1993 DVOA Ratings and Commentary

1993 DVOA Ratings and Commentary
1993 DVOA Ratings and Commentary
Photo: USA Today Sports Images

by Aaron Schatz

Let us travel now to football's distant past, a forgotten time known as "1993." It was a long-ago era when professional football was still played in antiquated small markets like "Los Angeles" and "Anaheim." Peyton Manning and Tom Brady were still in high school. The biggest album in America was Pearl Jam's Vs. and you could listen to it through a strange black strip known as "cassette tape." Dinosaurs still roamed the earth, by which I mean that Hank Stram was still on NFL broadcasts. FOX was a network of mostly UHF stations that nobody could ever imagine broadcasting professional football. And in these long ago days, two giants (who were not Giants) cast their shadows across the National Football League: Dallas and San Francisco. These two teams, as expected, top the newly-revealed Football Outsiders DVOA ratings for the 1993 season.

Of course, now that we go back and look at things with DVOA and other advanced stats, Dallas and San Francisco may not have been quite as dominant as we all thought -- at least, not during the regular season. As I noted on the site a few months ago, the standings were much more compact in the mid-nineties. No team in 1993 won more than 12 games, and only one team, Cincinnati, lost more than 12. The DVOA ratings are more compact as well; San Francisco and Dallas would have finished sixth and seventh in DVOA in 2009. Strangely, this is now the third season of play-by-play we have broken down during the Dallas/San Francisco heyday of the nineties, and in all three seasons, the lower of the two teams in total DVOA was the team that actually won the Super Bowl. The difference between Dallas and San Francisco in 1993 is less than one percentage point, but the 49ers had a terrible record in close games and ended up 10-6. San Francisco had four losses by a field goal but also won games by scores of 42-7 (over New Orleans), 55-17 (over Detroit), and 44-3 (over the New York Giants in the freakin' playoffs!!!). The Pythagorean projection based solely on points scored and allowed puts both Dallas and San Francisco slightly over 12 wins.

We remember the early nineties as a time still dominated by the NFC, but actually, the two conferences were pretty much even by 1993. The difference came in how the top of each conference was shaped. The NFC was dominated by two teams, while the AFC had a lot of good teams without any great ones. After San Francisco and Dallas, the next five teams in DVOA were AFC teams, although none of those teams represented the AFC in the Super Bowl. Instead, Buffalo went for a fourth straight year, after a 12-4 season that was very good by conventional stats but not particularly strong according to DVOA. Poor red-zone performance (23rd on offense) was the main reason behind Buffalo's poor DVOA. They didn't play a particularly tough schedule, and while they had bad fumble luck, they also had good fumble luck. The Bills recovered only five of 21 fumbles on offense but 20 of 30 fumbles on defense.

Here are the Football Outsiders team efficiency ratings for 1993, measured by our proprietary Defense-adjusted Value Over Average (DVOA) system that breaks down every single play and compares a team's performance to the league averaged based on situation and opponent in order to determine value over average.  (Explained further here.)

DVOA represents adjusted statistics. OFFENSE and DEFENSE DVOA are adjusted for opponent quality and to consider all fumbles, kept or lost, as equal value. SPECIAL TEAMS DVOA is adjusted for type of stadium (warm, cold, dome, Denver) and week of season. NON-ADJ TOTAL VOA does not include these adjustments. DVOA is a better indicator of team quality. VOA is a better indicator of actual wins. WEIGHTED DVOA gives a stronger consideration to games late in the season. Remember that, as always, defense is better when it is NEGATIVE.

1 SF 25.6% 26.9% 4 10-6 26.2% 1 -1.6% 19 -2.2% 22
2 DAL 25.0% 28.2% 2 12-4 16.5% 2 -5.2% 17 3.3% 7
3 KC 21.5% 20.0% 5 11-5 7.1% 4 -16.0% 4 -1.5% 19
4 SD 18.4% 26.9% 3 8-8 6.5% 5 -9.7% 11 2.3% 9
5 HOIL 17.1% 34.4% 1 12-4 -7.0% 15 -22.1% 2 2.0% 10
6 DEN 15.9% 15.4% 6 9-7 9.3% 3 -8.2% 12 -1.6% 20
7 PIT 14.4% 8.0% 11 9-7 -7.0% 16 -24.0% 1 -2.6% 23
8 NYG 13.4% 12.3% 7 11-5 5.9% 6 -6.6% 15 1.0% 12
9 PHI 11.0% 8.2% 10 8-8 1.1% 8 -14.7% 6 -4.8% 25
10 GB 10.9% 9.9% 9 9-7 -9.0% 17 -15.3% 5 4.5% 6
11 LARD 10.4% 4.0% 14 10-6 -1.1% 10 -4.6% 18 6.9% 1
12 BUF 9.4% -1.5% 18 12-4 -3.4% 13 -12.3% 8 0.5% 13
13 NYJ 7.9% 0.4% 16 8-8 0.3% 9 -13.6% 7 -6.0% 27
14 MIN 7.4% 11.8% 8 9-7 -5.0% 14 -19.0% 3 -6.7% 28
15 MIA 3.9% 2.5% 15 9-7 5.7% 7 1.8% 21 0.0% 15
16 CLE1 0.1% 6.2% 13 7-9 -12.6% 18 -8.1% 13 4.6% 4
17 ARI -1.9% 6.5% 12 7-9 -1.5% 11 1.8% 22 1.5% 11
18 NO -4.0% -23.6% 24 8-8 -19.6% 25 -10.0% 10 5.6% 3
19 DET -6.1% -1.4% 17 10-6 -16.7% 22 -6.0% 16 4.6% 5
20 CHI -7.0% -13.7% 22 7-9 -20.3% 26 -10.9% 9 2.4% 8
21 SEA -7.4% -11.2% 20 6-10 -13.0% 20 -7.4% 14 -1.8% 21
22 ATL -13.5% -3.4% 19 6-10 -12.8% 19 6.6% 25 5.8% 2
23 WAS -20.3% -24.6% 26 4-12 -15.0% 21 4.1% 23 -1.2% 18
24 LARM -21.0% -28.4% 27 5-11 -2.4% 12 13.8% 28 -4.8% 24
25 NE -24.0% -12.1% 21 5-11 -17.6% 23 1.3% 20 -5.1% 26
26 CIN -29.9% -24.3% 25 3-13 -19.3% 24 10.4% 26 -0.2% 16
27 TB -33.1% -18.4% 23 5-11 -26.1% 28 5.9% 24 -1.0% 17
28 IND -38.6% -37.3% 28 4-12 -26.1% 27 12.8% 27 0.2% 14

DVOA for 1993 is now listed in the stats pages:

It was the Houston Oilers, not the Cowboys or 49ers, who went into the 1993 playoffs as the hottest team in the league according to weighted DVOA. If NFL ever does a third America's Game season, they should follow up the Super Bowl Champions and "The Missing Rings" with a set of "The Greatest Rollercoaster Seasons in NFL History." This team would get the first episode. The Oilers were a veteran team with one last chance to win. Thirteen of their 22 starters were impending free agents, and nobody knew quite what that meant since this was still early in the history of NFL free agency. To help them get over the hump, the Oilers hired Buddy Ryan as defensive coordinator; he had been out of the league since his last year as Philadelphia head coach in 1990.

The Oilers started the season 1-4 and actually benched Warren Moon after he threw four picks in an 18-17 loss to San Diego in Week 3. Conventional wisdom said the team was suffering a massive hangover from their 38-35 loss to Buffalo in the last postseason, the infamous Frank Reich Comeback. The final loss of that 1-4 start was a 35-7 debacle in Buffalo in Week 6. Then the Oilers turned it around and won 11 straight. Ryan's defense gave up more than 17 points only once in that 11-game streak, when the Oilers beat Cleveland 27-20 in Week 12. The streak was tinged with both tragedy and comedy. When the Oilers were 9-4 in mid-December, backup defensive tackle Jeff Alm committed suicide, distraught because he had killed a friend in a drunk-driving accident. In the final game of the season, a 24-0 blowout of the Jets, Ryan attacked offensive coordinator Kevin Gilbride on the sidelines because he blamed the run-and-shoot philosophy for two late-game defensive injuries that came when the offense kept throwing the ball instead of just running out the clock. (And for the few who have not heard this story, let me make clear that "attacked" did not mean "yelled at." It meant "punched in the face." Hooray for YouTube.)

Do you believe in postseason momentum, that the hottest teams in the second half generally win big in January? What about the idea that teams that suffer through tragedy come together for emotional victories? Well, the 1993 Houston Oilers do a lot to disprove both of these theories. Riding that 11-game win streak, they went splat and lost their first playoff game at home to Kansas City -- a team with less momentum but the best DVOA in the conference over the full season. The next year, a bunch of players and coaches were gone, including Moon to Minnesota and Ryan to Arizona. The 1994 Oilers were the worst team in the league by DVOA.

The 1993 Oilers had a run-and-shoot offense that was fabulous at running (fourth in rushing DVOA) but not so great at the shooting (18th in passing). Part of the problem was an off-year for Moon (-8.9% passing DVOA) and part of the problem was that the wide receivers couldn't hold onto the ball (Webster Slaughter fumbled five times, Haywood Jeffires four -- by comparison, no wide receiver had more than three fumbles in 2009). The big change for the Oilers came not when they were 1-4 but when they were 4-4, because that's when starting running back Lorenzo White injured his hamstring and went out for the year. White averaged 3.5 yards per carry and ranked 34th among running backs with -11.6% DVOA. After White was done, 1991 eighth-round pick Gary Brown took over as the starter and was phenomenal. Behind the exact same offensive line, in the same run-and-shoot scheme, Brown (including his carries as a backup in the first half of the season) averaged 5.1 yards per carry and led all running backs with 28.4% DVOA, nine percentage points higher than any other back with at least 100 carries. He was third in rushing DYAR behind Rookie of the Year Jerome Bettis and veteran Emmitt Smith, and the gap between that top three and everyone else is pretty big.

I'm not sure people now remember just how weird Bettis' early career numbers are. Chosen tenth overall in the 1993 draft, Bettis was Rookie of the Year with 4.9 yards per carry. The very next year, he dropped to 3.2 yards per carry, gaining 200 fewer yards with 25 more carries. He was terrible again in 1995, so the Rams drafted Lawrence Phillips and traded him to the Steelers for a third-round pick. Phillips was a colossal bust and Bettis returned to glory in Pittsburgh with 4.5 yards per carry. So for four straight seasons, Bettis's rating in rushing DYAR goes from first to last to near last and then back to first.

Emmitt Smith, meanwhile, will probably end up as the best running back in DYAR history once we've got his whole career measured by our stats. Smith is within mere decimal points of leading the league in rushing DYAR for three straight seasons from 1993-1995; he was also fourth in 1998 and third in 1999.

The top running backs in receiving DYAR were both Miami Dolphins: Terry Kirby with 321 receiving DYAR and Keith Byars with 244 receiving DYAR. The 1993 Dolphins apparently ran an offense based entirely around the H-back.

One Hall of Famer who had a surprisingly poor 1993 was Barry Sanders. Sanders averaged 4.6 yards per carry, but in 1993 his boom-and-bust style led to too many busts and not enough booms. That season was his career-low with just three touchdowns. He was 15th among running backs with 60 rushing DYAR, and 34th (among 42 backs) with a 44 percent success rate. He also had a terrible year as a receiver, a career-low 5.7 yards per reception and -51 receiving DYAR.

Who was the top quarterback of 1993? That depends on your measure of choice. John Elway led the league with 1,200 passing DYAR, but part of the reason for his big season is that he had more than 600 pass plays, roughly 100 more than Steve Young (second with 1,132 passing DYAR) and 180 more than Troy Aikman (third with 1,077 passing DYAR). Young leads the league in total DYAR if we add passing and rushing value together, since he led all quarterbacks with 146 DYAR. (Young wasn't just a great scrambler, he was an effective one, ranking first or second in rushing DYAR for quarterbacks in every season from 1993-1996 as well as 1998). Aikman had an awesome 69.5 percent completion rate and the best DVOA rating of any quarterback with at least 150 passes that year.

One of the interesting things about quarterback stats in 1993 is how many young quarterbacks are down near the bottom of the league. Forty-five quarterbacks had at least 100 passes that season. Number one overall pick Drew Bledsoe was 30th with -95 DYAR. Number two overall pick Rick Mirer was 41st with -390 DYAR. Brett Favre, in his third year and second season as a starter, was 29th with -70 DYAR. David Klingler, in his second year and first season as a starter, was 37th with -271 DYAR. Going back in time, would the small difference between Favre/Bledsoe and Klingler/Mirer been enough to tell us which pair was headed for success and which pair was headed for failure? Favre's drop in performance from 1992 to 1993 actually reminds me of another young quarterback who regressed in his second season as a starter, then took a big leap forward and developed into a Super Bowl champion: Drew Brees.

  Drew Brees, 2002-2004   Brett Favre, 1992-1994
  GS Comp Att Yds TD INT Yd/Att C% xxx GS Comp Att Yds TD INT Yd/Att C%
Year 1 16 320 526 3284 17 16 6.24 60.8% xxx 13 302 471 3227 18 13 6.85 64.1%
Year 2 11 205 356 2108 11 15 5.92 57.6% xxx 16 318 522 3303 19 24 6.33 60.9%
Year 3 15 262 400 3159 27 7 7.90 65.5% xxx 16 363 582 3882 33 14 6.67 62.4%

Who were the best wide receivers of 1993? Well, the top two are the same players who ranked 1-2 in 1994 and 1-2 in 1995: Jerry Rice and Michael Irvin. As in 1995, Rice had the better traditional stats but Irvin surprisingly comes out ahead according to FO stats. There are two reasons for this. First, Rice had three fumbles and Irvin had zero. That's a career-long issue: In terms of fumbles per reception, Rice fumbled roughly twice as often as Irvin. Second, Irvin was much better than Rice at getting consistent mid-length gains. Here's a breakdown of their receptions by yards gained:

Split of Receptions by Yards Gained, 1993
Yards Gained J.Rice M.Irvin
5 yards or less 16% 6%
6-10 yards 24% 19%
11-15 yards 26% 38%
16-20 yards 12% 23%
21-25 yards 6% 7%
26+ yards 15% 8%

It would be nice if we had "air yards" tracked for each pass from 1993-1995, to see how much this difference had to do with the usage patterns for each receiver, but unfortunately, we don't have that data for any season before 2005. This consistent performance was enough to put Irvin ahead of Rice in 1993 DYAR even though Rice had better numbers on third down as well as in the red zone.

Maybe the best stats of any wide receiver in 1993 belonged to Rice's teammate John Taylor. Rice and Taylor each started all 16 games, but the 49ers threw to Rice twice as often as they threw to Taylor. Nonetheless, Taylor took advantage of the constant double-coverage on Rice to post a 49.7% DVOA and catch 76 percent of intended passes. Compare Taylor to Wes Welker in 2009. Both players had a catch rate of 76 percent. Welker averaged 11.0 yards per reception. Taylor averaged 16.8 yards per reception.

The worst wide receiver of 1993 was New England's Greg McMurtry, the one-time first-round pick of the Red Sox and Brockton High star who spurned Fenway Park for a Michigan football scholarship, but still got the opportunity to come home to play professionally when the Pats took him in the third round of the 1990 draft. McMurtry was not a great football player, but he gets points for being better than Drew Henson at both football and decisiveness.

Shannon Sharpe was the top tight end of 1993, and by a huge margin. He had 323 DYAR, the best season of any tight end in the nineties and one of only five tight end seasons above 300 DYAR. (The others: Antonio Gates in 2004 and 2009, and Tony Gonzalez in 2004 and 2000.) Sharpe had nearly twice as many DYAR as the tight end in second place, Brent Jones of the 49ers. Sharpe's catch rate of 74 percent was second among tight ends with at least 25 passes, behind Steve Jordan -- Brown University alert! -- leading to totals of 81 receptions, 995 yards, and nine touchdowns.

Some other comments on the 1993 season:

  • 1993 was the year the NFL experimented with two bye weeks per team during the season, which completely screws up our databases and required us to re-write the code that produces our stats and tables. Most people felt this went too far in breaking up the rhythm of the season, and the NFL went back to one bye per team in 1994.
  • 1993 was also the last year where teams kicked off from the 35-yard line rather than the 30. This required a completely new set of kickoff baselines for the special teams ratings.
  • The AFC West was really strong in 1993, one of the strongest divisions in NFL history. Kansas City, San Diego, and Denver all ranked in the top six, with Oakland 11th and last-place Seattle 21st. The Chargers went 8-8 because of a hard schedule and bad luck in close games; their last five losses all came by a touchdown or less.
  • Tampa Bay had the hardest schedule in the league -- yes, they played the AFC West in interconference play -- with the five AFC West teams ranked two through six.
  • Detroit had the easiest schedule in the league, which helped the Lions finish 10-6 despite ranking 19th in DVOA.
  • The top three defenses of 1993 (Pittsburgh, Houston, and Minnesota) were also the three most inconsistent defenses of 1993. Making the case for great coordinators, those defenses were run by Dom Capers (with Dick LeBeau as an assistant), Buddy Ryan, and Tony Dungy.
  • Memorable players making their first-ever appearance in our DYAR stats: Phil Simms, Gary Anderson (running back version), Neal Anderson, Roger Craig, Drew Hill, Mark Clayton (older Miami version), Willie Gault, James Lofton, and the corpse of the once-great Eric Dickerson (26 awful carries for 91 yards with the Atlanta Falcons).
  • Since Isaac Bruce was a rookie in 1994, Brett Favre is now the only "skill position" player who is listed in all 17 years of Football Outsiders stats. I believe the only other players to play every year from 1993 through 2009 are Jeff Feagles and Matt Stover.

We've made a number of other updates to our database today, along with posting the standard stats pages for 1993. Most player pages for "skill positions" have now been updated with stats from 1993 (for premium members) and 2009 (for free). Game charting data is not listed for 2009 yet, and neither are penalty totals. That data will be added in the near future, along with offensive linemen and defensive players from 1993 and 2009. In addition, the 1993 stats should be available in the Football Outsiders Premium database sometime this evening, except for the "DVOA Week-by-Week during the season" view, which I'll be able to run sometime in the next few days.

A big thanks to all the readers who participated in the 1993 transcription project: Stan Buck, Michael Bonner, Jeremy Billones, Kevin Mayo, Moshe Dachs, and the ridiculously prolific Jeremy Snyder. Jeremy not only did more games than anyone else, he also took care of translating the gamebooks for various teams whose official scorers were using non-standard play descriptions back in the pre-Internet Stone Age.

At our current rate, we're about 12 months away from having 20 full years of DVOA and DYAR; there's a good chance we can finish 1992 this summer, thanks in large part to Jeremy Snyder's amazing output of transcriptions, and then hit 1991 during the 2010 season. If you would like to help with the transcription project -- be aware, it isn't the most exciting work, but we're always looking for more help -- please drop us a line at Contact Us with "PBP transcription" as the subject.


71 comments, Last at 22 Jul 2010, 12:06pm

#1 by Dean // Feb 16, 2010 - 4:19pm

Gotta love that the Raiders are abbreviated LARD. Now I'll go back and read the rest.

Points: 0

#20 by young curmudgeon // Feb 16, 2010 - 9:02pm

And if you don't want to cook with "LARD," you can substitute "HOIL." Either way, you can make tasty doughnuts.

Points: 0

#41 by Dean // Feb 17, 2010 - 3:04pm

Mmmm.... Donuts.

(looks around to make sure he isn't trampled by Andy Reid, Rex Ryan, and Charlie Weiss.)

Points: 0

#52 by jebmak // Feb 18, 2010 - 11:26am

One of my favorite LOLs.


Points: 0

#2 by PerlStalker // Feb 16, 2010 - 4:24pm

Ah, the '90s and the Wild, Wild AFC West. I miss those days.

Points: 0

#3 by Tom Gower // Feb 16, 2010 - 4:46pm

Veteran Houston Chronicle scribe John McClain wrote he considered the 1993 Oilers (end of year version) the best team ever not to win the Super Bowl (pre 18-1); high claim, but the Weighted DVOA lends the argument some support. It was indeed a very odd year-you didn't note the Oilers started the year routing the Montana-less Chiefs 30-0, but did note the SD loss (on 6 FGs) and the Week 6 debacle on MNF in Buffalo before recovering to win the last 11. Reminiscent of 2001 that way (0-4 to 11-1), though 1993 was a much better team. 7 fumbles (though only 2 lost, I think) and a 4th quarter collapse left us pondering what might have been.

Gary Brown was somewhat of a revelation, in part for me personally. The Oilers had been a team that invested lots of high draft picks in running backs, spending 1st rounders on Mike Rozier, Alonzo Highsmith, and Lorenzo White over the previous decade, without quite the sort of success you'd expect, and here was an 8th rounder performing well. That it was scheme and rest of the team, rather than being especially good on his part, became apparent the next year when his numbers plummeted with the talent level around him. Still, he's one of a small number of backs in NFL history with 1000 yards on fewer than 200 carries.

The '93 Oilers had the league's highest payroll, in part because they were one of the few teams that refused to believe there actually would be a salary cap, so in addition to the free agents, they ended up trading other players, Moon included, after the season for pennies on the dollar, and were the worst team in the league in 1994. The year-on-year DVOA decline was 47.9%, and if you use the weighted figure for 1993 it's 65.2%. To put that in perspective, that latter figure is more than the difference between the Rams and the Vikings this year.

Points: 0

#9 by dedkrikit (not verified) // Feb 16, 2010 - 6:11pm

This is the first year I remember *really* being into football - and therefore the Oilers. What a heartbreaker that playoff game was. I still have a small collection of Gary Brown cards from that awesome year.

Points: 0

#42 by Tom Gower // Feb 17, 2010 - 3:08pm

Man, I can't believe I forgot to mention Baby-gate, as 1993 was the year RT David Williams skipped a game to attend the birth of his first child. Those of you who haven't yet should also read Mike Tanier's article from a couple years ago on Wilber Marshall and how he ended up on the Oilers for that year (beyond Buddy Ryan pretty much requiring the team pick him up).

Points: 0

#50 by Stravinsky (not verified) // Feb 18, 2010 - 4:33am

As I recall, things were very, very interesting with that 93 Oilers team.

Houston hired Jack Pardee from UH in 1990 and he brought the run-and-shoot to the Oilers. During the 90, 91 and 92 seasons, the team was fairly successful moving the ball and scoring points but they fell short in the playoffs. This led to a lot of grumbling with the fan base who were convinced that the Run-and-shoot passed too much and the Oilers should just run the ball like they did when Earl Campbell played.

In addition, the team itself was constantly in some turmoil or other. Pardee was not a strict disciplinarian and the team had a lot of outspoken players who were not afraid to stir the pot with the press. Relations between the team and the players were often strained especially at contract time. Everything the Oilers did was full of turmoil, the draft, spring camp and holdouts, training camp and holdouts, the games, the bye weeks, hiring front-office personnel - there was constantly some crisis or other going on. In addition, the fan base despised team owner Bud Adams, who was referred to as ‘Bottom Line Bud’ because of his reputation for not spending money.

One of the big things that frustrated everyone was that the team was constantly struggling to sell-out the home games in time so they would be on TV and Adams refused to do anything to help out. So you had the constant specter of the team playing at home and the local fans not being able to see it on TV because the team came up 100 tickets short of a sell-out.

So then came January 1993, the team had been under-achieving, the fans were grumbling, the owner was grumbling, the players were grumbling and then the team crapped out in Buffalo and things went from incendiary to full nuclear meltdown and in the aftermath things went from interesting to comical.

First, Bud Adams forced Pardee to fire Defensive Coordinator Jim Eddy.

Then Adams personally forced Pardee to add a tight end to the roster (John Henry Mills) even though the R&S does not use a tight end. According to several defensive players, the loss in Buffalo was because the team did not have a tight end and, as a result, the defense could not practice and prepare for a tight end on the opposing offense. I am not making this up. According the Adams, the Oilers could not move the ball on the ground in the second half because they did not have a tight-end to help the running game and if they’d had a tight end they would have won. So, by golly, going forward his team was going to have a tight end.

The fan-base, which was already grumbling about the R&S, was now convinced that the Oilers offensive-scheme sucked. Under fire from the press and fan base, Kevin Gilbride came to the defense of his R&S system pointing out how the Oilers were at or near the top of the league in many offensive categories. This only made things worse and convinced everyone that Gilbride was an idiot.

Then, as if this wasn’t enough, Bud Adams hired Buddy Ryan for Defensive coordinator and if I recall correctly he did it without consulting with Pardee. He just hired Ryan and told Pardee that Ryan was DC and that was that.

Ryan came to town and it took him about 15 minutes to get a feel for the city’s mood. Not one to miss a chance to promote himself, Ryan played up to the city’s angry mood by derisively referring to the offense as the Chuck & Duck.

This did not sit well at all with Gilbride and Pardee and when the team started the season in the toilet, Ryan’s comments further inflamed the already angry fan base. The players got into the act and a nice offense vs defense feud was brewing.

A big part of the problem was that opposing teams had learned how to defend the Oilers R&S attack and how to use outside blitzes to get pressure on Moon. The best way to attack a defense blitzing from the edge is to run the ball up the middle which the team tried with little success while the ineffective Lorenzo White was the starting back. When he got hurt and Gary Brown came in, the running attack came to life allowing the Oilers to move the ball and taking pressure off Moon and the passing game.

With the team having success, tensions between the team, the fans and the press eased somewhat though Babygate and the Jeff Alm tragedy served to keep the pot simmering.

Then came the Jets game, Ryan and Gilbride throwing punches and all the animosities that had been simmering below the surface boiled over. The two weeks between the end of the season and the play-off game with the Chiefs were dominated by “The Punch” and preparations for the Chiefs game were affected. When the Oilers took the field for the Chiefs game it was a different, divided team out there that looked nothing like the cohesive unit that had run off 11 straight wins. It was no surprise that they lost.

Points: 0

#4 by Travis // Feb 16, 2010 - 4:56pm

Junior Seau also played every season from 1994 to 2009, if only parts of the last two.

Buddy Ryan punched Kevin Gilbride at the end of the first half, not at the end of the game, mostly because the Oilers tried for extra scores when Ryan thought they should have sat on their 14-0 lead. Right before the punch, Cody Carlson was sacked and fumbled with 0:24 to go, giving the Jets the ball at the Oiler 18, setting the Jets up for a score after they couldn't do anything all half. (In true Jet fashion, they lost a yard in 3 plays and missed a field goal to preserve the shutout.)

Points: 0

#5 by Aaron Schatz // Feb 16, 2010 - 5:04pm

I meant to say "from 1993 to 2009," actually, but yes, Seau also qualifies, that makes four players.

Points: 0

#6 by Mike Y // Feb 16, 2010 - 5:30pm

This is also the year Super Tecmo Bowl is based on!

Points: 0

#12 by Gihyou (not verified) // Feb 16, 2010 - 7:02pm

No, Tecmo Super Bowl is based on the 1990 season, having been released in 1991.

Points: 0

#14 by Mike Y // Feb 16, 2010 - 7:56pm

Ah, I see you are confusing Super Tecmo Bowl for the SNES with Tecmo Super Bowl for the NES. Easy mistake to make. Tecmo Super Bowl is the one where Bo Jackson dominates. Super Tecmo Bowl is the one where Emmitt Smith and Barry Sanders dominates, and the one where you can play the 3-seasons mode. I have both of them on emulators, and I play them way too much.

EDIT: Nevermind, I thought the SNES version had the "Super" in front of its name, but it looks like they have the same name, one is the NES version based on the 1991 season, and the other is the SNES version based on the 1993 season, both called Tecmo Super Bowl.

Points: 0

#7 by MatMan // Feb 16, 2010 - 5:49pm

How in the world do your charters tally games dating back that far? The amount of video you have access to can't be more than a tape of the original TV broadcast.

Points: 0

#8 by Vincent Verhei // Feb 16, 2010 - 6:06pm

We don't have game charting for that era. We do have the official NFL gamebooks and play-by-play. That's why he have target info for wide receivers but not air yards vs. YAC, as Aaron noted in the Irvin-Rice discussion.

The people Aaron thanked at the end of the column aren't charting games; they're taking the play-by-play (usually just a photocopy or PDF) and translating it into a spreadsheet format.

Points: 0

#25 by AngryCoder (not verified) // Feb 17, 2010 - 5:37am

You do that manually? Can't you just write a script which parses the PDF and then run some limited postprocessing?

Points: 0

#27 by Arkaein // Feb 17, 2010 - 9:27am

If it's a PDF then it's an image PDF, not a traditional PDF with fully digital text.

FO first talked about this a couple of years ago when they ran out of fully transcribed play-by-play, and had to switch over to essentially paper archives, which are apparently bad enough that OCR (optical character recognition) software chokes pretty badly on it, not good for documents with lots of number you want to get right.

Points: 0

#30 by Kevin // Feb 17, 2010 - 12:11pm

Yep - I had tried that myself independently of that test and found that several different OCR softwares didn't do a good job at all - I ended up spending more time going over those and error checking than I did if I just did a straight transcription. Also by doing the straight transcription I'm more likely to find an error in the official PBP (spotting irregularities seem to me to be the most frequent problem)

Points: 0

#11 by deflated (not verified) // Feb 16, 2010 - 6:28pm

That was an amazing season for Elway and Sharpe. Ineffective running game with Rod Bernstine's 800 yards at 3.7 per carry. WRs led by Derek Russell (who?) for 13 games in his one half decent season. Everyone knew it was going to be Elway to Sharpe and somehow Denver still had a top 5 offence.

I miss the old AFC West.

Points: 0

#16 by The Guy You Do… (not verified) // Feb 16, 2010 - 8:23pm

Everyone was confused at this new Wade Phillips idea of letting John Elway pass. Even sometimes not on third downs or in the last two minutes of the game!

I remember spending the whole year wondering whether Wade would get fired mid-season or the offseason. (Then he ended up making it through 1994 as well before the inevitable Shanahan hiring.)

This season was also the only time I ever got to go see the Broncos. I still hate Glyn Milburn for that game. However, the game also gave me an excuse to say "Fuad Reveiz," the greatest name in the history of football.

Points: 0

#13 by Lou // Feb 16, 2010 - 7:41pm

The most dissapointing part of that Ryan-Gilbride video is that it cuts off before they go to Pam Oliver. I had no idea she had been around that long, this raised so many questions for me. Was she any worse then than she is now? can sideline reporters improve with time? most importantly, how did she look back then? my guess: smokin hot.

Points: 0

#17 by muteant // Feb 16, 2010 - 8:28pm

I can answer the last question: she was, indeed, smokin hot. As for evaluating sideline reporters, I'll leave that to someone who doesn't look at their feet to see if their toenails need clipped every time he hears the words "now let's go down to the field ..."

"We're the worst thing since sliced bread" - Steve Francis

Points: 0

#15 by Temo // Feb 16, 2010 - 8:18pm

Michael Irvin=most talented skill position player on those Cowboys teams. Yes, that includes Emmitt.

Points: 0

#18 by mm (not verified) // Feb 16, 2010 - 8:46pm

Ahh, the year that marks the sudden decline of Jim Mora's Saints. They started this year 5-0 before finishing 8-8. DVOA shows the collapse wasn't just a trick of schedule: defensive DVOA of -10%, but weighted was -5.4%; offensive DVOA of -19.6% but weighted was -34.5%!!!

(note: back then Jim Mora was a defensive genius and knew nothing about offense)

I'm really interested in what the DVOA numbers will come up with as you go to the years before this. The Saints were a really good team in that era with no playoff success. Jim Finks had assembled the team in the late 80s through the draft and with ex-USFL players, but retired in 1992. However, even under Finks the organization had shown an inability to adapt to changing ways of the league: in 1990 he let QB Bobby Hebert sit out a year rather than pay him a whole $1 million, yet could not find anyone else competent to play QB for a team with an elite defense.

Points: 0

#19 by RickD // Feb 16, 2010 - 8:47pm

The yardage comparison table between Rice and Irvin is not written in the most useful form. By itself, it means nothing to me that 26% of Rice's receptions were between 11-15 yards, while 38% of Irvin's receptions were in the same category.

It would be useful to instead list cumulative statistics for each yardage category. By which I mean, tell us that 66% of Rice's receptions were less than 15 yards, while 63% of Irvin's receptions were in that range.

Of course, percentages alone aren't terribly interesting either.

Oh - great job, BTW.

Points: 0

#21 by Vincent Verhei // Feb 16, 2010 - 9:28pm

Or we can draw the line at 10 yards. 40 percent of Rice's receptions gained 10 yards or less, compared to 25 percent of Irvin's receptions.

Points: 0

#22 by sanderrp@gmail.com // Feb 16, 2010 - 10:20pm

John Carney probably played every season from '88-2009. You won't be rid of players in every year before 2010 for a while. I hope he plays next season, too.
Though apparently he didn't actually do anything in '89, just suited up for one game (if that's how I should interpret NFL.com's 1 Game 0 Games Started).

Points: 0

#23 by Travis // Feb 16, 2010 - 10:52pm

In 1989, Carney kicked off once for the Bucs in the Week 15 game against the Lions.


Jason Elam, Jason Hanson, and Jeff Zgonina all played every season from 1993 to 2009.

John Kasay and Mark Brunell were on active NFL rosters/injured reserve in each season from 1993-2009, but didn't play a single game in at least one of those seasons.

Craig Hentrich was on the Packers' practice squad in 1993.

Points: 0

#24 by ammek // Feb 17, 2010 - 3:16am

It is tempting to view the current, late-2000s NFL as exceptional, with pass offense on a seemingly endless upward spiral, and parity wilting in the face of persistent Pats-Colts-Steelers dominance and Lions-Raiders-Browns futility. But in many statistical categories — scoring, offensive yardage, cumulative wins by playoff teams — the modern era is quite unremarkable. (2009 may be an exception.)

The period 1991-93, however, was genuinely remarkable: it was the last time the NFL worried it had a boring product. DVOA is now wading into the strangest period in recent NFL history. Here are some stats to prove the point:

• Over the three years 1991-93, teams averaged 18.9 points per game. That’s the third-lowest average since 1942-44. Only 1976-78 and 1977-79 were lower — and those years prompted wholesale changes to the rules. Six teams scored fewer than 200 points in a year during 1991-93, something that had never happened in a 16-game season until 1990, and has only happened four times since.

• The decline in scoring happened despite increases in field goals and defensive and special teams scores. Rushing and passing touchdowns plummeted to unfathomable depths. Teams averaged fewer than 0.8 rushing TDs/game in two straight years, 1992-93, for the first and only time since 1936-37. Passing TDs/game dipped to levels last seen before the rule changes in 1977-78: the 1544 TDs thrown in 1991-93 were 19% fewer than in the peak three-year period of 1983-85, and more than one-third below 2007-09. Combined rushing and receiving TDs averaged 1.9 per game, the lowest three-year total since 1940-42. The nadir was 1993: with only 825 rushing and receiving TDs in 448 games, it became the lowest-scoring season for offense since 1938.

• Yardage was down, too. Here, 1993 was the best of the three years — and marginally better than 1990, too — but its yards/game average of 317.9 was still lower than any year of the 1980s. (It was higher than 2005 or 2001, though, and equal to the 1996-2000 average). Taking a four-year period from 1990-93, the average yards/game of 306.1 was the lowest since 1973-76, some 7.6% down from a peak of 329.5 in 1983-86 — though only 1.2 % below the 2006-09 average of 309.6.

• When you think of early 1990s offense, you think West Coast and you think Run’n’Shoot. Sure enough, rushing attempts and yardage shrank significantly. Team rush attempts/game fell below 30 for the first time ever in 1989 (29.2) and have not exceeded 28.4 since. Indeed, 1990-93 can be seen as the beginning of the modern rush-o-phobic era: the average of 27.8 attempts/game is very similar to 2006-09 (27.6). However, the average gain remained at 1980s levels (3.9 yards/carry) rather than today’s 4.2.

• But wait! What’s this about passing attempts? They were down too? After averaging between 31.4 and 32.3 passes/game every season from 1981 to 1989, suddenly the numbers dropped: 30.2 att/g in 1990, 31.1 in 1991, 29.9 in 1992. (Attempts in 1993 bounced back close to 1980s levels at 32.2.) Only one season since then has been below 32 att/g (2004).

• Passing yardage stumbled as a result, dipping below 200 yards/game in three straight years, 1990-92, before ‘recovering’ to 200.6 in 1993. The last full season below 200 pass y/g had been 1980. None has been below 200 y/g since; the average of the last decade has been 208.8, with a low of 200.4 in 2003.

• This despite record completion percentages. 1991 saw the league average top 57% for the first time ever, and 1992 and 1993 were better still: the three-year average of 57.6% was up almost three percentage points on 1985-87, and one whole percentage point better than the previous three-year high (1982-84).

• Consequently yards/attempt were down only marginally on the 1980s. Here, 1993 was the worst year — only five seasons in the last 64 years have had worse than its 6.7 y/a average: four of them in the mid-1970s, the other being 2003. Nevertheless, the three-year average for y/a from 1991-93 (6.8%) is only fractionally lower than 2007-09 (6.9%) and better than any three-year stretch between 1994 and 2004.

• The other improvement was the interception rate. This dipped to 3.3% in 1993, the best rate ever at the time, and more than 20% lower than it had been only a decade earlier. The rate had hit record lows about every other year from 1978 to 1992, but even the 1991 rate of 3.5% has not been reached since. The rest of the 1990s averaged 3.2 ints/att, as did 2000-2006. The three-year period 2007-09 has set a new low at less than 3.0%.

• The sack rate for 1991-93 is pretty hard to read: there was one very bad year (1992: sack rate of 7.8%, worst since 1985) following a record best (1991, 6.6%). 1993 was between the two (6.8%) although it set a record for fewest yards lost to sacks, bettered only once between 1994 and 2001. It has been equalled or bettered every year since.

So what happened? Why did scoring and offense hit a nadir in 1991-93? And how were they resuscitated?

• The freefall in yards per completion and interception rate points to the spread of the high-percentage west-coast offense across the league. Mike Holmgren and Denny Green took their first head coaching jobs in 1992.

• Even so, this was an era dominated by defensive coaches. Almost all the best teams were either coached or built by defensive minds: Jimmy Johnson, Marty Schottenheimer, George Seifert, Buddy Ryan, Chuck Noll-Bill Cowher, Jim Mora, Jerry Glanville-Jack Pardee, Bill Parcells, Bill Belichick. It saw the debut of ‘Blitzburgh’ and the Dungy Two.

• The QBs might have been prudent, but they weren’t very good. As Aaron notes, this was a terrible era for young quarterbacking talent. There were 28 QBs drafted in rounds 1-3 between 1987 and 1993. From whatever statistical angle you choose, the careers of David Klingler, Cody Carlson, Rick Mirer and Billy Joe Tolliver were all in the better half. The rest is a mess of Stouffers, Marinovichs, McGwires and Dilwegs. Even the summit is underwhelming: after Favre and Aikman, the best QBs on the list are Bledsoe, Testaverde and Chris Chandler. Was there a massive disconnect between the college game and the NFL? Or was it just a freakily hopeless crop? (The later rounds were deeper, but the best of these were not starting in 1993 — Gannon, Brad Johnson, Brunell, Trent Green. Among those who were, Beuerlein, Peete, Richardson, Mitchell, Majkowski and Humphries were hardly setting the world on fire.)

• Even though rushing attempts were down, there was an unusually low number of plays per game in 1990-92. The average had been around 65 in the mid-80s; in 1990, it fell below 61 for the first time since 1972, and bottomed out two years later at 59.9, the lowest since 1946. There was a significant rebound in 1993, to 62.8, which has been about the average ever since. Does anyone know what happened? Was there a change in the rules governing the play-clock?

• What was the consequence of great defensive coaching, conservative offense, fewer plays, and low scoring? Why, it was parity, of course. 1993 was a prime example: of the 28 teams, some 12 finished within one game of .500. The following years would be even more balanced: 13 in 1994, 15 (of 30) in 1995, 13 again in 1996. This remains the most middle-heavy four-year streak in history.

• How much of that mediocrity can we attribute to free agency? My guess is: not much. 1993 was the first year of free agency as we know it, and only 121 players changed teams (that’s 4.3 per team). More significant was the lack of really dominant or really bad teams: no team won more than 12 games, and only one team (3-13 Cincinnati) lost more than 12. That’s only happened one other time in a 16-game season, in 1988. Again the continuity here is with the late 1980s rather than the ‘free agency period’.

• Nowhere was that continuity more evident than in the Superbowl, where the AFC representative crashed for a 10th straight year. Although the conferences were evenly balanced in total points differential, the AFC was huddled around average while the NFC had greater variance. The median distance from zero, plus or minus, in points differential for 1993 was 35.5 for the AFC, 71.0 for the NFC. That was the ninth year in a decade that the NFC had a greater disparity.

• Splitting the numbers by conference indicates that the early 1990s’ trends were very much in line with those emanating from the mid-1980s; all that had happened is the NFC had started to look a little more like the AFC. From 1986 to 1996, AFC teams produced just 13 seasons of 12 or more wins; the NFC had 21. From 1988 to 1993, AFC teams totalled 26 seasons of 10 or more wins; the NFC had 36. So the ‘mediocrity’ of the free-agency period was not new in the AFC. But the NFC had only recently begun trending down. By 1993-96 its top teams were not winning many more games than the AFC’s.

The offensive revival began in 1994 with rule changes designed to improve starting field position (and destined to provoke much wrath in overtime games). In 1995, the radio transmitter appeared in QBs’ helmets, and in 1996, officials began enforcing “more stringently” the five-yard contact rule. Offense and scoring perked up: passing yardage peaked in 1995, and offensive TDs have recovered to 1980s levels for most of the last decade-and-a-half.

Tempting as it is to read the beginning of free agency as the dawn of a new era, the numbers don’t back it up. 1993 looked and felt like the previous few years, with bad quarterbacking, low scoring, and a dreadful Superbowl.

Points: 0

#26 by Travis // Feb 17, 2010 - 7:56am

The low number of plays per game mainly resulted from a change in some timing rules before the 1990 season. The biggest of these was the restart of the game clock after out-of-bounds plays outside of the last 2 (or 5) minutes.

Points: 0

#35 by The Other Ben … (not verified) // Feb 17, 2010 - 1:59pm

When did the NFL switch from a 45 second play clock to a 40 second play clock? 1992, or was it later? Am I just pulling this out of my ass? It's in the back of my brain from watching the number of wasted timeouts from year one of Joe Gibbs II. He could NOT adjust to the loss of that 5 seconds.

Points: 0

#49 by ammek // Feb 18, 2010 - 4:08am

Interesting from your link that the overtime effect was already at the heart of the debate about moving kickoffs back to the 30-yard line. I didn't realize the NFL had been fudging this issue for more than 15 years.

Points: 0

#29 by Aaron Schatz // Feb 17, 2010 - 11:13am

Yo, Ammek. Great post. Remind me to re-post this when I put up 1992... whenever that happens. (August?)

Points: 0

#31 by ammek // Feb 17, 2010 - 12:48pm

Thanks. I dig this old stuff. Looking forward (backward?) to 1992 and some more fabulous Andre Reed rushing stats.

Points: 0

#44 by Jeff Fogle // Feb 17, 2010 - 3:35pm

Great post ammek, off a great article by aaron. What do you think about the possibility that 91-93 was basically a key transitional point in the "predator and prey" flow of league history...representing the defensive peak after "the '85 Bears taught everyone how to play modern defense" and before the combination of offensive strategy and rules changes that helped get the equilibrium back? I had always viewed that sequence in that light...would be interesting to see what you guys and DVOA think about that.

Interesting to note that KC was fairly dominant statistically in the playoff win at Houston, even though Houston had a bye week to get ready. Vegas spread was only -3 based on the data at covers. That means the market saw the teams as even even with Houston's very strong close to the season.

Total Yardage: Kansas City 354, Houston 277
Yards-per-play: Kansas City 6.3, Houston 4.9
Net Passing: Kansas City 22-38-2-283, Houston 32-43-1-238
Rushing: Kansas City 71, Houston 39
Turnovers: Kansas City 2, Houston 3

Moon was sacked 9 times for 68 yards, helping to kill stat production. Montana, a veteran perhaps less likely to be shaken by what Ryan was doing defensively, was only sacked twice. He did have two picks though, and more incomplete passes than Moon. Marcus Allen rushed for 74 yards on 14 carries (rest of team was -3). Longhorn Keith Cash had 80 yards receiving on four catches for the Chiefs.

Remember either Birden (11 yd TD pass) or Davis (18 yd TD pass) just drilling the ball at a Buddy Ryan banner hanging on the Astrodome outfield fence after his score. Felt like the whole city of Houston curled up after that. The feeling of invincibility from the prior few months was gone.

Houston sure didn't look like championship material in this particular outing. Not sure if it's been mentioned already. Kansas City hasn't won a playoff game since.

Points: 0

#51 by ammek // Feb 18, 2010 - 4:38am

Neither has Marty.

The Chiefs' loss in the AFC Championship Game must have been among the most mourned in history. Not only did Montana look terrible in his last big game, throwing 2 picks and completing barely 50% of his passes, but the result left us with a fourth successive Buffalo Superbowl berth.

It will be interesting to see how DVOA rates those near-miss Bills. Although it did grab the top seed, the 1993 model felt like the feeblest of the four. Emphasizing the importance of takeaways in DVOA, the Bills' '93 defense ranked second-last in yards allowed, but 8th in DVOA thanks to a league-leading 41 forced turnovers. As is often the case, the turnover streak didn't extend to the Superbowl. With the offense struggling — viz Thurman Thomas' replacement-level season — another beatdown was inevitable.

DVOA doesn't yet cover the peak of his career, but the Jim Kelly Hall of Fame ticket is looking ever more questionable. Over the last four seasons of his career, he never featured among the top ten QBs in either DYAR or DVOA. His season-average DVOA was -6.4, and his cumulative DYAR is 647 so far.

Points: 0

#56 by Jeff Fogle // Feb 18, 2010 - 2:14pm

Went to look up the box. Looks like Montana was replaced midway through. He's credited with just one pick.

Montana 9-23-1-125
Krieg 16-29-1-198

KC was down 20-6 at the half. Don't recall off the top of my head if there was an injury or if KC switched to Krieg to try to jumpstart things. Ugly memory for Montana fans either way.

Was Thurman Thomas really "replacement level?" I guess I could see that on per-play averages because they buried him in workload. He rushed for over 1,300 and added more than 400 more on receptions. Big volume in the league rankings, but not great per-play stuff with more than 350 carries and almost 50 catches. Wonder if they were running out the clock with him when ahead and it hurt his stats (but they probably did that every year). Definitely a step backward from what he had been doing in previous seasons. Tough to call a performance "replacement level" with that much workload though I think. Most replacement guys can't handle that workload, or get 1,700 yards and finish 2nd in yards from scrimmage in the league if they had it.

Respectable playoffs for the Bills...winning 29-23 and 30-13 with yardage edges vs. teams from the tough AFC West (as Aaron outlined). That would seem better than DVOA or weighted would have projected. Some general similarities to the Colts this year in that regard given where Indy '09/Buffalo '93 ended up in DVOA stuff...and Baltimore '09/KC '93 ended up. Indy/Buffalo beat their most dangerous playoff threat (based on DVOA) by 17, but failed for the AFC on the big stage.

Agree that Kelly doesn't really have Hall of Fame stat production. But, how do you weigh being the "team leader" of a squad that won four conference championships and went 66-24 when he was a starter during their six season peak. Averaging 11-4 per 15 games and winning four conference titles trumps production in the minds of some voters. The goal is to win championships, not compile stats. Tough call with a guy like this I think. They went 12-4 once when he had a 15-17 TD/INT ratio, then again later when he was at 18-18. Surprised his stats were so soft when I went back and looked (lived some in Houston when he was a Gambler, lol). Looks like he was getting more credit at the time than he deserved for Buffalo's success...and that took hold in everyone's mindset. Still, HOF's in general smile on prominent leaders on champions...and Buffalo was a consistent AFC champ if not NFL champ during his time there. I'm guessing he won't grade out as elite as the DVOA studies drift further back into his career.

Points: 0

#60 by capt. Anonymous (not verified) // Feb 19, 2010 - 7:23am

Went on pro football reference and Jim Kelly's career is comparable to 4 other hof'ers according to their similarity scores metric. Maybe we've lost track of how tough it was to play qb in that day.

Points: 0

#63 by Eddo // Feb 19, 2010 - 11:04am

It was indeed harder to play QB in the early 1990s, but Kelly's DYAR and DVOA don't even rank well amongst his peers; he finished 13th in DYAR and 22nd in DVOA in 1993.

The 13th-ranked QB by DYAR this season was Matt Ryan, and the 22nd by DVOA was Byron Leftwich. Regardless of what you think of Ryan moving forward, he didn't play like a Hall-of-Fame QB this season.

Points: 0

#67 by Thomas_beardown // Feb 19, 2010 - 7:12pm

Look at the similar players after 10 years: Warren Moon*, John Elway*, Fran Tarkenton*, Donovan McNabb, Drew Bledsoe, Bob Griese*, Mark Brunell, Troy Aikman*, Johnny Unitas*, Ken Anderson.

That's a pretty good list.

Points: 0

#61 by Travis // Feb 19, 2010 - 7:45am

Montana was injured 3 plays into the 3rd quarter and didn't return.

Points: 0

#64 by Jeff Fogle // Feb 19, 2010 - 1:38pm

Thanks Travis.

Regarding Capt Anon and Kelly...three of the four similar guys were from prior era's. Passer ratings and TD/INT ratios have generally improved over time (particularly as you transition from the 70's through the 90's).

Namath: last year as a regular '76, career rating of 65.5 with a very ugly 173/220 TD/INT ratio (47 more INT's than TD's!).

Griese: last year as a regular '79, career rating of 77.1, with a 192-172 TD/INT ratio.

Bradshaw: last year as a regular '82, career rating of 70.9, with a 212-210 TD/INT ratio.

(Note how Greise jumps out of phase with the time line in a way that speaks well of him).

Kelly's last year as a regular was '96, but he's grading out as most similar by pro-football reference to guys centered around '79-80. Interesting that the other three were AFC champions too. That's just something that gets respect from voters (though the other three won SB's obviously and Kelly didn't).

The fourth similar guy was Aikman, a contemporary.

I could see why a stat approach would be skeptical about Kelly in that light as a true HOF candidate. I think I'm on the fence myself. Stuff to respect about Kelly, stuff that doesn't shine very brightly when you dig deep.

Points: 0

#65 by Eddo // Feb 19, 2010 - 2:17pm

Nice comment, Jeff.

Kelly's interesting; like you say, on the surface, he appears to be a solid Hall of Famer. However, I think we have that perception because of the Bills' success, specifically the four straight conference titles.

Without that unique feat, does Kelly waltz into the Hall of Fame? I'm not sure.

Now, the better question is, if the Bills only go to three of those Super Bowls, is Kelly a Hall of Famer? I think it depends which one of the four they miss.

If they miss the first or fourth, they have still made three in a row; this would tie them with the 1971-73 Dolphins for most consecutive Super Bowl appearances. Kelly's probably in, though without the close loss to the Giants, the Bills are more of a joke, being blown out all three times they reached the Super Bowl.

If they miss the second or third, they don't share any record with the Dolphins. They're the equivalent of the 1973-76 Vikings or 1986-89 Broncos. Now, Tarkenton and Elway were easy Hall of Famers (though Tarkenton inexplicably had to wait several years), but they also have much more statistically strong and less championship-dependent cases than Kelly. It's not a stretch to say that, if the Bills "only" make three of four Super Bowls, instead of three or four consecutively, Kelly's Hall of Fame chances decrease dramatically.

Ironically, then, Kelly's backup may be quite responsible for him being in the Hall of Fame. If Frank Reich doesn't lead the great comeback against the Oilers following the 1992 season, the Bills never reach their third straight Super Bowl, don't hold a unique place in history, and Kelly might have missed out on the Hall of Fame.

The "What if" game can be fun in some cases.

Points: 0

#66 by Jeff Fogle // Feb 19, 2010 - 3:55pm

Agree with you eddo...Kelly's a fun case. We should probably also throw in his two years with the USFL. It is the "Pro Football" Hall of Fame after all. He won an MVP in 1984...and had 83 TD's and 45 Int's in 2 seasons according to his Wikipedia page. A lot of both for just 2-years (Mouse Davis the coach). Wiki page notes that Steve Young and Doug Williams were QB's in the league at the time...so he did beat out some quality for the award. Other future NFL stars too as I recall.

Having to pass during winters in Buffalo may have hurt his stats as well. Still, the numbers weren't great if you adjust for that in my view...and his SB numbers weren't great either on neutral fields in huge games (2 TD's, 7 INTs).

Interesting point about Reich...not that those of us with some roots in Houston want to think any more about that game! (lol)

Points: 0

#69 by Mike19531 (not verified) // Feb 20, 2010 - 4:04am

Montana's pick was actually a pass that deflected off of the hands of fullback Kimble Anders and was intercepted inside the Bills endzone. On that play, Montana was sandwiched by two defenders and suffered a concussion.

Dave Krieg played the rst of the way.

Points: 0

#45 by bravehoptoad // Feb 17, 2010 - 4:36pm

What was the consequence of great defensive coaching, conservative offense, fewer plays, and low scoring? Why, it was parity, of course.

Also crucial to parity was that the salary cap was much, much smaller then, so that premier free agents hit the market every year.

After 1993, Deion Sanders became a free agent, was snapped up by the 49ers, who won the Superbowl with him, then became a free agent again and was snapped up by the Cowboys, who won the Superbowl with him. And his signing wasn't particularly distinctive.

These days, a premier player would only be available like that if he were a huge character risk. Then, it happened all the time.

Points: 0

#28 by Ashley Tate (not verified) // Feb 17, 2010 - 10:46am

1993 was my first year playing fantasy football, and I was lucky enough to pick up both Gary Brown and Natrone Means (#3 and #5 in DYAR) as free agent running backs. Good times!

Points: 0

#37 by Lance // Feb 17, 2010 - 2:19pm

My first year was 1991, but reading this brings back all sorts of memories... I got lots of points drafting those Run-and-Shoot Oilers. Even if I could never understand how Heyward Jeffires was said "Jeffries".

Points: 0

#32 by Boots Day (not verified) // Feb 17, 2010 - 1:19pm

Excellent post, Amnek. I wonder if QB scouting just went horribly wrong for a few years there. One high draft choice you didn't mention was Jeff George, who went No. 1 overall in 1990 after famously wowing them at the combine. The six-foot-eight Dan McGwire went in the first round in 1991, while 5-10 Doug Flutie had gone in the 11th a few years before the scope of your study, after a brilliant college career. And of course, the draft whiffed entirely on Kurt Warner in 1994.

Points: 0

#36 by Travis // Feb 17, 2010 - 2:01pm

Doug Flutie wasn't drafted until the 11th round of the 1985 NFL draft because he had already signed with and was playing for the USFL's New Jersey Generals. Herschel Walker went in the 5th for the same reason.

Points: 0

#33 by WCH (not verified) // Feb 17, 2010 - 1:37pm

I'm very surprised by how low Sterling Sharpe ranks in the 1993 and 1994 ratings. My perception, at the time, was that he was a truly elite WR, sure handed and a strong runner after the catch, and a guy clearly outpacing Michael Irvin and Tim Brown for the title of "Best WR from the 1988 Draft." After he broke the reception record in consecutive seasons, I even remember some "best WR in the game" talk taking place.

We'll see what earlier years bring, but he's starting to look like a solid but unspectacular WR, who's numbers skyrocketed when a young Brett Favre entered the lineup and started firing him the ball 12 times per game.

Points: 0

#34 by strannix (not verified) // Feb 17, 2010 - 1:43pm

So the Bears were 26-9-8 in offense/defense/ST. That's really sort of funny.

Points: 0

#38 by johonny (not verified) // Feb 17, 2010 - 2:20pm

I thought Shula ran pretty much the same system his whole time in the league. The Dolphins used Byars and Kirby the same way the used Bennett and Nathans in the 1980s. Marino threw a lot of passes to the backs throughout his career.

Points: 0

#40 by NWeb (not verified) // Feb 17, 2010 - 2:59pm

Not sure if it's a terms of use violation of some sort - not sure how the PBP's were sourced - but is there any reason why (even if for a nominal fee) FO couldn't share the underlying PBP's themselves. Could be delimited, Word, whatever. I'd love to get that to add to my collection - which stops at 1996 right now.

Points: 0

#43 by DP (not verified) // Feb 17, 2010 - 3:20pm

In case this isn't mentioned elsewhere, it's worth noting that 1993 was the first of two years when Joe Montana was the QB for the Chiefs, and his performance probably had a lot to do with their #3 ranking in DVOA, as well their appearance in the AFC title game (the Chiefs hadn't even won their division in over 20 years at that point, I believe). On Championship Sunday, when the Cowboys ultimately beat the 49ers and the Bills beat KC, we were oh-so-close to a Montana vs. the 49ers Super Bowl...can you imagine?

Points: 0

#46 by R. Carney (not verified) // Feb 17, 2010 - 7:42pm

One of the interesting notes branching off Annex's post (which is the most in depth and informative I've ever seen on this site by a non-contributing writer) is that if you look at the Hall of Fame qbs of this era, whom people at the time set as the gold standard of qbs (Aikman, Elway, Kelly, all Hall of Famers), their numbers aren't thrilling. The one that stands out is completion percentage, which is a tell tale sign of how offense has evolved. League wide completion percentage in '93 was 57.7%, while in '09 it was 60.9%. I think this is a testament to how numbers from era to era can't be compared (in HOF disucssions, etc.), because to say that Matt Schaub is a better QB than any of the guys parenthetcally mentioned earlier would be obsene.

Points: 0

#47 by R. Carney (not verified) // Feb 17, 2010 - 7:44pm

Ammek* sorry...that was an accident, not me being a dick

Points: 0

#48 by Anonymous Jones // Feb 17, 2010 - 8:11pm

It's amazing how everyone forgets the struggling Drew Brees. Even in his "better" year of 2004, I remember how I felt all hope was lost in the SD home opener, Game 2 against the NY Jets. He was 8 of 19 for 146 yards and that doesn't even begin to tell the story of how awful he was. He had *no* touch on the ball, just rifling it when he could have lofted it with ease. It was literally a Jamarcusian performance. With all seriousness, if you has asked me the odds that he could develop into a great QB after that outing in September '04, I would have said 100-1 against. It seemed absolutely impossible that this person had any idea how to be a decent quarterback. Obviously, the Chargers agreed; that's why they drafted Rivers. Then, slowly but surely, and steadily, Brees just got better and better; now, even though he's been playing for a different team than mine for four years, he's still my favorite player.

Seriously, the lesson is some quarterbacks need time. Of course, the problem is some quarterbacks are indeed terrible and must be jettisoned. Tough thing this job of talent evaluator is.

Points: 0

#53 by td (not verified) // Feb 18, 2010 - 12:31pm

Dallas was easily the best team in 1993 once Emmitt Smith ended his holdout. They started 0-2 without him, and still ended with the best record in the league.

Points: 0

#55 by JIPanick // Feb 18, 2010 - 1:19pm

Half a point of DVOA is practically nothing (the two teams will probably switch spots back and forth over the next few DVOA updates, like DEN, GB, KC, and SF have done in the 1997 ratings) and I think that the Cowboys blasting the 49ers in the NFC championship is a pretty convincing case that the Cowboys were the best team of 1993.

I guess, however, you could counter-argue that regular season DVOA also doesn't include the 49ers annihilating the Giants in the divisional game.

Points: 0

#57 by Lance // Feb 18, 2010 - 7:54pm

Re Smith and the holdout-- yes, I was wondering if this would get mentioned, but it doesn't seem to have been. Smith was the league MVP that year and led the league in rushing even though he missed the first two games of the season. I wonder how Dallas would look if one managed to factor out those two games.

Points: 0

#58 by Mike19531 (not verified) // Feb 19, 2010 - 5:39am

Of the three Aikman/Emmitt/Irvin championship squads, I have long felt that the 1993 version is the best of the three.

If memory serves, not only did Emmitt miss the first two games because of his holdout, he missed most, if not all of Dallas other 2 losses because of a leg injury.

I believe that Emmitt got a charley horse or an injury of that nature after his first carry at Atlanta and had to leave the game. Naturally, Emmitt's abscence was a major factor in Dallas' loss.

Dallas' next, and last loss of the season, came vs Miami on Thanksgiving day vs. Miami, a game best remembered for Leon Lett's misadventure with a blocked fg late in the game. But if I recall, Emmitt missed all of that game because of his injury.

If my memory's accurate, Dallas could have won 14+ games if Emmitt had been available for the entire season. With him, they shellacked an outstanding 49er team twice (once in the regular season, then again in the playoffs). That's why I have always believed that the '93 Cowboys were the best of their 3 champions. With a healthy Emmitt, the '93 Cowboys had about an airtight team as you could imagine.

To Aaron or any of the other Outsiders, am I right in remembering that Emmitt missed the majority of the Falcons game and the entire Miami game?

Points: 0

#59 by ammek // Feb 19, 2010 - 6:35am

I'm not an Outsider but you can check that kind of information here and here.

Links don't seem to work. They referred you to pro football reference, which has box scores for historical games.

Points: 0

#62 by Travis // Feb 19, 2010 - 8:04am

Emmitt suffered a bruised right thigh late in the first half of the Falcons game and didn't play the rest of the way. It's possible he came into the game with an injury of some sort - he only rushed once (mid-first quarter), but then again he did catch 4 passes. The lack of runs (12 Dallas passes to 2 runs in the first half) could also be explained by Bernie Kosar making his only start of the year.

Emmitt was doubtful going into the Dolphins game, but got the majority of the carries (and 9 receptions).

Points: 0

#68 by Anonym (not verified) // Feb 19, 2010 - 8:41pm

The Oilers-Bills playoff game was 41-38 not 38-35

Points: 0

#71 by Majorajam (not verified) // Jul 22, 2010 - 12:06pm

On the Taylor/Rice/Irving comparison, I wonder whether receivers aren't getting credit for being targeted more, and if not, why not. If a basketball player gets the ball only on plays in which they're favorably positioned or set up, but otherwise stays on the periphery of plays, that's obviously going to flatter their efficiency ratios (like FG%, assists/turnovers, etc.). But these clearly carry a lesser burden then the guy whose hands every play has to pass through, and are concomitantly less important to their team and crucial to its success. Likewise with any focal point of an NFL offence.

As to the Rice/Irving comparison, I question any weighting schema that has the distribution of yardage as more important than performance on 3rd down and in the red zone, not least due to the fact that the latter are (far) more important to winning actual football games, but also due to the fact that big plays are frequently determinative of game outcomes, (in any case far more so than the 20 and 30 yard variety). One wonders if a barbell is not a good thing.

This points up a second critical dimension to football performances that I don't believe is captured here, which is whether 'successful plays' and big plays come amidst blowouts or in close games where the outcome is in doubt. The difference is one of the things that separates the Michael Jordans of the world from the Allen Iversons. So that would be a glaring omission, potentially as significant as a 'strength of the opposition' omission, if indeed it is an omission.

In any case, the numbers certainly are an upgrade over raw fantasy type stats, but I'm under the impression there's non-trivial room for improvement.

Points: 0

Save 10%
& Support Aaron
Support Football Outsiders' independent media and Aaron Schatz. Use promo code SCHATZ to save 10% on any FO+ membership and give half the cost of your membership to tip Aaron.