Captain Comeback: Clutch NFL Teams Regress to the Mean

Cold, Hard Football Facts for Aug 28, 2012



By Scott Kacsmar
Cold Hard Football Facts' Comeback King (@CaptainComeback)

Captain Comeback makes an official return for the 2012 season. But before we get back to analyzing the week’s close finishes and cracking skulls over semantics, it is time for a regression study utilizing some of the advanced historical data we did not have last season.

It was an especially thrilling year with a few teams—namely the Super Bowl champion New York Giants, San Francisco 49ers, Denver Broncos and Arizona Cardinals—pulling out a rare and large number of close wins in the fourth quarter and overtime.

After studying similar teams from NFL history, the chances of a repeat performance in the clutch in 2012 are slim to none. As if repeating wasn’t already going to be hard enough for the Giants.

ccWhen a NFL team has a season with a lot of close wins, the media loves to crown them as “a team that knows how to finish.” For example, the Giants tied NFL records with Eli Manning leading seven fourth quarter comebacks and eight game-winning drives.

But what happens when the following season comes, and that same team is now struggling to win the close ones?

Did they “forget” how to finish that fast?

The fact is that winning a lot of close games consistently is very difficult, because so often the outcome can hinge on one play. Whether it is Billy Cundiff blowing the field goal (and apparently his career) or Kyle Williams fumbling another punt in overtime, the luck factor can really outweigh the skill.

Here we will look at the regression study for the teams that loaded up on close wins, and what that means specifically for 2012.

The four aforementioned 2011 teams with a lot of close wins could be in for a major regression this season when the games get tight again.

And we know most games are close in the NFL.

 

The Study

This study should give a better sense of the close-game quandary than analysis that relies on the final score of games. Too often garbage-time scores can make a game closer than it really was, or create the mirage of a comfortable win a la Super Bowl XLIV.

Combing through the database, there have been 48 teams in NFL history that met the desired criteria of having at least six wins (no ties) result from a fourth quarter or overtime game-winning score.

Two key caveats: the playoffs were included for all teams, and the wins can be in any form, whether it is a normal offensive drive, a return touchdown, a non-offensive field goal drive, or a rare game-winning safety.

With Captain Comeback we are always talking about the comeback and game-winning drive being done by the quarterback/offense. In this study, we are looking at teams, so any way the team could win the game with a fourth quarter or overtime score counts.

Besides, one could definitely argue it is more improbable to win on a punt return touchdown in overtime than it is to lead a common game-winning drive for a field goal.

The 48 teams include our four from last year (NY Giants, San Francisco, Arizona and Denver), so they were obviously excluded from the regression study. Also, since opportunity data is not up to par for pre-1980, eight more teams were excluded, giving us 36 in the sample.

The results are very telling of the regression.

Year N is the initial season when the team had at least six late-game wins, and N + 1 is the following season.

The “Win %” is the overall season record for these teams, which again includes postseason.

“4Q/OT Wins” is the average number of wins the teams had via a fourth quarter or overtime game-winning score.

The “4QC Win %” is the record the teams had in fourth quarter comeback opportunities, defined here as having the ball when trailing by one score.

The “4Q/OT Win %” is the team’s record in all opportunities in the fourth quarter or overtime when the score is tied or a one-score difference.

Year

Win %

4Q/OT Wins

4QC Win %

4Q/OT Win %

N

0.683

6.38

0.615

0.676

N+1

0.581

2.75

0.349

0.399

While the teams regress overall in record by about 1.6 wins the following season, the standout numbers are the large declines in their ability to win the tough games with comebacks and game-winning drives the following season.

At fourth quarter comebacks, the teams went from a record of 163-102 (.615) down to 74-138 (.349) the next season.

Just like that they went from the top of the elite to roughly average. Regression to the mean at its finest.

Here are the 36 teams with three comparisons (Year N vs. N + 1) between their season record (Rec.), fourth quarter comeback opportunity record (4QC), and overall fourth quarter/overtime opportunity record (4Q/OT).

The last three columns show the change in win percentage in each category, with any improvements in red.

 

Year N: 6+ 4Q/OT Wins

Year N + 1

Regression

Team

Year

Rec.

4QC

4Q/OT

Rec.

4QC

4Q/OT

Rec.

4QC

4Q/OT

Miami

1985

13-5

4-2

6-2

8-8

1-1

2-2

-0.222

-0.167

-0.250

Denver

1985

11-5

6-3

7-4

13-6

2-2

2-2

-0.003

-0.167

-0.136

Cleveland

1986

13-5

5-2

7-2

11-6

1-5

1-6

-0.075

-0.548

-0.635

New Orleans

1987

12-4

4-2

6-2

10-6

1-4

2-4

-0.125

-0.467

-0.417

Green Bay

1989

10-6

5-4

7-4

6-10

2-6

2-6

-0.250

-0.306

-0.386

San Francisco

1989

17-2

4-2

6-2

15-3

4-3

6-3

-0.061

-0.095

-0.083

Dallas

1990

7-9

4-4

6-4

12-6

5-2

5-2

0.229

0.214

0.114

San Francisco

1990

15-3

4-3

6-3

10-6

2-5

2-5

-0.208

-0.286

-0.381

Atlanta

1991

11-7

5-3

6-3

6-10

1-5

2-5

-0.236

-0.458

-0.381

Indianapolis

1992

9-7

4-0

6-1

4-12

2-2

4-2

-0.313

-0.500

-0.190

Miami

1992

12-6

6-2

6-2

9-7

3-3

3-3

-0.104

-0.250

-0.250

Detroit

1993

10-7

3-4

6-4

9-8

2-5

4-5

-0.059

-0.143

-0.156

Buffalo

1993

14-5

5-2

6-2

7-9

0-6

1-6

-0.299

-0.714

-0.607

Arizona

1994

8-8

4-5

6-5

4-12

3-3

3-4

-0.250

0.056

-0.117

San Diego

1994

13-6

6-3

6-3

9-8

2-4

3-4

-0.155

-0.333

-0.238

Pittsburgh

1995

13-6

3-3

6-3

11-7

0-4

2-4

-0.073

-0.500

-0.333

Buffalo

1997

6-10

2-5

6-5

10-7

2-6

2-6

0.213

-0.036

-0.295

Arizona

1998

10-8

5-3

7-3

6-10

2-5

2-5

-0.181

-0.339

-0.414

Indianapolis

1999

13-4

6-3

7-3

10-7

2-5

3-6

-0.176

-0.381

-0.367

Baltimore

2001

11-7

4-4

6-4

7-9

1-5

1-5

-0.174

-0.333

-0.433

Tampa Bay

2001

9-8

3-5

6-5

15-4

1-3

2-3

0.260

-0.125

-0.145

Philadelphia

2003

13-5

4-2

6-2

15-4

2-1

2-1

0.067

0.000

-0.083

New England

2003

17-2

4-1

7-1

17-2

0-1

2-1

0.000

-0.800

-0.208

Carolina

2003

14-6

5-4

8-5

7-9

2-5

3-6

-0.263

-0.270

-0.282

Pittsburgh

2004

16-2

6-0

8-0

15-5

2-4

2-4

-0.139

-0.667

-0.667

Dallas

2005

9-7

5-5

6-6

9-8

2-4

3-5

-0.033

-0.167

-0.125

Jacksonville

2005

12-5

4-3

6-3

8-8

0-6

1-6

-0.206

-0.571

-0.524

Seattle

2006

10-8

5-5

6-6

11-7

3-3

3-4

0.056

0.000

-0.071

NY Giants

2007

14-6

6-2

7-2

12-5

3-1

3-1

0.006

0.000

-0.028

Indianapolis

2008

12-5

5-2

8-3

16-3

7-2

7-2

0.136

0.063

0.051

Pittsburgh

2008

15-4

5-3

6-3

9-7

2-5

3-7

-0.227

-0.339

-0.367

Tennessee

2009

8-8

3-2

6-3

6-10

1-5

2-7

-0.125

-0.433

-0.444

Indianapolis

2009

16-3

7-2

7-2

10-7

0-5

2-5

-0.254

-0.778

-0.492

New Orleans

2009

16-3

3-1

6-2

11-6

4-5

5-5

-0.195

-0.306

-0.250

NY Jets

2010

13-6

4-4

6-4

8-8

4-3

4-3

-0.184

0.071

-0.029

Atlanta

2010

13-4

5-2

6-2

10-7

3-4

3-4

-0.176

-0.286

-0.321

 

Here are some facts:

  • In Year N, 26 of the 36 teams had a winning record at fourth quarter comebacks. Six were .500, and four had a losing record.
  • In Year N + 1, only seven of the 36 teams had a winning record at fourth quarter comebacks. There were 23 teams with a losing record, and six were .500.
  • In Year N, 34 of the 36 teams had a winning record at overall fourth quarter/overtime wins. Only the 2005 Dallas Cowboys (6-6) and 2006 Seattle Seahawks (6-6) were stuck at .500.
  • In Year N+ 1, only eight of the 36 teams had a winning record at overall fourth quarter/overtime wins. There were 24 teams with a losing record, and four teams were .500.

 

Even if you remove the teams that failed to make the playoffs the following season (leaving 20 teams), the overall fourth quarter/overtime record drops from 130-61 (.681) down to 47-69 (.405).

Trying to find improvement is very difficult. Only seven teams had a better overall record on the season the following year, while the 2004 New England Patriots had a consecutive 17-2 campaign as the last team to win back-to-back titles.

Only four teams had a better win percentage at fourth quarter comebacks. Slight improvements by the 1995 Arizona Cardinals and 2011 New York Jets are meaningless. Despite losing more games than in the previous season, they had fewer comeback opportunities overall, which meant they were playing too poorly to be close enough in the fourth quarter.

The 1991 Dallas Cowboys and 2009 Indianapolis Colts are the only two teams to see improvement across the board in all three records. That was the start of Dallas’ six-year playoff run, while Peyton Manning earned his last two MVP awards with a ridiculous two-year stretch of pulling out 15 clutch wins.

Only the 2008-09 Colts and 1989-90 49ers, which similarly featured Joe Montana winning consecutive MVP awards, managed to have back-to-back seasons with at least six fourth quarter/overtime wins.

While improvement is rare, there are some fascinating declines.

Despite “The Drive” wiping out an 8th game-winning drive, the 1986 Cleveland Browns were 7-2 at fourth quarter/overtime wins. Even with Bernie Kosar’s career season in 1987, Cleveland regressed to just 1-6 in such games, including “The Fumble” against Denver in another AFC Championship loss.

After a fourth consecutive loss in the Super Bowl, the Buffalo Bills went from 6-2 in their fourth quarter/overtime opportunities in 1993 to 1-6 in 1994, missing the playoffs in the process. Their 5-2 record at fourth quarter comebacks turned into 0-6.

Even though they would rally for a Super Bowl win, the 2005 Pittsburgh Steelers were a far cry from their dominant 8-0 record in the tough games in 2004. They were only 2-4 the next season, but played with the lead throughout an eight-game winning streak to end the season. Tommy Maddox, third-string quarterback, did have two awful performances in overtime losses to bring the record down in 2005.

Finally, while Peyton Manning makes the list three different times, the Colts suffered one of the greatest regressions after starting the 2009 season 14-0 and 7-0 in fourth quarter comebacks. After Curtis Painter failed to save the day in their first loss against the Jets, Manning’s failed comeback in the Super Bowl left the Colts at 7-2. The next year, they would go 0-5 at comebacks, preventing Manning from tying Dan Marino for the all-time record.

If that wasn’t enough, the 2011 Colts went 0-7 to start the season at comebacks, keeping up a franchise-worst streak of 0-14 at fourth quarter comebacks. They finally snapped it with a Week 16 win over Houston.

From 7-0 to 0-14, that is a 7-14 record (.333) over the last three seasons for the Colts. In other words, regression to the mean has been well at work. Your turn, Andrew Luck.

 

General Season Results

Here is data on how all of the teams generally performed in the regular season and postseason.

6+ 4th QT/OT Wins

Next Season

Record: 560-274-3 (.671)

Record: 427-313 (.577)

Teams

48

Pct.

Teams

44

Pct.

Won Championship

7

14.6%

Won Championship

3

6.8%

Lost Championship

5

10.4%

Lost Championship

4

9.1%

Lost Conf-Champ.

10

20.8%

Lost Conf-Champ.

3

6.8%

Lost Divisional

10

20.8%

Lost Divisional

5

11.4%

Lost Wild Card

5

10.4%

Lost Wild Card

8

18.2%

No Playoffs

11

22.9%

No Playoffs

21

47.7%

 

A few noteworthy facts:

  • Overall, the win percentage dropped about 10 percentage points the following season.
  • Four of the last five Super Bowl champions had at least six 4th QT/OT wins, and the Captain should not have to tell you which one did not.
  • In the next season, 29 of the 44 teams (65.9 percent) failed to make the playoffs or get past the Wild Card round. Only 13 of the teams (29.5 percent) won at least one playoff game.

 

Excluding this year’s Cardinals, half of the 10 teams that failed to make the playoffs despite at least six of these wins made the playoffs the next season. The 1986 Denver Broncos made it to the Super Bowl, losing to the New York Giants.

That means 16 teams made the playoffs the first year and failed to return the next season.

 

Pre-1980 Teams

These are the eight teams without complete data for the regression study.

 

Team

Year

Record

4Q/OT Wins

Next Year

4Q/OT Wins

San Francisco

1957

8-5

6

6-6

3

Philadelphia

1960

11-2

6

10-4

3

Pittsburgh

1963

7-4-3

6

5-9

2

Atlanta

1978

10-8

7

6-10

5

New England

1978

11-6

6

9-7

2

LA Rams

1978

13-5

7

11-8

3

Houston

1978

12-7

7

13-6

4

Cleveland

1979

9-7

7

11-6

4

 

Norm Van Brocklin became the first quarterback to lead six fourth quarter comebacks in a season in 1960 with the Philadelphia Eagles, including a NFL Championship comeback win over Vince Lombardi and the Packers. He retired and gave way to Sonny Jurgensen, who managed three comebacks and game-winning drives the following season.

The 1963 Pittsburgh Steelers are a real oddity. They had nine games that featured a go-ahead (eight times) or game-tying score in the fourth quarter. But without overtime, they went 6-0-3 in those games, and 7-4-3 on the season.

Ed Brown led the offense to four traditional comeback wins, while also coming back two more times to force a tie. The Steelers had two more comebacks done with return scores. Finally, they had a go-ahead field goal drive in a game that eventually ended in a tie.

The rules to open up the passing game in 1978—not to mention the expansion to a 16-game season—sparked an interesting spike in fourth quarter wins that season. A whopping four teams made the playoffs on the back of close wins.

Believe it or not, Dan Pastorini is the first quarterback in NFL history to lead seven game-winning drives in a season (1978). The Houston Oilers would reach the AFC Championship in back-to-back years, but lost to Pittsburgh each time.

 

2012 Regression Warnings

We have four teams this season that will try to be one of the outliers that beat the close-game regression the following season. Here is a recap of what they face, and there may actually be some hope for two of them.

 

New York Giants (13-7 overall, 7-5 at 4QC, 8-5 overall 4Q/OT wins)

The Giants played a lot of close games last season, which is part of the reason they had a record-tying amount of comebacks and game-winning drives, including such wins in the NFC Championship and Super Bowl XLVI. However, this season also included failed comebacks over teams like Washington and Seattle.

If there is a silver lining for the champs, it is the fact that they have been here before. In 2008, the defending champions had a much stronger team in the regular season that did not play so many close games. They even managed a 3-1 record at comebacks, though that was misleading because of their bad performances in losses.

Of the 36 teams with complete data, they averaged 59.9 percent of their losses in Year N +1 as a result of failed comebacks/game-winning drives. The 2008 Giants had just one of five losses that way, which only ranked ahead of the 1993 Colts (two of 12).

It is hard to imagine they will lose seven games again. Though, it is hard to imagine they will sniff seven comebacks again as well.

Still, the Giants have at least shown they can get better after winning a Super Bowl, and this team has potential to do great things in 2012.

This is assuming Victor Cruz will not visit a New York nightclub this November in sweatpants and attempt to Salsa while being strapped.

 

Denver Broncos (9-9 overall, 5-4 at 4QC, 7-4 overall 4Q/OT wins)

If you wanted to boost your record in close games, adding Peyton Manning is a smart move. After having a down year in 2010 in that department, Manning should be back to playing and winning some close games with a Denver team that is hardly ready to dominate the league this year.

After all, the 2011 Broncos can lay claim to being the worst 8-8 team in NFL history, and their six-game winning streak with Tim Tebow is one of the most improbable streaks ever.

The 36-year-old Manning, after a long layoff with four neck procedures, will have his work cut out for him. But leading the team to a better record than .500 sounds more than doable for a quarterback with 11 seasons of double-digit wins, and one that often exceeds expectations for what his team’s record should be statistically.

There may not be any Tebow Zone-type miracles in Denver again this year, but fans can count on consistent quarterback play for a change.

 

San Francisco 49ers (14-4 overall, 6-3 at 4QC, 6-4 overall 4Q/OT wins)

The 49ers had a great season in Jim Harbaugh’s first year as a head coach in the NFL. However, they did a lot of things that make you very leery of them to repeat that success in 2012.

We will be looking at more of those things in great detail this week, but on the close game front, the 49ers got a career season out of Alex Smith.

Before last season, Smith had six career game-winning drives since his rookie season in 2005. In 2011, he had six to double his career total. The only other player to do that this late in his career was Steve Young.

No one is ready to claim Smith as the next Young, though he was very impressive in the comeback win over New Orleans in the playoffs. But based on past history, Smith usually fails in these situations. Maybe a coach like Harbaugh and a lot of talent was all he needed, but he remains a question mark, or at least someone that is not a sure bet this season.

The 2011 49ers were one of those rare teams that actually held a lead or at least a tie in the fourth quarter of every game they played last season, including the playoffs.

It is hard to pick the 49ers to improve any of their records, overall or close games, but they should still be the favorites in the NFC West because of how much better they are compared to the competition.

Still, the perfect storm of regression is there for San Francisco, and it will be coming this season.

 

Arizona Cardinals (8-8 overall, 6-5 at 4QC, 8-5 overall 4Q/OT wins)

Trouble is brewing in Arizona. Kevin Kolb still looks like a huge waste of money (as expected), and John Skelton is so statistically bizarre that no one should trust him this season.

Last season the Cardinals tied a NFL record with four overtime wins. Patrick Peterson had two game-winning punt return touchdowns, which is another oddity that will not be repeated.

Skelton has only played 13 games with 11 starts in his career, but already has seven fourth quarter/overtime wins, which makes no real sense. Neither does his 3-1 record when attempting at least 40 passes. That rarely happens too.

Arizona became just one of five teams since 1940 to play in 13 games decided by seven points or less. They were 8-5. All of their wins were decided on scores in the fourth quarter and overtime, which has never happened before for a team with that many wins.

Oh it sure was exciting, but this season looks more likely to be a disaster, and while that may be some of the bad preseason display talking, we know that statistically they are due for a steep decline.

Skelton has a 5-1 record at fourth quarter comebacks, but there is no reason to expect that to continue. If it does, then he is even stranger than Tebow with the way he plays so poorly for three quarters before coming alive late.

It is hard to imagine the Cardinals win more than five games in 2012.

 

Conclusion: Enjoy the Rare Seasons

For a change, a hypothesis was proven right with results that showed a clear regression to the mean for teams that win an elite-level number of close games in a season. It is not sustainable for most teams.

This does not diminish what the teams accomplished in those seasons, but it also speaks more to having good results and luck rather than real skills to “finish teams off.”

If you have an incredibly elite quarterback then you might be able to duplicate the success the following year, but those players are hard to come by.

Instead, most teams will just have to focus on becoming a better team overall, relying on fewer close wins in the process. Having to consistently pull out a lot of comebacks and game-winning drives is not a feasible, long-term strategy for a team.

The teams that have these kinds of seasons bring their fans a ton of excitement, and in some recent years, a championship as well. But enjoy it while you can, because you may never see another season like it.

Think of it as someone having a streak of good luck at Vegas. That will not last forever. Soon you will be broke, beat up, and holding a sign that says “Used to be John Skelton, Statistically Improbable Clutch Quarterback.”

 

Next week: a bonanza of quarterback career opportunity records for fourth quarter comebacks and game-winning drives. They were largely missing from last year’s articles, but will be a weekly feature. Also expect more drive stats in the only place that tries to quantify clutch with facts and not perception: Captain Comeback.

 

Scott Kacsmar is a football writer/researcher who has contributed large quantities of data to Pro-Football-Reference.com, including the only standardized database of fourth quarter comebacks and game-winning drives. He ponders: Keystone Cops or Keystone Kolb? You can visit his blog for a complete writing archive. Please send any questions or comments to Scott at smk_42@yahoo.com, or you can follow him on Twitter at @CaptainComeback.


From our partners




Team Pages
AFC East NFC
South
North
West

Connect With Us
Sign up for our newsletter to recieve all the latest news and updates...




The Football Nation Network

© Copyright 2014 Football Nation LLC. Privacy Policy & Terms of Use   Logout
Some images property of Getty Images or Icon/SMI