1987 DVOA Ratings and Commentary

by Aaron Schatz

With 30 years of hindsight, the 1987 NFL season now looks even more ridiculous than it looked at the time. Trying to explain it to an NFL fan who wasn't alive at the time seems even more absurd.

Did you enjoy last week as the news of free-agent signings came fast and furious? Well, there was no NFL free agency in 1987. There was no collective bargaining agreement. When a player contract ended, that team still had the rights to that player unless they specifically cut him. This messed-up labor market is part of what created the opportunity for the USFL in 1983, but by 1987 the USFL was toast.

Yes, the NFL owners really did bring in a bunch of replacement players and stuck them on the field for three weeks and pretended they were real NFL teams. Week 3 of the season was cancelled, and then the NFL played three weeks of the schedule with rosters that were almost entirely different. The death of the USFL left a lot of unsigned players floating around with a year or two of professional football experience. Some of these players took the field during the strikebreaker games in Weeks 4-6 of the 1987 season. But there were nowhere near enough of these players to actually fill all 28 NFL rosters. So for three weeks, NFL teams fielded rosters with a mix of USFL and CFL refugees, washed-out low-round draft picks, undrafted camp bodies, random weekend warrior athletes, and future rap mogul Suge Knight. (Yes, the man behind Death Row records had played defensive end for UNLV and had two games with the 1987 Rams.)

Making this all even more absurd was the fact that the strikebreaker rosters also included a number of actual NFL starters, because the cohesiveness of the NFL Players Association rapidly broke down. Some players crossed the picket line immediately. Others waited until the second week of the strike. By the third week of the strike, a bunch of players were crossing because everyone knew the strike was going to be over. So you ended up with Joe Montana throwing to wide receivers you had never heard of, but he also had Roger Craig in the backfield, and the other team's quarterback would be some eighth-round pick who was cut before the season started. And then in Week 7, everyone was back and the NFL just went on with the schedule as usual.

(For a good history of all this craziness, I recommend buying A Good Walkthrough Spoiled: The Best of Mike Tanier at Football Outsiders, which includes the 50-page "Year of the Scab" mini-novella about the 1987 season. "Year of the Scab" is going to be the basis for -- and title of -- an ESPN 30-for-30 documentary currently in development.)

All this nonsense means the official statistics for 1987 are completely screwy. Most players had a 12-game season, but a lot of stars played 13 games, and some guys played in 14 or even 15 games. Each team's record is 20 percent the responsibility of a mostly different group of players, and the NFL actually determined playoff positions based on this. For example, the Indianapolis Colts and Miami Dolphins both went 7-5 with their regular players, but the Colts won the AFC East because their strikebreakers went 2-1 while Miami's went 1-2.

At Football Outsiders, we've solved this problem by pretending that those three games weren't really part of the 1987 season. For our official 1987 DVOA ratings, we have completely ignored the strikebreaker games of Weeks 4-6. Every team's rating is based solely on a 12-game season. Opponent adjustments are also based on regular games only, with Weeks 4-6 removed. But if you're curious about the three strikebreaker games, we've computed ratings for them too. All the 1987 stats pages will have additional tables for the strike games, without opponent adjustments.

The ironic thing about all the work we did to separate out the 1987 strike games is that the team that is No. 1 in our ratings for the regular 1987 games was also No. 1 for the three strike games. In fact, the 1987 San Francisco 49ers may be one of the greatest regular-season teams in NFL history. Their dominance has mostly been hidden from history because of the weirdness of the 1987 season, and because of an egregious postseason loss to the Minnesota Vikings. Five other 49ers teams won the Super Bowl, but the 1987 49ers may have put up the franchise's finest regular-season performance.

The 1987 49ers were not one of the best offenses in DVOA history, and they were not one of the best defenses in DVOA history. But they were the first team we've ever measured to finish No. 1 for the year in both offense and defense. They also had above-average special teams. When you put it all together, San Francisco had 47.2% DVOA for the 12 regular-season games of the 1987 season, ranking them as the third-best regular-season team in DVOA history behind only the 1991 Washington Redskins (56.9%) and the 2007 New England Patriots (52.9%).

Unlike those other two teams, the 49ers did not start the season strong. They led off the season with one of their two losses, falling in Pittsburgh 30-17. In retrospect, this game was a good sign of how insane the 1987 season was going to be. Lower down on the page we'll get into a discussion of just how bad Pittsburgh quarterback Mark Malone was in 1987, but it was a sub-Osweiler performance from the very first week. Malone went 9-for-33 against the 49ers with just 99 passing yards and the Steelers scored 30 points anyway. I know we're really digressing here, but this game was kind of amazing.

Joe Montana threw an interception on San Francisco's first drive, but Malone threw three straight incomplete passes to Louis Lipps and the Steelers settled for a field goal attempt. Gary Anderson missed from 42 yards. On the next drive, Roger Craig fumbled on the 50 and Pittsburgh rookie cornerback Delton Hall returned it for a touchdown. In the second quarter, Malone did manage to lead a touchdown drive and a field goal drive, and the Steelers went into halftime up 17-3. However, Malone imploded after halftime, completing 2 of 13 pass attempts. One of those completions was a dumpoff to running back Frank Pollard for a loss of 5 yards. Malone's only positive passing play in the second half was a 15-yard completion to Lipps after Montana threw another pick in the fourth quarter. That was Montana's third pick of the game, and San Francisco's fourth turnover. The 49ers fumbled three times and recovered two of them. The Steelers fumbled four times but recovered all of them. So even though the Steelers offense couldn't move the ball through the air, the 49ers handed them the game with turnovers.

Honestly, I only remember one other game where a quarterback played this badly and his team won anyway. One man was part of the losing coaching staff in both games. I can't believe that San Francisco wide receivers coach Dennis Green didn't come out after this game and scream at the press that "the Steelers are who we thought they were and we let 'em off the hook!"

The second game of the season was even stranger than the first. The 49ers went to Cincinnati and were trailing the Bengals 26-20 when they got the ball back with two minutes left. Joe Montana is one of the greatest comeback artists in NFL history, right? He threw two incomplete passes to John Taylor and then took an 8-yard sack, and the 49ers punted the ball back with 54 seconds left. The play-by-play does not list timeouts, but it's clear that the 49ers had two timeouts left when they punted. Their only hope was to somehow get the ball back with less than 10 seconds left for a miracle play.

And then the Cincinnati Bengals somehow screwed up the process of kneeling out the clock.

Boomer Esiason kneeled down for a 2-yard loss, and took five seconds off the clock. Timeout. He kneeled down for a 3-yard loss, and took four more seconds off the clock. Timeout. On third-and-15, he stepped back really far and took a 5-yard loss and then let the clock run all the way down to six seconds left. The Bengals took a delay of game penalty there, giving them fourth-and-25 from their own 30 with six seconds left. This is a pretty easy situation, right? It seems obvious the Bengals should a punt, but there's always the possibility of a block. A better option would be to have the punter run out of the end zone for an intentional safety, which would then make the last play of the game a free kick from the 20-yard line. Sam Wyche did neither of these things. Instead, he thought keeping the ball on the ground could somehow use six seconds off the clock. The Bengals ran a sweep with James Brooks and he got slammed by 49ers defensive end Kevin Fagan, tackled five yards behind the line of scrimmage with two seconds on the clock. That left room for one play from the Cincinnati 25, and Montana hit Rice leaping in the end zone for a touchdown. The extra point was good and the 49ers won 27-26.

So the 49ers had a DVOA of about 0% and a 1-1 record when the strike came. But Bill Walsh was ready. Walsh had brought 120 players to training camp in 1987, expecting that labor talks would break down and the owners would prove to be serious about their replacement-player plan. Roughly 30 of those players in training camp were cut specifically so they could be signed again when the strike came, including third-string quarterback Bob Gagliano. It turned out that the 49ers didn't even need Gagliano that much, because some of their best players crossed the picket line after just one week. The 49ers took the field in Atlanta for Week 5 with Joe Montana, Roger Craig, Dwight Clark, Russ Francis, and Joe Cribbs all in the starting lineup. Eventually they won all three of their strike game, with a league-leading 38.6% VOA. They were the No. 1 offense during the strike, though closer to average on defense.

But when all the regular players came back in Week 7, the 49ers marched easily through the rest of the schedule. They split two games with New Orleans, the No. 2 team in DVOA, that were each decided by two points. But they won every other game by a touchdown or more. In the final three weeks of the regular season, they beat the Bears, Falcons, and Rams by the combined score of 124-7. That Bears team finished 9-3 and won the NFC Central, and the 49ers beat them 41-0.

As usual, the 49ers were led by their passing attack. Montana finished third in passing DYAR despite missing three games. This was the first season after the 49ers acquired Steve Young from Tampa Bay to be their backup, and he was out of this world in the three games he started in Montana's place: 54.0% DVOA, 10 touchdowns with no interceptions. The defense was No. 1 against the pass and No. 3 against the run, with safety Ronnie Lott and nose tackle Michael Carter both earning All-Pro honors. But no player dominated the 1987 season the way Jerry Rice did. In just his third NFL season, Rice set a new NFL record with 22 receiving touchdowns -- in only 12 games. Randy Moss broke the record in 2007, but it took him 16 games to do it.

Of course, Jerry Rice easily finished No. 1 in receiving DYAR for 1987. Pro-rated to 16 games, Rice would have 519 DYAR, which would rank No. 8 all-time. Despite all the touchdowns, Rice wouldn't rank No. 1 because of a relatively low 58 percent catch rate. Every other season in the all-time top 20 had a catch rate of at least 61 percent. Rice himself had a catch rate of 64 percent in 1989, which is actually his best season by FO stats (563 DYAR, fourth all-time).

Unfortunately for Rice, in the postseason he had to take a backseat to the player who finished No. 2 in receiving DYAR: Anthony Carter of the Minnesota Vikings. The Vikings were 8-4 during the regular part of the 1987 season but almost didn't make the playoffs because their replacements went 0-3 during the strike. However, the DVOA ratings suggest that bad luck was balanced by good luck the rest of the year, as the Vikings were just 12th in overall DVOA. The defense that dominated the 1988 season was only seventh in 1987, and the offense was average. When the season was over, the officially 8-7 Vikings snuck into the playoffs with the second wild card because only five of the 14 NFC teams ended up with a winning record.

And then they nearly marched themselves to the Super Bowl, beating the top two teams in DVOA in the process.

First, the Vikings went into the Superdome and curb-stomped the Saints by a final score of 44-10. The Saints had one of the league's best pass rushes, so the Vikings ran all over them. Bobby Hebert threw two picks, so Jim Mora benched him for backup Dave Wilson, and Wilson was even worse: 2-of-12 with two picks of his own. The upset spoiled the first playoff game in the history of the New Orleans Saints franchise.

That win earned the Vikings a trip to San Francisco. The 49ers were favored by 11 points. Removing strike records suggested that the 49ers were overrated by the public, since they went 3-0 during the strike while the Vikings were 0-3. But a closer look at the play-by-play data through DVOA suggests the 49ers were actually underrated by the public, because they were dominating the league so much over the last 10 weeks of the season. The game started 3-3 after one quarter, but then the Vikings defense shut down Joe Montana. He finished 12-of-26 for just 109 yards with a pick and four sacks. Steve Young came in during the second half but by that point it was too late because Anthony Carter had destroyed the 49ers secondary. Carter caught 10 passes for 227 yards, which was a playoff record at the time, and added a 30-yard end-around run. Amazingly, he had no touchdowns, though nine of his catches earned first downs. The Vikings bogged down in the red zone and had to settle for five Chuck Nelson field goals -- which itself was somewhat incredible because Nelson had gone just 13-of-24 during the regular season.

The Vikings' win was featured on NFL Network's Top 10, on an episode about the ten greatest playoff upsets ever, and the entire game is available on YouTube.

Finally, the Vikings went to Washington. The Redskins had themselves upset the Chicago Bears the week before. Although Washington had only been an average defense during the regular season, the NFC Championship Game was a tight defensive battle. Washington finally took the lead with a touchdown from Doug Williams to Gary Clark with five minutes left, and they stopped the Vikings' last drive on fourth-and-4 from the Washington 6 with 56 seconds left when Darrin Nelson dropped what would have been the game-tying touchdown. (Thanks to reader Will Allen for enlightening me on this, see comment 4 below.)

The 1991 Washington Redskins may have the best DVOA in the history of our metrics, but the 1987 Washington Redskins had the lowest DVOA of any Super Bowl champion until the Cinderella run of the 2007 Giants. Only seven teams since 1987 have won the Super Bowl despite a regular-season DVOA rating under 20%.

Washington had a lower DVOA rating than three of the other four NFC playoff teams as well as the defending champion New York Giants. (The Giants' season was seen as a colossal failure at the time because they started 0-5, but remember -- three of those losses were strike games. The actual Giants went 6-6 against one of the tougher schedules in the league.) Although the NFC was better than the AFC throughout the '80s, the 1987 NFC Championship Game featured two weaker regular-season teams while the AFC Championship Game featured the No. 3 Cleveland Browns and the No. 4 Denver Broncos. The Broncos won thanks to The Fumble by Earnest Byner, and the Las Vegas oddsmakers agreed with what DVOA would have said at the time: the Broncos were favored by 3 in Super Bowl XXII. Instead, Washington wiped the floor with them, 42-10.

* * * * *

Here are the Football Outsiders team efficiency ratings for 1987, 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.)

OFFENSE and DEFENSE DVOA are adjusted based on strength of opponent as well as 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-ADJUSTED TOTAL VOA does not include these adjustments. Official 1987 Football Outsiders stats do not include Week 4-6 strikebreaker games.

As always, positive numbers represent more points so DEFENSE is better when it is NEGATIVE.

• ESTIMATED WINS uses a statistic known as "Forest Index" that emphasizes consistency as well as DVOA in the most important specific situations: red zone defense, first quarter offense, and performance in the second half when the score is close. It then projects a number of wins adjusted to a league-average schedule and a league-average rate of recovering fumbles. In 1987, this number is based on a 12-game season.
• WEIGHTED DVOA is adjusted so that earlier games in the season become gradually less important. It better reflects how the team was playing at the end of the season.
• 1987 SCHEDULE lists average DVOA of opponents played this season, ranked from hardest schedule (#1, most positive) to easiest schedule (#28, most negative).
• PYTHAGOREAN WINS represent a projection of the team's expected wins based solely on points scored and allowed. In 1987, this number is based on a 12-game season.
• VARIANCE measures the statistical variance of the team's weekly DVOA performance. Teams are ranked from most consistent (#1, lowest variance) to least consistent (#28, highest variance).

DVOA for 1987 is now listed in the stats pages:

• OFFENSE
• DEFENSE
• SPECIAL TEAMS
• QUARTERBACKS
• RUNNING BACKS
• WIDE RECEIVERS
• TIGHT ENDS

The truncated 1987 schedule gave the season more teams than usual where DVOA deviated from win-loss record. The Los Angeles Raiders are the best example of this, going 0-6 in games decided by a touchdown or less. With four more games, perhaps they do more to balance that out, and the No. 9 team in DVOA is 7-9 instead of 4-8. In fact, the Raiders were 7-9 in 1988 despite being a worse team than they were in the strike year, because the 1988 Raiders were 4-4 in games decided by a touchdown or less.

The top offenses of 1987 generally match the top quarterbacks of 1987. The Miami Dolphins finished No. 2 on offense and dead last on defense in both 1987 and 1988, which is something to show anyone who ever says Dan Marino wasn't great because he never won a championship. The list of top defenses features a couple of the usual NFC East suspects: the Giants were No. 2, and the Eagles were No. 5 with Reggie White racking up 21 quarterback sacks in just 12 games. That stat may be as impressive as Rice's 22 touchdowns. The Patriots were a surprising No. 3 in defensive DVOA, and Pittsburgh was No. 4. Both defenses flew a bit under the radar since they were paired with poor offenses. The Patriots were led by Hall of Fame linebacker Andre Tippett and veterans who probably aren't well known outside of New England: linebacker Steve Nelson, cornerback Raymond Clayborn, safety Fred Marion. Tippett was the only Pro Bowler, but Pittsburgh didn't even have a Pro Bowler on defense. The Steelers still had a few players left from the final Super Bowl years, such as cornerback Dwayne Woodruff, safety Donnie Shell, and nose tackle Gary Dunn. The defensive coordinator was Tony Dungy, but the defense was not how you think of a Tony Dungy defense now; it was a 3-4 alignment that played strong pass defense despite very little pass pressure. The Steelers were tied for 25th out of 28 teams with just 23 quarterback sacks in non-strike games.

The team with the fewest quarterback sacks was the worst team in the NFL in 1987, the Atlanta Falcons. They had just 11 sacks in non-strike games, so not even an average of one per game. They also had the worst offense in the league, and only threw 12 passing touchdowns in the non-strike games. The combination of terrible offense and terrible defense put the Falcons way behind the rest of the league in DVOA, essentially the opposite of the 49ers -- which means both teams rank among the most extreme in DVOA history.

You might also notice New Orleans as the best special teams unit of 1987. That's mostly recently elected Hall of Famer Morten Andersen, ranking third in field goal value and second on kickoffs despite being penalized for playing so many games indoors. The 1987 Saints also got strong punting from Brian Hansen and punt returns from USFL refugee Mel Gray, who was eventually ranked fifth when NFL Network did a show on the Top 10 Return Aces.

So, what about those strikebreaker games of Weeks 4-6? Well, here are those numbers as well. OFFENSE and DEFENSE VOA are not adjusted for opponents but do 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.

REGULAR GAMES RANK lists the team's Total DVOA rank during Weeks 1-2 and 7-16, from the table above.

There's a surprisingly high correlation coefficient of .34 between each team's rank in DVOA in the regular games and their rank in VOA in the strike games. That's some combination of star players crossing the picket line and better-managed teams also preparing better for the strike and the use of replacement players. There are three clear exceptions at the bottom of the table. The Giants completely ignored scouting players for the replacement roster, apparently banking on the strike not lasting more than a couple of weeks. Buddy Ryan of the Eagles, as Mike Tanier detailed in his "Year of the Scab" essay, was philosophically on the side of labor and really didn't want to participate in building a scab team. Marv Levy was probably so busy rebuilding the actual Bills roster that he didn't pay much attention to the replacements; Buffalo's one win was over the Giants, 6-3 in overtime in the last week of the strike.

On the other hand, Tom Landry and Joe Gibbs both devoted resources to scouting replacement players and got good results out of those three weeks. Meanwhile, the Colts had Gary Hogeboom crossing the picket line right from the get-go, and walloped Buffalo 47-6 in the first strike game. Head coach Ron Meyer actually took Hogeboom out of the game in the second half to preserve Johnny Unitas' then-team record of 401 passing yards. San Diego, on the other hand, mostly just got lucky during their strike games. All three games were on the road, but the Chargers won all three by less than a touchdown: 10-9, 17-13, and 23-17. The Chargers started the season 8-1 including the strike games, then lost six straight and missed the playoffs.

In order to emphasize the quality of the football with replacement players, computing VOA for the strike games uses the same baselines that I used to normalize the average of the regular 1987 season at 0%. That means that for the strike games, the league average is below 0% for all three units as well as total VOA. Overall, offensive VOA for the entire league was 14.0% lower during the strike games, while defensive VOA was 10.0% lower. (The difference comes because there were more of the plays I count only for offense such as aborted snaps and false starts.)

However, the units that took the biggest hit during the strike were special teams. That makes sense for a number of reasons. With practice time limited for new teammates, I have to imagine that there was almost no time available to practice special teams. There's also a steep slide in the quality of kickers after the first few. When I talk to people about the quality of kickers in college, one of my go-to lines is "There are about 40 guys good enough to kick in the NFL, and there are only 32 jobs. But only two or three of the other eight guys are currently between the ages of 18 and 23." It's a bit of an exaggeration, but not far off. So the 1987 strike showed what happened when you took a bunch of those guys who weren't good enough to kick in the NFL and stuck them on teams anyway. Special teams already were much worse in 1987 than they are now. In the regular 1987 games, kickers hit 71.5 percent of their field goals. In the strike games, that dropped to 63.9 percent. Punting was also horrible, with only five of the 28 teams coming out with net punting value above the average baseline for the rest of 1987. The averge net punt went from 32.9 yards during the rest of 1987 to 30.1 yards during the strike. Overall, the average strikebreaker team had -7.2% DVOA on special teams.

The Houston Oilers dominated special teams during the strike primarily by forcing four fumbles on opponent punt returns and getting a phenomenal performance in two games from scab kicker John Diettrich. Diettrich had the advantage of playing one of the two games in Denver, but the Ball State product went 6-for-6 on field goals, including three from 40 yards out or longer. In a high-scoring game with the Broncos, Diettrich put 8 of 9 kickoffs into the end zone with four touchbacks; the next week against Cleveland, he got a touchback and two kickoffs deep enough that the Oilers pinned the Browns at the 20 and 17. However, regular kicker Tony Zendejas and punter/kickoff man Lee Johnson crossed the picket line for the final week, and Diettrich never played in the NFL again.

Now let's take a look at the best and worst players by position:

Quarterbacks: Bernie Kosar narrowly edges out Dan Marino on the top of our quarterbacks table in both DVOA and DYAR. Both quarterbacks excelled even though Cleveland and Miami faced two of the five toughest offensive schedules in the league. (Miami's was tougher against the run than the pass, which is why Kosar gets a bigger boost from opponent adjustments than Marino does.)

The DVOA table is a who's who of legendary 1987 quarterbacks, with a DYAR top 10 that includes Montana, John Elway, Neil Lomax, Boomer Esiason, Phil Simms, and Jim Kelly. The surprise is Bill Kenney, who ranks seventh. Bill Kenney? I admit I don't know much about this guy, but there's definitely a story there. Kenney was a 12th round pick in 1978 and found his way to Kansas City in 1980. From 1980 he started 77 games in nine seasons for the Chiefs, but only once did he actually start all 16 games in a season. The Chiefs drafted Todd Blackledge in 1983 to take Kenney's job, and Kenney ended up having his best season and made the Pro Bowl. It looks like every year after that the Chiefs would try to make it work with Blackledge, but Blackledge would be awful, and Kenney would end up starting, and then Kenney would get hurt, and Blackledge would end up starting, and back and forth it would go. Sometimes when both quarterbacks were healthy, the Chiefs would swap them out throughout the game, playing Blackledge on running downs and Kenney on passing downs. They finally gave up in 1988 and traded for Steve DeBerg.

Down the list you'll find a couple of legends at the end of their careers. Dan Fouts was slightly below average in his final season, and Danny White was a bit below that in his last year as a starter. (He did play in three games as a backup in 1988.) No. 1 overall pick Vinny Testaverde was horrendous, completing just 43 percent of passes in his first six NFL games for Tampa Bay. Chuck Long, the 12th overall pick of the 1986 draft, wasn't much better for the Detroit Lions. But no quarterback in 1987 could come close to the brutal performance put up by Mark Malone. Malone completed just 46 percent of his passes and threw 19 interceptions (one on a Hail Mary) but just 6 touchdowns. John Stallworth had 521 receiving yards in his final season, and nobody else on the Steelers had more than 350. Even with just 12 games in the season, that's awful. Malone and the Steelers passing attack put up these numbers despite facing the No. 26-ranked schedule of opposing defenses. As a result, Malone had minus-39.0% passing DVOA when no other quarterback with at least 200 passes was below minus-25%. and he had minus-621 passing DYAR which was almost triple that of any other below-replacement level quarterback. It ranks as one of the 20 worst quarterback seasons in our database. Steelers fans booed Malone mercilessly in the first game after the strike, chanting "Go home, Malone." There was a fan banner that said "Malone, Stay on Strike." I have no idea how the Steelers went the whole year without putting second-year quarterback Bubby Brister in the lineup. They had drafted him in the third round the year before, and he replaced Malone the following season.

Two quarterbacks dominated during the strike games in Weeks 4-6. One of them was Joe Montana. You know, one of the top five quarterbacks to ever play the game, crossing the picket line to take on a bunch of random working stiffs and CFL refugees. No surprise. The other was Ed Rubbert. Who? Well, Ed Rubbert was a three-year starter at Louisville, but he didn't get selected in the 1987 draft. This is not a huge surprise, given that during his college career Rubbert completed 49 percent of his passes and threw 50 interceptions including 28 in one 11-game season. But Rubbert started three games for the Redscabs and all three ended up as wins, though Rubbert himself was injured early in the Week 6 game. Mostly, Rubbert and former Atlanta sixth-round pick Anthony Allen just went off in a Week 4 win over St. Louis. Allen had touchdown catches of 88, 48, and 34 yards plus other catches of 43 and 27 yards. The next week, the Giants did a better job of covering Allen but Rubbert still found Ted Wilson for a 64-yard touchdown. And somehow, Rubbert only threw one pick in his three scab starts, though it was a pick-six.

I also should acknowledge one great game that Cleveland backup quarterback Gary Danielson (yes, Verne Lundquist's longtime SEC broadcast partner) had when he crossed the picket line for Week 6 against the Bengals. The Browns won 34-0 and Danielson went 25-for-31 for 274 yards with four touchdowns and no picks, a single game worth 230 YAR. And you'll find a couple of interesting names at the bottom of the "secondary quarterbacks" table: New Orleans head coach Sean Payton in his limited action for the Spare Bears and wide receiver Guido Merkens attempting to play quarterback for the Eagles.

Running Backs: Looking just at advanced stats and the 12 games with regular players, the best running back of 1987 was Curt Warner of the Seattle Seahawks. The Seattle running game was overall the best in the league, thanks to the combination of Warner and fullback John L. Williams. Warner led the league in rushing DYAR, also ranking fifth in success rate and sixth in DVOA among qualifying backs (minimum 80 carries). He also did well in limited receiving targets, with 17 catches on 20 targets for 167 yards and 48.4% DVOA. Williams was fourth in the league in DYAR, third in DVOA and No. 1 with a fabulous 69 percent success rate. He also was the No. 2 most valuable receiving running back, trailing only Herschel Walker. People may forget now what an outstanding receiving back Walker was: he had 837 receiving yards in 1986, his first NFL season, and then 715 receiving yards in only 12 games in 1987. (That pro-rates to 953 yards over 16 games, which would be the fifth-highest receiving yardage total ever for a running back.)

Warner may have led the league in rushing DYAR, but he didn't lead the league in rushing yardage. That's a complicated story revolving around what, in retrospect, looks like a colossal error by the Los Angeles Rams. The story actually starts back in 1979, when USC running back Charles White won the Heisman Trophy. The Browns selected him 27th overall in the 1980 draft, but he was a massive bust. White never ran for more than 350 yards in a season for the Browns, and had significant drug problems. (It was the '80s, so that meant cocaine, not marijuana.) White's former USC head coach John Robinson was in charge of the Rams, so when the Browns cut White in 1985, Robinson brought him to Los Angeles to back up superstar Eric Dickerson.

The cocaine use didn't stop. On August 21, 1987, the police arrested White in a cocaine haze in a vacant lot near the Rams training complex in Fullerton. ("Brandishing a trash can lid at the time," reported the Los Angeles Times.) Robinson somehow talked the NFL out of a major suspension and so White started the season as Dickerson's backup. In the first two weeks, White carried the ball twice, once for a loss of 4 yards and once for a loss of 5 yards. Then the strike came. White crossed the picket line from the get-go and started all three strike games. He put up big totals in those games, with a combined 339 rushing yards. And somehow... the Rams decided based on those three strike games that White would be good enough to replace Dickerson, and they dealt Dickerson to Indianapolis a week after the strike ended. When White went on to lead the league in every rushing category, it looked like they made the right decision. But the advanced stats show that White's big totals were entirely based on big workloads. White had a below-average DVOA and success rate. He only averaged 4.1 yards per carry in non-strike games. He had negative receiving value. He wasn't even that great in the three strike games! He didn't lead running backs in either DYAR or total yardage during the strike. Roger Craig crossed the picket line after one week and his two games gave him the most DYAR during the strike. And some guy named Lionel Vital actually outgained him during the strike, 346 yards to 339 yards. Vital was a 1985 seventh-round pick of the Redskins out of Nicholls St. who only never played a regular-season game other than the three 1987 strike games.

White followed up his big Pro Bowl comeback year with a positive alcohol test, which got him a four-game suspension because of his prior cocaine offenses. When he came back, he had lost the starting job to ex-Bills starter Greg Bell, and White retired after the 1988 season. Eric Dickerson, meanwhile, was the real rushing yardage champion of the 1987 season, with 1,288 yards in the 12 non-strike games. He only finished 10th in DYAR because he played a very easy schedule of opposing defenses.

There were a number of other notable running back performances in 1987, starting with Bo Jackson's first NFL season. Jackson had 81 carries, which barely made it over our reduced qualifying baseline to be ranked. Still, he led all qualifying running backs in rushing DVOA and finished sixth in rushing DYAR. Johnny Hector of the Jets finished second in rushing DYAR despite an awful 46 percent success rate, thanks to 11 touchdowns against a difficult schedule. Bengals fullback Larry Kinnebrew finished third in rushing DYAR. You may ask, "Why didn't I read about Kinnebrew when Football Outsiders reviewed the powerful Bengals offense that led the league in DVOA in 1988?" Well, Kinnebrew thought the Bengals weren't respecting him enough in contract negotiations so he held out of 1988 training camp. The Bengals decided Ickey Woods was better than Kinnebrew and he missed out on a Super Bowl appearance. By the way, the contract that Kinnebrew turned down for 1988 was worth $250,000. He probably should have held out for 30 years.

The bottom of the rushing DYAR table has a very surprising name: Roger Craig finished 47th out of 48 running backs in DYAR. His rushing performance was the weakness of that otherwise powerful San Francisco offense, as he averaged just 3.8 yards per carry with five fumbles. Craig had an offseason as a receiver, too, barely coming out above replacement level. However, the worst running back of 1987 was Pittsburgh's Walter Abercrombie. Like Mark Malone, he had a subpar performance against a very easy schedule. The Steelers offense really was hot garbage that season.

Other fun names: Walter Payton finished 32nd in rushing DYAR in his final season. And down on the list of strikebreaker running backs, you'll find Joe "What The Heck, Why Not" Dudek, made famous by a 1985 Sports Illustrated cover that promoted the record-breaking running back of Division III Plymouth State over Bo Jackson and Chuck Long for the Heisman Trophy.

Wide Receivers: We covered Jerry Rice and Anthony Carter above. Al Toon of the Jets was third in DYAR, with Eric Martin of the Saints and Mark Clayton of the Dolphins rounding out the top five. It will not surprise you to learn that the lowest-rated receiver of the year came from Pittsburgh: it was John Stallworth, at minus-135 DYAR. Two other receivers were below minus-100 DYAR despite not even making it to the 40-target baseline we used for our main ranking table. One was Pittsburgh's No. 3 receiver, a third-round rookie named Charles Lockett who caught just 7 of his 35 targets. The other was former CFL star Mervyn Fernandez, in his first NFL season with the Los Angeles Raiders. Fernandez got better, but Lockett was out of the league after 1988.

The top-rated wide receiver from the strikebreaker games was Kelvin Edwards of the Cowboys. Edwards was a fourth-round Saints draft pick out of Liberty University in 1986, but the Saints cut him before the 1987 season. He signed up with the Cowboys when the strike happened and starred with 14 catches for 272 yards and 3 touchdowns during the strike. It was enough to keep him on the roster for the rest of the year, and he started six more games after the strike... but you may notice that he finished next-to-last in DYAR among qualifying wide receivers. His NFL career was over after five more catches in 1988.

A number of the top receivers of the strike games disappeared afterwards, and sometimes it was a bit of a surprise. Most of the strike players were low-round picks or undrafted, but some had strong draft or USFL pedigrees. San Diego's Al Williams was a 1,000-yard receiver in the USFL for the Oklahoma/Arizona Outlaws and the 20th pick in the 1984 supplemental draft, but his three strike games (12 catches, 247 yards, 1 TD) were the only time he ever played in the NFL. Walter Murray was a 1986 second-round pick out of Hawaii, traded by Washington to Indianapolis when he wouldn't sign with the Redskins. He had 14 catches, 256 yards and 3 touchdowns during the strike, but only 8 other career catches and he never played again after 1987.

But the best wide receiver performance of the strike games belonged to the legendary Seattle wide receiver Steve Largent. Edwards earned 117 YAR in three games during the strike, but Largent earned 104 YAR in less than three quarters of football. With the strike essentially over in Week 6, Largent crossed the picket line along with backup quarterback Jeff Kemp, and together they thoroughly embarassed the replacement Detroit Lions defense. Kemp threw three touchdowns to Largent before the first quarter was even over. When Largent came out of the game with 9:26 left in the third quarter, he had caught 15 of 19 targets for 261 yards. Thirteen of his 15 catches earned first downs or touchdowns. Asked after the game why he had not used double coverage on Largent, Lions coach Darryl Rogers told the press, “Why embarrass two guys when you can just embarrass one?” So, where would Largent's game would fit among the best DYAR games of all-time? Obviously, if we did count an opponent adjustment for playing the scab Lions defense it would be huge, but even without an adjustment this game surprisingly doesn't come in among all-time top 25. The games that come in higher generally have a higher yards per reception average or get a big boost from opponent adjustments. The two best DYAR games in history qualify both ways (Flipper Anderson's 15-of-20, 336 yards, 1 TD against the 1989 Saints and Jimmy Smith's 15-of-21, 291 yards, 3 TD against the 2000 Ravens).

Tight Ends: As great as Jerry Rice was in 1987, the best receiving performance may have belonged to New York Giants tight end Mark Bavaro. Bavaro blew away the rest of the league's tight ends that season with 343 receiving DYAR and 65.9% DVOA. He had 55 catches for 865 yards and eight touchdowns plus a 75 percent catch rate against a tougher than actual schedule. By the way, he was also one of the top blocking tight ends in football, capable of taking on Reggie White without help. It's no wonder Bill Belichick talks about Bavaro in the same glowing language he uses to talk about Rob Gronkowski. Bavaro's total of 343 receiving DYAR is enough to rank as the No. 5 tight end season of the DVOA era even without pro-rating it for 16 games. Pro-rated, it would be worth 457 DYAR, coming just short of topping the No. 1 season in DYAR history, Gronk's 461 DYAR in 2011. In terms of DVOA (per play) rather than DYAR (total), the only season better than Bavaro's 1987 was Antonio Gates' injury-shortened 10-game 2010 campaign at 77.1% DVOA.

Other than Bavaro, the top tight end of 1987 was the late Todd Christensen of the Raiders. Christensen re-defined what a tight end could look like. I'm not talking about size and speed, I'm talking about this amazing perm-and-moustache combination. Third in receiving DYAR for tight ends was Minnesota's Steve Jordan, the pride of Brown University (and father of Cameron). Pat Beach of the Colts and Mickey Shuler of the Jets were at the bottom of the ratings.

During the strike, teams shied away from using tight ends very much. Only seven different tight ends had at least 12 targets during the three strike games. One tight end was far ahead of the rest: Bobby Micho of Denver, a 1984 tenth-round pick who had never caught a regular-season pass until the strike, although he did have one 20-yard reception during the 1986 postseason. Micho crossed the picket line for all three games and caught 24 of 28 passes for 236 yards and 2 touchdowns. The rest of the season in Denver, he caught one pass for six yards, and that was the end of his NFL career.

Here are a few more fun tidbits about the 1987 season:

• The Raiders and Chiefs played in Week 4 in 106-degree temperatures in Los Angeles. Between the heat and the strike, attendence was only 10,708, which is low enough for everybody to fit into the StubHub Center.
• Buddy Ryan was so pissed off about the strike and a 41-22 loss in Week 5 that two weeks later, in the first game after the strike, he had Randall Cunningham in the game in the final seconds to run up the score on Tom Landry and the Cowboys. The Eagles were up 30-20, and scored to finish the game at 37-20.
• Kellen Winslow was roughly average in DVOA in his final NFL season, but was involved in a very strange play when the Chargers lost to the Broncos in Week 12. The Chargers drove 64 yards to the Denver 1-yard line in the final minute of the first half, yet they did not score. From the Los Angeles Times:

On third-and-goal from the Bronco seven-yard-line, with 17 seconds and no Charger timeouts left, here's what happened: Charger tight end Kellen Winslow ran to the end zone, Fouts threw a ball into the end zone, and Winslow caught the ball. But not all at the same time. Running a simple outside pattern, Winslow leaped and seemed to catch the ball at the six-point side of the goal line. But Denver defensive backs Mike Harden and Randy Robbins were hanging on his back, and pushed him, in mid-air, back to the other side of the goal line. Winslow landed on the one-yard line and began fighting to get across. Denver defenders Steve Wilson and Jeremiah Castille joined the fray, making it four on one, with Winslow going nowhere. By now, there were just five seconds left and, with no timeouts, the Chargers couldn't get their field goal unit on the field in time. "I thought I scored," said Winslow, "but what I think doesn't matter. I was in the end zone. The only thing that could have happened was, I might stepped back to the ball and gone back over the line." Saunders said: "We were yelling for a replay, but they were so slow to spot the ball afterward, we had to stop yelling and get our field goal team out there, and by then it was too late. We practice that emergency field goal team thing everyday. If they had spotted the ball quicker, we might have had a chance."

• Once again, Jeremy Snyder put together a Year in Quotes from 1987 for us. You can read it all here. Thanks again to Jeremy for all his work in transcribing these old gamebooks and watching old game tapes to fill in play-by-play on games that no longer have available gamebooks.

Housekeeping notes:

• The next step will be to finish up 1986 and get that posted. I'm guessing we'll have that in mid-April, along with an extra article about stat errors and the experience of trying to make old play-by-play logs agree with currently listed career totals. If you see our numbers disagree with the official 1987 totals, even for players who only played in 12 non-strike games, this is probably the reason why.
• As I noted when I put up 1988 DVOA a month ago: The 1989-1990 DVOA ratings were never added to the FO player pages, so at some point this offseason there will be a gigantic upload that will add 1986-1990 player DVOA and DYAR to all the player pages, along with 2016 DVOA and DYAR, and also create new player pages for those players who retired prior to 1991. This is why many players do not have their names in the 1987 player stats tables link to player pages.
• Once 1986 is finished, we will also add all three of the new seasons (1986-1988) into the FO Premium database.
• We still need to add LAST YEAR listings to 1988 and 1989 and make some fixes on 1989 special teams. There are also going to be some small fixes to 1988, issues that I found with certain aborted snap plays that had their yardage counted wrong.
• Once all of that is done, we move on to the next project: 1983-1985. We have all of 1985 except for one quarter: the first quarter of the Week 6 Colts-Broncos game has an unreadable mimeograph in the official gamebooks in both the Broncos and NFL archives. If we can't find a record of it elsewhere or a videotape, we'll need to piece it back together based on game stats, but we'll do our best. As for 1984, there are two games where we do not have any gamebook or video: Week 13 Bengals-Falcons and Week 14 Bills-Colts. If you know anything about a tape of one of these games, please let us know. For 1983, apparently we have everything except the first drive of the Week 15 Bengals-Lions game.