This has been quite the week. Four years after first quantifying a quarterback’s record at fourth-quarter comeback opportunities, I finally saw that work transfer to the TV set this week on ESPN’s First Take with this graphic:
Little did I expect what would follow. In true First Take style, right after debating whether or not Peyton Manning was the greatest QB in the history of the NFL, the next segment was fully devoted to whether or not Aaron Rodgers was still the best QB in today’s NFL. You know, ahead of the guy they just said might be the GOAT.
The surreal event of watching Stephen A. Smith and Skip Bayless hold a printout copy of my Insider article on Rodgers so they could argue about it is something I never would have expected and never will forget.
The fruits of my labor made it like Christmas morning for Bayless, as he has argued his ridiculous “lack of clutch gene” narrative — ridiculous in that no gene exists for anyone — on Rodgers for years without doing the research to support it. He has something now, just as anyone should when I first wrote about the front-running Packers before the 2011 season started. This is nothing new to long-time readers, but it took a push by ESPN to finally get the numbers out there.
So if Green Bay’s historic struggles to win these games is a story going forward, then I have done my job.
The problem is when a large audience catches on to something completely new to them, there’s going to be a strong negative reaction too. That’s what I want to address here. You can consider this version 2.0 of “The Truth About the Front-Running Green Bay Packers”
First, allow me to expose a little secret: Monday’s article was a last-second backup plan after the events of Sunday’s early games made a piece I did on the AFC null and void. So after the dramatic game ended between Green Bay and Cincinnati, I pitched a topic I’m very familiar with and have plenty of research on already.
Now, let’s understand this is a business. You need some controversial headlines that will generate clicks. Any good business will tell you that, not just ESPN. People can twist headlines all they want, but if you read the article:
I never said Rodgers is not clutch. I don’t write about the “clutchiness” of QBs. I write about what happened in clutch situations. Clutch is a history, not a skill.
I never said the 5-24 record at comebacks or 9-26 record at game-winning drive opportunities is all Rodgers’ fault. In fact, my first mention of this goes right to head coach Mike McCarthy.
“These close-game failures have been the hush-hush hallmark of coach Mike McCarthy’s otherwise successful tenure as Packers head coach. While the blame should be distributed everywhere, why are we not looking at the quarterback more?”
Here are some other direct quotes from the article that do not put the blame all on Rodgers:
“It’s always the same story for Green Bay: win big or lose close”
“Sunday was a perfect opportunity, but it was the latest in a long line of failures for the league’s best front-running quarterback and team.”
“There is some historical data to show the crunch-time disconnect in Green Bay.”
I understand the article is behind a pay wall, so not everyone was able to read it (hint: try Google). But there are claims out there on things I never wrote in the piece.
I also did not write the line “Recurring fourth-quarter failures prevent him from being NFL’s top QB” under the title, however I agree with it 100 percent. I’m not going to put Rodgers ahead of Peyton and Tom Brady, who have the gaudy stats, records, MVP awards and Super Bowl rings too. They also have a larger body of work. But the main difference comes in that I can still trust those QBs when the game does not start as planned and they have to win it late. I don’t trust Rodgers in the same fashion, which is why I had little faith he would get the go-ahead drive on Sunday in Cincinnati.
I’ve written thousands upon thousands of words on this topic before, so anyone thinking this was a knee-jerk reaction to Sunday’s game just doesn’t know my work on the topic. By the way, I’m limited to around 1,500 words on Insider, so any thought to being able to fully explain away every loss in the 9-26 record is a pipe dream.
Stephen A. Smith said he didn’t see a list of the games where Rodgers led the Packers to a fourth-quarter lead, but the defense gave it back. HOWEVVVVA, it does state this in the article:
“Of course, some of the 26 losses speak well for him. He has put Green Bay ahead seven times in the fourth quarter when trailing, only for the team to go on to lose the game. The defense is certainly deserving of blame for this.”
I make sure I cover my bases. So that’s what I wanted to say about the Insider piece itself.
As for any fan criticism or written defenses that have come from other writers this week, now I will respond to those.
I’m not as nonchalant about things as Rodgers, who responded with “Yeah, I’m not worried about that at all” when ESPN’s Jason Wilde asked him point blank about the lack of success in these games. I probably need to get that way to survive in this business, but I probably like arguing with people too much to stop completely.
There were many comments, e-mails and articles this week in response to my work. I’m not going to link to any of the articles as I didn’t see any that attacked me personally. If I did, I would have responded accordingly. I’m just going to go over some of the general faults I found.
No one’s done the same study I have done. It’s hard to compare (straight up) any past study of close games if you’re not looking at things the way I do, which is 4th quarter/OT, tied or down by one score. What I do takes an eternity for one person to compile, so I don’t think anyone could have accomplished that the last few days.
Stats in the final 5:00 – Sure, we can look at these, but that leaves out a lot of what goes into the 5-24/9-26 records. It’s not just about what you do when you’re behind, but it’s how you protect that lead or how you avoid getting into these situations late in the first place.
Win-loss record at 4QC/GWD should not be thrown away like trash – You can read my rant on this from FO here. We can take these stats and just look at how good a guy is at scoring a TD when he’s down 4-8 points in the 4Q, or scoring a FG when he’s tied or down 1-3. We can break them up that way and maybe get something useful out of that. The only reason I haven’t done it is because I’m still trying to put together a full database for every single opportunity in the last 30+ years. That takes time.
However, the record, the wins and losses (and sometimes ties), is the starting point for knowing which games to look at. We can’t just ignore it. While we can break the games down and see why the team won or lost, we need to be taking 4QC/GWD, which are situational drive stats at the heart of it all, and not just focus on the scoring drive(s).
Rodgers probably could have avoided last Sunday’s 4QC opportunity if he didn’t throw a bad INT early in the quarter in scoring territory. And people talk about the Johnathan Franklin fumble on 4th-and-1 losing the game, but I can tell you any advanced stat (DVOA, QBR, WPA, EPA) will give Rodgers two negatives for the sack on 2nd-and-6 and the 11-yard pass on 3rd-and-12 that set up that 4th-and-1 in the first place. He’s still accountable in that loss for things that took place before he was even trailing in the 4Q.
With a stat like TD passes, we don’t care about what happened on the drive before and after. It is what it is. These 4QC/GWD stats are different because what happens before and after them will usually decide if they stand up or not. Just taking a 1-point lead with 14:50 left to play does not put you in good position for a GWD. You will likely need to do something the rest of the game too.
Even before I became the guy who corrected 4QC stats for people like Elway and Marino, I was tracking successes and failures for active QBs for years. Eventually I started combining the two files to develop records for how successful QBs/teams are at such games. It was only natural for me to start quantifying things like one-minute drills, two-minute offense and the four-minute offense. I want to develop a new win probability model this offseason so I can use things like WPA and Expected Points Added (EPA) for QBs in these situations. I want to quantify late-game performance and strategy as well as anyone ever has, but it’s a process and you’ll just have to bear with me.
I don’t think the W-L record, especially for a QB, is the best way to judge these things, but I know it’s not meaningless either, especially for those who sit at the extreme ends of the chart. There’s something there that’s worth exploring and talking about.
Final-score analysis is heavily flawed to study the closeness of games. Because it takes too long to do this, most close-game studies have always been about the final score. Those can be very misleading. The Colts/49ers from last Sunday played a game that was a tie or one-score difference for 93% of the game before the Colts pulled away 27-7. A final-score study would reject that as a close game, but it would accept trash like MNF Eagles/Redskins from Week 1 when Washington made it 33-27 late and failed to recover the onside kick. That game was not close and the only drive involving a one-score game in the 4Q that night was Michael Vick taking two knees. Forget about the final score.
Rodgers is 20-22 (.476) in games decided by one score, and I hope it’s assumed when I say Rodgers I mean “the Packers with Rodgers at QB”. Because the record with Matt Flynn or Brett Favre (under McCarthy) would be different.
Anyways, 20-22 is a hell of a difference from 9-26 (.257) at GWDs, so you can see it’s two completely different studies. That’s the one thing I would like to change in how I’ve been writing about this. It’s not so much a close-game issue for Green Bay as it is a failure to win games when they have to score the winning points in the 4Q/OT. Behind Rodgers they’re 9-26 at doing that, but 49-5 in all other games. No one has been able to explain that absurd gap in winning percentage, which is the largest in NFL history.
There is no simple explanation as teams lose games for various reasons. Sometimes it’s the QB, sometimes it’s the defense and once in a while it’s a kicker. You can count how many times Mason Crosby missed a clutch kick (four games and three were long attempts) that led to a loss, but what about Tony Romo (5) or Tom Brady (once)? You can’t just adjust Rodgers’ record for these things, because they happen to all other QBs too. If you want the article that will show that, stay tuned to Football Outsiders this season.
No matter who you want to blame, the Packers are 9-26 at GWDs with Rodgers at QB and that is a terrible record, especially for such a good team. Rodgers is the headline, but the Packers’ problems are the real story, and too many people are glossing over that aspect of this.
As for criticism of my “Phil Simms analysis” that 4QC show the cream rising to the top, well you find fault with the 10 guys who have held the record for most 4QC wins since 1950: Sammy Baugh, Sid Luckman, Bob Waterfield, Bobby Layne, Otto Graham, Y.A. Tittle, Johnny Unitas, John Elway, Dan Marino and Peyton Manning. That’s a who’s who of the best QBs through the years with Joe Montana (5th all time) only excluded because he missed too many games in his career. The 1970s are not represented, but wouldn’t you know Terry Bradshaw, Roger Staubach and Ken Stabler all lead the decade with 15 4QC wins. Throughout NFL history, the best QBs dominate this stat as much as any other stat you can pick. But for Aaron Rodgers, he’s still somehow behind John Skelton and Tim Tebow. If that doesn’t make you scratch your head, nothing will.
Enough with the “lack of opportunity” argument – I hammered on this before, but again some people think Rodgers has a lack of 4QC/GWD for a lack of opportunity. 29-35 games is plenty of opportunity. It’s not the opportunity, it’s the bad winning percentage. Here’s an updated list with a few more notable QBs and how many 4QC opportunities they have had by start.
Rodgers is just above average at 32.6%, so stop it.
Statistical significance vs. real significance – I want to tread lightly on this topic as this alone could be 5,000 words out of me. I fully understand the small sample size issues with covering football. I’ve done hundreds of articles and looked at many things over the years, so I know as well as anyone when we don’t have enough data to make good conclusions. How many comeback opportunities does Rodgers need before we can statistically conclude his record is bad? 30? 50? 100? I don’t know, but I will work to find out in the offseason.
In the meantime, I’m going to keep doing my job as a football analyst to present the patterns and trends that aid our coverage of the game. They may or may not have statistical significance, but once you start talking about 29-35 games, that seems rather foolish to brush everything off as being random.
We can all agree the final minutes of a close NFL game are different from the rest of the game, right? The rule book changes in regards to clock stoppages and things like advancing the ball after a fumble. Time actually becomes a factor with using timeouts and managing the clock. No one cares about the game clock unless it’s the end of a half. Offenses will use all four downs while playing three-down football most of the time otherwise. There’s that sense of “if this drive is not successful, we will lose the game” that just does not come early in the game. It’s a different experience in crunch time.
So how many times does a team need to experience this before they learn how to adapt to the situation? Think of your own real-life experiences in adverse situations: driving up an icy hill on your way home from work, flying on an airplane or going to a funeral parlor. Yeah, I’m going to go with the darkest analogy I could think of.
Do you have to go see 80 dead people before it becomes statistically significant in how you will handle the situation? Or does it take a few trips before we know what to expect and act accordingly? That could be anything from the smell of the place, the demeanor of mourners, dealing with the image of the person in the casket, proper dress attire, etc. Sometimes we may get thrown a curve ball like a person laughing hysterically or someone throwing themselves onto the casket. In football, some unexpected things can come up too like a seven-man blitz or a dropped pass.
In other sports we have seen teams like Michael Jordan’s Bulls or Sidney Crosby’s Penguins have to climb the ladder of success before winning a championship. That means getting your feet wet in the playoffs, learning how to adjust for a best-of-7 series and going further each time before eventually completing the journey to the top.
Why can’t it be the same in the NFL where you have to learn to adjust to adverse situations? It shouldn’t take years upon years to do that either. I think we’ve seen enough from the Packers to reasonably conclude they struggle a lot in these types of games.
If you honestly see zero significance and only randomness to the Packers being 5-24 at 4QC behind Rodgers — possibly 0-20 against winning teams — then maybe following the NFL is not right for you. That record is unlike anyone else’s record when we’re talking about an annual SB contending team. Now if you want me to break the records down to adjust for opponent, or dig deeper into the causes, then that’s fine. I’ve done such things in the past. I know the few wins the Packers do have have often been unimpressive (bad opponents, small deficits). There are patterns. I’ve done enough to know something is not right with how the Packers win and lose football games.
Not to harp on it, but the comments made this offseason by Greg Jennings and Donald Driver about Rodgers’ leadership is another layer to this story. Cue the smoke/fire line. We don’t see receivers for QBs like Peyton, Brady and Matt Ryan question their leadership. We also see those QBs with great success in these close games. Maybe there’s something there, but let’s stick to numbers.
I have seen all 26 losses by GB. They happened and it didn’t take a stroke of bad luck every time. This team has issues late whether it’s the QB’s unwillingness to throw interceptions so he takes drive-killing sacks, the lack of a running game, the struggling OL, McCarthy’s playcalling, Dom Capers’ defense or Mr. Crosby’s kicking. There are baselines already established. For an elite QB, a 9-26 record at GWDs is bad and no one will convince me to say otherwise. Should it improve, then credit to the Packers.
But as long as it stays where it is, we have a problem here, and remember it’s a problem that has already and will continue to cost the Packers wins, division titles, higher playoff seeds, playoff wins and Super Bowl rings.
2013 NFL Week 4 Predictions
After hesitantly picking the 49ers, that makes me 4-0 on the Thursday games this season. My record’s much better than the quality of those games. I’m still stinging from another difficult Week 3 that saw an 8-8 record. Onward and upward this week as we try to figure these teams out.
Winners in bold:
- Giants at Chiefs
- Cardinals at Buccaneers
- Steelers at Vikings
- Ravens at Bills
- Bears at Lions
- Bengals at Browns
- Colts at Jaguars
- Seahawks at Texans
- Jets at Titans
- Eagles at Broncos
- Cowboys at Chargers
- Redskins at Raiders
- Patriots at Falcons
- Dolphins at Saints
- Week 1: 11-5
- Week 2: 12-4
- Week 3: 8-8
- Season: 31-17
Good god I have 10/14 road teams winning this week. Even if we don’t count Pittsburgh (neutral site), that sounds like trouble. Upset watch for Seattle, Cincy, Baltimore and Chicago?
Also, back in April I had Pittsburgh beating Minnesota in London with the premonition of Adrian Peterson being contained, Christian Ponder coughing over some turnovers, Big Ben finding Sanders/Brown deep down the sideline for scores. Just a good day for the Steelers in London. Now with both teams at 0-3, I barely feel like watching this one. Though with Matt Cassel stepping in at QB, I can’t imagine the takeaway-less Steelers do not get a few this week. And I still expect the Steelers to win, dropping a Minnesota team I railed on more than any other team this offseason to 0-4.
With Carolina and Green Bay on the bye week, there’s no chance to blow a late lead this week. But if there’s anyone I don’t want to see need a fourth-quarter comeback in Week 4, it will be Breaking Bad. I’ve noticed a lot of big-time series finales in recent years (Dexter and Big Love especially) waited too long to get things going and tried to rush it for a botched ending. I’m counting on big things from AMC here.
If Walter White escapes the country to become a lumberjack, I’m going to lose my sanity and quit watching these series since we never get closure or final satisfaction anymore.