58
1984 5.73
1983 5.64
1982 5.47
1981 5.51
1980 5.38
1979 5.20
1978 4.69
1977 4.31
1976 4.86
1975 4.69
1974 4.52
1973 4.59
1972 4.85
1971 4.51
1970 4.81

Ryan Leaf played in an era where passing numbers were more impressive than Jim Plunkett did. So I normalized each QB's stats by subtracting the league average AY/A from each QB's AY/A. There's just one more step: multiply that difference by the number of attempts for each quarterback.

Let's run through one quick example. In 1985, Ken O'Brien averaged 7.742 AY/A on 488 passes. The league average was just 5.578 AY/A, so O'Brien was much better than average and had a large number of attempts. His "Value Added" number was 1056. That number doesn't have any significance, of course, until you rank him against other QBs. O'Brien's 1056 ranks as the thirteenth best season of all time.

But we don't care about the best seasons ever. We care about the worst. So here's the list of the 50 worst seasons of all time. Remember, "worst" is just a proxy for his statistical performance relative to the league average in any given year, and in no way is an indication of the QB's talent relative to the abilities of his teammates. A terrible QB with great teammates would look ok in my system, and a good QB with a very bad supporting cast would likely look below average.

Name		Year	Tm	Att	AY/A	Value Added
Jake Plummer 1999 ari 381 2.94 -1025
Joey Harrington 2003 det 554 3.72 -1024
Jon Kitna 2001 cin 581 4.04 -936
Kerry Collins 1997 car 381 3.38 -886
Chris Weinke 2001 car 540 4.05 -865
Vince Evans 1981 chi 436 3.59 -840
Jack Trudeau 1986 ind 417 3.59 -836
A.J. Feeley 2004 mia 356 3.73 -833
Craig Whelihan 1998 sdg 320 3.21 -826
Mark Malone 1987 pit 336 3.28 -823
Kordell Stewart 1998 pit 458 4.06 -794
Jake Plummer 2002 ari 530 4.25 -794
Vinny Testaverde1988 tam 466 3.85 -789
Ryan Leaf 1998 sdg 245 2.59 -786
Gary Marangi 1976 buf 232 1.50 -781
Joey Harrington 2002 det 429 3.95 -771
Drew Bledsoe 1995 nwe 636 4.59 -766
Steve Deberg 1978 sfo 302 2.19 -756
Kyle Orton 2005 chi 368 3.73 -755
Bobby Hoying 1998 phi 224 2.48 -742
Brett Favre 2005 gnb 607 4.57 -736
Joe Ferguson 1983 buf 508 4.19 -734
Archie Manning 1975 nor 338 2.52 -732
Dan Pastorini 1981 ram 152 0.72 -729
Joe Namath 1976 nyj 230 1.78 -709
Boomer Esiason 1992 cin 278 3.03 -690
Joe Kapp 1970 nwe 219 1.68 -685
Mark Rypien 1993 was 319 3.46 -679
Troy Aikman 1989 dal 293 3.51 -673
Jim Zorn 1976 sea 439 3.36 -660
Terry Bradshaw 1970 pit 218 1.79 -660
Ryan Leaf 2000 sdg 322 3.67 -646
Vince Ferragamo 1985 buf 287 3.35 -639
Joe Ferguson 1984 buf 344 3.91 -626
Kyle Boller 2004 bal 464 4.73 -622
Rusty Hilger 1988 det 306 3.56 -609
Jeff Komlo 1979 det 368 3.57 -599
Randy Hedberg 1977 tam 90 -2.29 -594
Peyton Manning 1998 ind 575 4.76 -593
Mark Malone 1986 pit 425 4.20 -592
Bubby Brister 1995 nyj 170 2.39 -579
Dave Brown 1996 nyg 398 4.10 -574
Chuck Long 1987 det 416 4.35 -574
Jim Plunkett 1972 nwe 355 3.24 -570
Kim McQuilken 1976 atl 121 0.17 -569
Jake Plummer 2000 ari 475 4.49 -567
Dan Pastorini 1973 hou 290 2.64 -565
Alex Smith 2005 sfo 165 2.36 -565
Scott Brunner 1983 nyg 386 4.19 -560
Dennis Shaw 1971 buf 291 2.59 -560

That's not a typo next to Randy Hedberg's name. He actually averaged -2.29 AY/A in 1977, which not surprisingly, was the only year he played. But there's only so much damage you can do with 90 attempts, so he didn't hurt his team as much as Jake Plummer did in 1999.

Plummer and Harrington stand alone as the only two QBs to ever produce a negative value in the thousands. What made them stand out was the large number of attempts, especially for Harrington. To play that badly for that long -- and to avoid the bench -- is apparently unprecedented. Not surprisingly, Harrington and Plummer both had their historic seasons when they were considered young players that their respective franchises could build around, so Arizona and Detroit let their inexperienced QBs learn on the fly. Despite both QBs still being active, neither remains with the team that drafted them.

I promised an answer to the question of who is the worst QB of all time. To do this, we need to add the "value added" number for each QB, for each season of his career. Now we can compare Ryan Leaf's miserable few seasons, to the sustained ineptitude of Dan Pastorini. The results are a bit surprising.


Player Name Value Career Attempts
Rick Mirer -2512 2043
Mike Phipps -2447 1799
Joey Harrington -2374 1802
Dan Pastorini -2272 3055
Mark Malone -2146 1648
Jack Trudeau -1780 1644
Trent Dilfer -1719 2952
Kordell Stewart -1676 2358
Ryan Leaf -1560 655
Steve Walsh -1548 1317
Danny Kanell -1540 956
Scott Brunner -1474 1046
Jon Kitna -1451 2837
Randy Wright -1435 1119
Mike Pagel -1415 1509
Kim McQuilken -1413 272
Gary Huff -1343 788
Billy Joe Tolliver -1314 1707
Bobby Douglass -1307 1030
Kyle Boller -1270 981
Kerry Collins -1226 5082
Dave Brown -1218 1634
Craig Whelihan -1212 557
Richard Todd -1194 2967
Marc Wilson -1117 2081
Todd Blackledge -1114 881
Jake Plummer -1098 4033
Vinny Testaverde -1083 6526
Heath Shuler -1049 593
Jack Thompson -1043 845
Chris Weinke -1039 591
Mike Tomczak -1030 2337
Vince Evans -1029 1390
David Woodley -970 1300
Stan Gelbaugh -970 391
Tim Couch -953 1714
Joe Ferguson -946 4519
Dave Wilson -942 1039
Doug Pederson -940 522
Akili Smith -936 461
Bob Avellini -929 1110
Kelly Stouffer -887 437
Mike McMahon -884 515
Vince Ferragamo -864 1615
Bubby Brister -861 2212
Kent Graham -849 1340
David Klingler -835 718
A.J. Feeley -833 524
Jeff Komlo -830 437
Gary Marangi -817 283

So there you have it. Rick Mirer is the worst QB ever. Once again, remember when I say "worst QB ever" I mean Rick Mirer "accumulated more below average statistical production over the course of his career than any QB since 1970." I'm not surprised to see a former high draft pick at the top of this list, because you need to compile a lot of attempts to really be the worst. And Mirer did just that.

Some of our younger readers might be wondering who Mike Phipps is. You're not alone -- I had never heard of him either. I e-mailed PFR blog reader and Browns fan Ace Davis for some insight on Mike Phipps. Here's what Ace told me.

Well, he was bad, that's for sure. Surely no worse than hundreds of quarterbacks who have come and gone over the years, but he was a special disappointment. Art Modell traded away Hall of Fame receiver Paul Warfield, a native Ohioan, to Miami for the right to draft Phipps third overall in 1970. So the price was dear, but that's not Phipps' fault. In fact, without Warfield, the Browns were left with a subpar receiving corps at his disposal.

Also, the Phipps era really coincided with the end of the Browns as a dominant NFL franchise, the beginning of their first terrible decade, and the first-ever ascendancy of the dreaded Pittsburgh Steelers and their QB, Terry Bradshaw, drafted first overall the same year as Phipps. Still, Phipps did establish himself as a starter, and he flashed enough to lead the Browns to the 1972 playoffs, where -- but for his five interceptions -- they would've toppled the perfect Miami Dolphins. The contrast between Warfield's elegance and Phipps' inconsistency was apparent to all in that game.

He was athletic, a good runner, and a decent guy. He just was a terribly inaccurate passer who melted in the crucible of Cleveland's high expectations. Check out his completion percentage and TD/INT ratios, particularly in '74 and '75, the first time since their founding that the Browns had consecutive losing seasons.

The transition from the pressurized, overrated Phipps to a laid-back, unheralded QB with less physical skill but a go-for-broke playmaking ability was slow to develop, but fruitful nonetheless. Brian Sipe became a fan favorite and league MVP, while the Browns somehow managed to salvage a first-round pick from the Bears in exchange for Phipps. They eventually drafted Ozzie Newsome -- another HOF pass-catcher --as a result.

Tomorrow we'll use this same system to analyze the best QBs ever. I'll let you guys comment on some of the surprising names on this list, such as Trent Dilfer (7th worst ever), Jon Kitna (14th) and Vinny Testaverde (28th). Here are Dilfer's career numbers. Note that even in 2000, when his Ravens won the Super Bowl, Dilfer was below the league average in adjusted yards per attempt.


Year Tm Att AY/A Value Added
1994 tam 82 2.11 -298
1995 tam 415 4.83 -399
1996 tam 482 4.41 -548
1997 tam 386 5.88 66
1998 tam 429 5.28 -222
1999 tam 244 5.06 -141
2000 bal 225 5.01 -151
2001 sea 122 7.41 215
2002 sea 168 5.67 -13
2003 sea 8 -0.50 -49
2004 sea 58 3.59 -144
2005 cle 333 5.68 -36

67 Comments | Posted in Statgeekery

Deion Branch and spreading it around

Posted by Doug on September 5, 2006

Of all the interesting stories that will unfold during the next six months, the one that intrigues me most is that of the Deion Branch-less New England Patriots. It's an interesting situation, both on the field and off it.

The Patriots have cap space, but have apparently decided that they'd rather spend that cap space on, well, nothing, than to pay Branch more than what they think he is worth. I understand their position. It's one that might help them in the long run and it's certainly hard to argue with they way they've done business over the last half-decade. But I have to believe that attitudes will change very quickly --- in the media, in the bleachers, and in the locker room --- if and when the Patriots start losing.

I know that The System in New England is more important than any individual, but there has to be a limit to the number of defections it can withstand. The down year in 2005 seems like a hint that they are near that limit. I happen to agree that Branch probably isn't worth what the Jets are pretending they want to pay him, and what Seattle apparently really is willing to pay him. But he is a talented player, and the Patriots might be too close to a tipping point to be able to afford the loss of another talented player.

Now on to the trivial, but interesting, matter of who is going to be catching passes while Branch watches. The Patriots have always been a spread-it-around team, and this year they will probably take that to a whole new level. Just for fun, I created a spread-it-around index and applied it to all teams since 1978. According to my measure, the team with the highest index was the 1989 Chicago Bears. Here were their top 6 pass receivers (by yards):


Ron Morris 486
Dennis Gentry 463
Neal Anderson 434
Dennis McKinnon 418
Wendell Davis 397
James Thornton 392

On the other end of the spectrum is the 2001 Denver Broncos:


Rod Smith 1343
Desmond Clark 566
Dwayne Carswell 299
Eddie Kennison 169
Chris Cole 128
Scottie Montgomery 99

[If you must know how I ranked them, here are the details...

I looked at the top six receivers on each team, and computed a simple average of their receiving yards. Then I took a weighted average of the same six receivers, with the top guy getting a weight of one, the next guy a weight of two, and so on down to the sixth guy, who gets a weight of six. Divide the weighted average by the simple average (and multiply by 100 to make it easier on the eyes) and you've got the index. The maximum possible index --- which would be attained if all six players had the same receiving yardage --- is 100.]

Last year, the teams that spread it around most and least were Minnesota and Washington, respectively. New England had the 8th-highest index in the league last year. Here are the top 10 and bottom 10 teams from 1978--2005, and you can see the full list right here.


chi 1989 96.2
Ron Morris 486
Dennis Gentry 463
Neal Anderson 434
Dennis McKinnon 418
Wendell Davis 397
James Thornton 392

kan 1995 94.0
Willie Davis 527
Webster Slaughter 514
Lake Dawson 513
Todd McNair 501
Keith Cash 419
Kimble Anders 349

was 1997 93.9
Leslie Shepherd 562
Michael Westbrook 559
Henry Ellard 485
Jamie Asher 474
Brian Mitchell 438
Larry Bowie 388

sdg 1984 93.0
Charlie Joiner 793
Pete Holohan 734
Bobby Duckworth 715
Wes Chandler 708
Kellen Winslow 663
Eric Sievers 438

atl 2001 91.6
Terance Mathis 564
Tony Martin 548
Shawn Jefferson 539
Brian Finneran 491
Bob Christian 392
Alge Crumpler 330

ind 1985 91.3
Wayne Capers 438
Matt Bouza 381
Pat Beach 376
Ray Butler 345
Mark Boyer 274
George Wonsley 257

det 2003 90.9
Az-Zahir Hakim 449
Mikhael Ricks 434
Bill Schroeder 397
Shawn Bryson 340
Scotty Anderson 325
Cory Schlesinger 247

mia 1992 90.9
Mark Duper 762
Mark Clayton 619
Keith Jackson 594
Tony Martin 553
Bobby Humphrey 507
Tony Paige 399

bal 2004 90.6
Travis Taylor 421
Kevin Johnson 373
Randy Hymes 323
Todd Heap 303
Clarence Moore 293
Daniel Wilcox 219

ari 2002 90.6
David Boston 512
Marcel Shipp 413
Frank Sanders 400
Jason McAddley 362
Freddie Jones 358
MarTay Jenkins 250

[. . .]

hou 1988 61.5
Drew Hill 1141
Ernest Givins 976
Curtis Duncan 302
Willie Drewrey 172
Leonard Harris 136
Alonzo Highsmith 131

was 1984 61.3
Art Monk 1372
Calvin Muhammad 729
Clint Didier 350
Charlie Brown 200
Don Warren 192
Mark McGrath 118

pit 1996 61.2
Charles Johnson 1008
Andre Hastings 739
Yancey Thigpen 244
Mark Bruener 141
Jerome Bettis 122
Erric Pegram 112

det 1995 60.9
Herman Moore 1686
Brett Perriman 1488
Johnnie Morton 590
Barry Sanders 398
David Sloan 184
Ron Hall 81

det 1996 60.1
Herman Moore 1296
Brett Perriman 1021
Johnnie Morton 714
Barry Sanders 147
Pete Metzelaars 146
David Sloan 51

pit 1997 59.9
Yancey Thigpen 1398
Charles Johnson 568
Courtney Hawkins 555
Will Blackwell 168
Mark Bruener 117
Jerome Bettis 110

was 2005 59.7
Santana Moss 1483
Chris Cooley 774
David Patten 217
Clinton Portis 216
James Thrash 194
Robert Royal 131

jax 1999 59.7
Jimmy Smith 1636
Keenan McCardell 891
Kyle Brady 346
Damon Jones 221
Reggie Barlow 202
Tavian Banks 137

stl 1996 58.9
Isaac Bruce 1338
Eddie Kennison 924
Harold Green 246
Ernie Conwell 164
Jermaine Ross 160
Aaron Laing 116

den 2001 58.0
Rod Smith 1343
Desmond Clark 566
Dwayne Carswell 299
Eddie Kennison 169
Chris Cole 128
Scottie Montgomery 99

12 Comments | Posted in General

Page 1 of 212