Clutch Encounters: Super Bowl LI

Clutch Encounters: Super Bowl LI
Clutch Encounters: Super Bowl LI
Photo: USA Today Sports Images

by Scott Kacsmar

Exhilarating. Inconceivable. Triumphant. The New England Patriots completed the comeback of all Super Bowl comebacks, erasing a 25-point second-half deficit to win 34-28 in overtime. Three teams had previously won Super Bowls after trailing by 10 points, but this was unlike anything we have ever seen with the championship directly on the line. Bill Belichick and Tom Brady moved past Chuck Noll and Terry Bradshaw with their fifth Super Bowl win together, and Brady won his fourth Super Bowl MVP, setting Super Bowl records for attempts (62), completions (43), and passing yards (466). This was the 50th fourth-quarter comeback win for a Belichick-coached team, and Brady's 10th game-winning drive in postseason play.

Deflating. Inexcusable. Tragic. The Atlanta Falcons seemingly could do no wrong early, but then failed to do anything right after taking a 28-3 lead with 8:31 left in the third quarter. Even if we forget the peak of the lead and the fact that this was the Super Bowl, the 19-point blown lead in the fourth quarter is now the largest blown fourth-quarter lead in NFL postseason history. Head coach Dan Quinn was Seattle's defensive coordinator in Super Bowl XLIX, so he has watched the Patriots pull off the two largest fourth-quarter comebacks in Super Bowl history against his defenses. Matt Ryan had the highest passer rating ever in a Super Bowl loss (144.1). Ryan and Brady have now been on the losing end of the three biggest blown leads in NFL championship game history.

Biggest Comeback Wins in NFL Championship Game History
Rk Winner Loser Season Game Deficit Final
1 Patriots Falcons 2016 SB 25 (28-3) W 34-28 OT
2 Colts Patriots 2006 AFC-CG 18 (21-3) W 38-34
3 49ers Falcons 2012 NFC-CG 17 (17-0) W 28-24
4 Seahawks Packers 2014 NFC-CG 16 (16-0) W 28-22 OT
5 Falcons Vikings 1998 NFC-CG 13 (20-7) W 30-27 OT
6 Colts Jets 2009 AFC-CG 11 (17-6) W 30-17

Just when I thought I would have Sunday night off due to another one-sided rout this postseason, that happened. Depending on your stance as a fan, you can choose which story fits the outcome better. Given that leads of that size so late in the game almost always stand up, especially with the trophy in the building, I cannot help but feel that the Falcons gave this one away more than anything.

That said, New England also did its share to take the game back. After the 28-3 start, the Patriots gained 24 first downs to three for Atlanta the rest of the way. New England's 37 first downs are a new playoff record. We have seen some great rally attempts from huge deficits in recent Super Bowls, including the 2010 Steelers (trailed Green Bay 21-3; got as close as 28-25; ultimately lost 31-25) and 2012 49ers (trailed Baltimore 28-6; got as close as 31-29; ultimately lost 34-31). Usually, the team with the big lead still finds a way to close things out, whether it's a goal-line stop or a takeaway. The Falcons produced none of those things, and even lost the overtime coin toss to boot, never to see the ball again.

Overall, the 2016 NFL season was a bit of a disappointment in terms of the quality of play, with some bad injuries that weakened the playoff field. However, an unforgettable Super Bowl will sure help its legacy. Our final recap of the year looks at one of the more perplexing collapses in NFL history.

Game of the Week

New England Patriots 34 vs. Atlanta Falcons 28

Type: 4QC/GWD
Largest Fourth-Quarter Deficit: 19
Head Coach: Bill Belichick (50-77 at 4QC and 65-78 overall 4QC/GWD record)
Quarterback: Tom Brady (39-35 at 4QC and 51-37 overall 4QC/GWD record)

The Path to 28-3

Incredibly, the Patriots are now 13-0 in the playoffs since 2001 when facing a team for the first time that season, but 12-9 in rematches. Almost as incredible is the fact that the Patriots have gone scoreless in the first quarter of all seven Super Bowls in that time. This was no exception, with Brady getting sacked twice in a scoreless first quarter that featured more defensive pressure than expected on both sides. Atlanta jumped out to a 14-0 lead, the first time the Patriots had trailed since Week 12, which ended the longest such streak the league has seen since the 2005 Colts.

[ad placeholder 3]

It looked like the Patriots would get back into it with the help of three third-down conversions via defensive holding penalties on one possession. However, on the drive's fourth third-down play, Brady was picked off by Robert Alford for an 82-yard touchdown to give Atlanta a stunning 21-0 lead. That lead was trimmed to 21-3 at halftime, but still, if there was any concern for the Falcons at that point, it was that the offense had only run 19 plays to 42 for New England. The Patriots held the ball for 19:35, but at least the long halftime show gave Atlanta's young defense some rest.

After getting the best of cornerback Malcolm Butler a few times, Atlanta's passing game was back in the end zone, with Ryan tossing a 6-yard score to Tevin Coleman to go up 28-3 nearly halfway through the third quarter.

From there, New England scored the game's final 31 points.

The Slow Comeback

Every big comeback has to have an inciting incident to get things started. Down 28-3, the Patriots looked awfully desperate when Julian Edelman threw an incomplete pass to Dion Lewis after some trickery on third-and-3. Some coaches probably would have punted from their own 46, but Belichick absolutely made the right call by going for it here. Brady found Danny Amendola for a 17-yard gain and that was the inciting incident. Brady also scrambled for a 15-yard gain on third-and-8 after the Red Sea parted in the middle of the field, another unexpected play the Falcons failed to stop during this collapse. Still, even after a touchdown pass to James White, Stephen Gostkowski missed an extra point, and then illegally touched his onside kick attempt before it had traveled 10 yards on the ensuing play.

The Patriots now trailed 28-9 with just over 17 minutes left, and the Falcons were in great field position at the New England 41. A low-key big play was a holding penalty on left tackle Jake Matthews that knocked the Falcons out of field goal range and set up a second-and-11. Ryan ended up taking a sack on third down and the Falcons punted. That was a missed opportunity, but Atlanta still took a 19-point lead into the final quarter. When Grady Jarrett sacked Brady twice to force a field goal, it felt like the Patriots were stuck way behind the eight ball with a 28-12 deficit and 9:44 left.

Where the Falcons Blew It

Let's get to the meat of the loss. At this point of the game, any additional score would have likely sealed the deal for Atlanta. Even a drive that consumed four or five minutes of clock before flipping field position likely would have done the trick. Coming into this game, some compared Belichick's potential strategy to the one that he infamously used against the 1990 Bills when he was the Giants' defensive coordinator in Super Bowl XXV. In some ways, the statistics look very similar, right down to the number of drives, points (especially if Scott Norwood had made the game-winning field goal), yards, the miserable 1-for-8 on third-down conversions, and the fact that Belichick's team held the ball for over 40 minutes in both games.

Bill Belichick: Super Bowl Defense Déjà Vu?
Game DEF OFF Drives Points Yards 1st Downs 3rd Down Rush YPC Sacks Takeaways TOP
SB XXV NYG BUF 10 17 371 18 1-for-8 6.64 1 0 19:27
SB LI NE ATL 10 21 344 17 1-for-8 5.78 5 1 23:37

However, there is a huge difference when it comes to the sacks produced, including one massive turnover that changed this game entirely. After Coleman left with an injury, Atlanta faced a third-and-1 from its own 36 with 8:31 left. The first problem was flexing Devonta Freeman into the backfield and showing a pass look, because that took away from the play's unpredictability. I personally would have just run the ball with Freeman, and would have considered doing it again on fourth down if he had been stopped short, but offensive coordinator Kyle Shanahan had other ideas. The second problem was that the snap came with 15 seconds left on the play clock. This thing needed to be milked as much as possible given the situation.

Then there was the play itself. Freeman barely even chipped Dont'a Hightower, who had a clear path to Ryan for a strip-sack that the Patriots recovered at the Atlanta 25. Sure, had Ryan been looking for Freeman as a quick outlet pass, the play would have been great, but that was not the result. Ryan was looking to his left and had no idea Hightower was coming so cleanly. Given that the dink-and-dunk Patriots had moved at an Andy Reid pace down the field on the previous two drives, this was a huge turning point to cough up the ball at the 25 instead of forcing the Patriots to drive 75 yards or more. The play call was nonsensical in a power-running situation with Atlanta's best power runner still available and having a good night.

Naturally, the Patriots turned the short field into a touchdown, and pulled out the old direct-snap play to James White for a key two-point conversion. Suddenly, we had a one-score game at 28-20 with 5:56 left, and the comeback started to feel inevitable at this point.

When protecting a 16-point lead in the fourth quarter, you expect to get that last defensive stop, or to prevent one of those two-point conversions, or to add a field goal that ices the game. The Falcons did none of these things, despite putting themselves in a great position to do get more points. An uncovered Freeman immediately gained 39 yards on a pass. Julio Jones made a fantastic 27-yard grab near the sideline with 4:40 left. Jones had a season-low four targets, but caught them all in impressive fashion for 87 yards. The fact that Ryan only threw 22 non-spiked passes had a lot to do with Jones' low output, but it was still puzzling to see Jones with four targets while rookie tight end Austin Hooper led the team with six.

Still, the Jones catch put the ball at the New England 22, which should have all but iced this one. Using the Game Play Finder at Pro Football Reference, I looked at offenses with a 6- to 8-point lead in the final five minutes of the fourth quarter and a first down inside the opponent 30. That is the margin and field position where a team should be expected to make it a two-score game with at least a field goal. Any game where the first-down play was a kneeldown was excluded. Since 2000, these teams were 221-3 (.987), so Atlanta is in the rarest company with this loss. I did not have time to check 224 games to see how often the team extended to a two-score lead versus how often the defense just held up, but the previous three losses show Atlanta's offensive failure to be the worst of them all.

  • 2000 Eagles at Steelers: Up 20-13 with 4:07 left, Pittsburgh ran Jerome Bettis three times, which led to a 40-yard field goal and 23-13 lead. Fair enough. The Eagles got a quick touchdown and recovered an onside kick to get the game-tying field goal with no time left before winning 26-23 in overtime.
  • 2005 Texans at Rams: Up 24-17 with 3:19 left, the Texans failed to get another first down, but Kris Brown kicked a 35-yard field goal for a 27-17 lead with 2:49 left. (Oddly enough, Brown had been Pittsburgh's kicker in the 2000 Eagles game.) This was Ryan Fitzpatrick's NFL debut for the Rams, and he led one of the craziest comebacks ever. After a touchdown with 26 seconds left, the Rams recovered an onside kick, kicked a 47-yard field goal, and then won in overtime. Again, an onside kick was needed to keep hope alive.
  • 2009 Saints at Redskins: Here is a very memorable game for Saints fans in their Super Bowl season. The Redskins actually had three first downs inside the 30 in the final five minutes, and a 23-yard field goal by Shaun Suisham should have put Drew Brees and the Saints away in a 30-23 game. Shockingly, Suisham was wide right on the kick with 1:52 left, keeping hope alive. Brees led the game-tying touchdown drive and the Saints eventually won 33-30 in overtime.

The fact that Atlanta did not even get off a field goal attempt is just mind-numbing. This was easy field goal range for Matt Bryant, one of the game's best kickers. When a field goal is almost as good as a touchdown in this situation, why wouldn't Atlanta simply run the ball three times and kick a field goal? Hey, Atlanta might have even earned a first down via the run with the way things were going on the night. With New England having four clock stoppages left, completely running out the clock was very unlikely. The Patriots were going to get the ball back, but if they only had a little over three minutes left in a 31-20 game, then who cares? That makes it very hard for them to win without an onside kick recovery to extend the game like those other teams did in 2000 and 2005.

You would think with the way that Super Bowl XLIX ended at the 1-yard line that Quinn would understand the importance of running the ball. You would like to think that a defensive-minded coach would make sure his offensive coordinator understood the situation here, but this is where Shanahan really botched things. After Freeman lost a yard on a first-down carry, Shanahan took a risk by calling a pass play on second-and-11. Trey Flowers destroyed center Alex Mack, who was playing on a fractured fibula, and Ryan lost 12 yards on a sack. The Falcons then completed a short pass that would have given Bryant a shot at a field goal, but it was wiped out by another holding penalty on Matthews, forcing a third-and-33 situation. Shanahan should have called something simple like a screen to try getting 10 yards or so, but Ryan's intermediate pass to the sideline missed Taylor Gabriel, stopping the clock in the process. The Falcons ended up punting the ball back after completely botching the situation.

Quinn and Shanahan failed Game Management 101 in the fourth quarter. Maybe Ryan can take some blame for this too, but all three should have known that the clock was an ally. Yet on this drive, the Falcons snapped the ball with 19 seconds left on the play clock after the Freeman catch, and with 13 seconds left before and after the Jones catch. That's roughly 45 seconds saved for the Patriots by Atlanta. Add the 15 seconds from the third-and-1 fumble disaster, and that is a full minute of game time. Even though Brady had to drive 91 yards, 3:30 with two timeouts was an eternity of time.

In our look at sneaky Super Bowl stats, one of the topics was generating pressure. Ryan's pressure rate this season was the highest of his career at over 30 percent. He is usually hard to sack, but went down five times on just 27 actual dropbacks in this game. That is a terrible sack rate of 18.5 percent. The sacks were not always blitzes, but Ryan took four of his sacks on third down.

The sacks in situations where the Falcons should have just been running the ball cost Atlanta the Super Bowl.

The Tying Drive

No player in NFL history has had more doors left open for him than Brady, and he usually makes teams pay for that. One big third-and-10 conversion to Chris Hogan got the drive moving, and then once again we saw a crazy late catch in a Super Bowl involving the Patriots. This time the catch went for New England, and that was only after Alford dropped an interception that could have been the dagger. Instead, the tipped ball bounced off an Atlanta's player leg, then Edelman was just able to grip the ball before it hit the ground. Incredible, and a 23-yard gain.

Brady was not as sharp to Edelman as usual (5-for-13), but that unexpected catch was a good boost. It was also instrumental in leading to a piss-poor challenge by Quinn, who did not get the clock to stop until 2:03 remained. Had he let things go to the two-minute warning, he would have seen that the catch was good and that there was no need to blow his final timeout. Instead, the Falcons were out of timeouts and the Patriots had time to get another 20-yard gain to Amendola with 1:57 left.

White took two short passes down to the 1-yard line. I singled out the receiving backs in this game for Film Room, because I thought Dion Lewis could challenge the Super Bowl record for catches. Well, that record was broken, but it was by White, who caught 14 of his 16 targets for 110 yards while also scoring 20 points in the game. When the Patriots had the ball at the 1-yard line in the final minute of the Super Bowl, they did the smart thing and ran the ball. White scored, setting up what FOX's Joe Buck really missed out on calling the biggest two-point conversion in NFL history. This was basically the game, and the Patriots went with a screen to Amendola, who muscled his way past the plane with the ball for the conversion. Dwight Freeney was offsides, so it would have been a do-over even if Amendola had been stopped short.

The Game-Winning Drive Attempts

So after an epic collapse by Atlanta, the game was tied at 28. We were finally going to get overtime in a Super Bowl, unless Ryan could extend his NFL record with a sixth one-minute drill to win a game. However, Eric Weems made a terrible decision to take the kickoff out of the end zone, eventually getting tackled at the 11-yard line. With no timeouts remaining, this made things very difficult for Atlanta to do something here. Ryan had to be a bit aggressive for fear of never seeing the ball again, but he also couldn't throw a pick to lose the game.

The drive had little rhyme or reason, with a short gain to Hooper followed by a spike. Ryan's deep shot for Hooper was nearly intercepted, leading to an Atlanta punt. The offense never touched the ball again. The Patriots tried a little razzle-dazzle from their own 35, but that just led to a hamstring injury for Lewis.

[ad placeholder 4]

In overtime, the Patriots won the coin toss, and you better believe this was a game to take the ball first. The ending almost felt inevitable from that point. Brady threw some of his best passes of the night in overtime, and the Falcons showed very little resistance. The pressure that had been so evident early was nonexistent down the stretch after the Patriots finished up with 93 snaps, the second most in playoff history. Martellus Bennett almost had a game-winning touchdown catch, but he was interfered with by De'Vondre Campbell on a good call. With the ball at the 2-yard line, only some silly pass really could have saved Atlanta, and Brady almost obliged on first down. Vic Beasley got one hand on the ball, but never had any real shot for a pick. The Patriots wised up on second down, and White took a toss and cracked the end zone for the game-winning touchdown.

Comeback completed. Collapse completed. The No. 1 offense never had a chance to answer, because that is the overtime system the league has always stuck with. I was always hoping that a competitive, back-and-forth Super Bowl would go to overtime where this happens to promote changing the system so that both offenses touch the ball, but I think the overall comeback/collapse theme here was too strong to get people fired up enough about that. If the Falcons got the ball back, did anyone expect them to score after such a gut punch? Still, it sure would have been nice to see, and maybe that is something that can be brought up in the rule changes this offseason. The object of this game has never been about who can score first, but who can score the most. Only letting one side of the ball dictate that after a coin flip is not fair. If they want to impose rules that only apply to the postseason, then so be it. After all, that's how modified overtime started before it was applied to the regular season.

Gassed and Gasps: Atlanta's Dream Season Ends

How can a franchise and its fans ever get over a loss this bad? The Falcons were trying to become a very unique Super Bowl winner. They would have scored more points (540) and allowed more points (406) than any Super Bowl winner ever has. They would have had the lowest DVOA on one side of the ball (No. 27 defense) for any Super Bowl winner since 1988. Ryan would have been the first MVP winner since Kurt Warner in 1999 to win a Super Bowl in the same season, and yet it was almost predictable that this would not work out. These highly imbalanced teams tend to lose in the playoffs, and in the end, New England's defense made the big sacks and the offense dominated the second half.

Still, how do you lose a game when a quarterback averages 12.35 yards per pass attempt? Well, the Patriots also beat Russell Wilson when he averaged 11.76 yards per attempt in Super Bowl XLIX. In fact, those are the two highest YPA averages in playoff history in a loss (minimum 15 attempts).

Highest Passing Yards per Attempt in Playoff Loss (Min. 15 Attempts)
Rk Player Year Team Opp. Round Result Att. Cmp Pct. Yds TD INT PR YPA
1 Matt Ryan 2016 ATL NE SB L 34-28 OT 23 17 73.9 284 2 0 144.1 12.35
2 Russell Wilson 2014 SEA NE SB L 28-24 21 12 57.1 247 2 1 110.6 11.76
3 Jeff Hostetler 1993 LARD at BUF AFC-DIV L 29-23 20 14 70.0 230 1 0 125.0 11.50
4 Harry Newman 1933 NYG at CHI NFL-CG L 23-21 19 13 68.4 209 2 1 118.1 11.00
5 Colin Kaepernick 2012 SF BAL SB L 34-31 28 16 57.1 302 1 1 91.7 10.79

Moral of the story: maybe run the ball next time instead of looking for a big pass play? Of course, the misleading part here for Ryan are those five brutal sacks for 44 yards and a lost fumble that will haunt him forever should he never make it back to a Super Bowl. As this game has proven many times, you never know when this is going to be your only chance, so you have to make the most of it. Atlanta made the most of this one for about 37 minutes, but failed to get the job done.

It is a good thing that Dan Quinn is bald, because he should be pulling out his hair wondering why his offensive coordinators just won't run the ball in situations that scream for it. He had a chance to speak up this time though.

Never Count Out the Patriots

Do you know how many times in NFL history before Super Bowl LI that a team was down by 16 points in the fourth quarter, scored two touchdowns and two two-point conversions to tie, and won the game in overtime? The answer is one, back when Tim Rattay led the 49ers to a 31-28 win against the Cardinals in 2004. Yeah, that's about how rare it was for the Patriots to win the way they did here.

Super Bowl LI was not the best night for the various NFL win probability models. Many had the Falcons in the high 90s for much of the second half, including 99.8 percent at one point in the third quarter according to Brian Burke.

Given that we had a 28-point comeback by the Colts in the 2013 playoffs against Kansas City, and some other wild games in recent years, it is fair to say that these models have some work to do in keeping up with the current game. Specifically, since 2011 there has been a rise in 17-point comebacks, as well as one-minute drills where teams have at most 60 seconds to score. This is a very offense-oriented era, and it makes for improbable comebacks to be ever so slightly less improbable. Why 2011? That was the year of the lockout and new CBA. Passing numbers have just continued to go up each year since, even if the media is often questioning the talent level of quarterbacks around the league. And the new touchback rule that starts many drives at the 25 also makes comebacks a little bit more likely.

Even with Brady and the Patriots, this change can be seen in how big comeback attempts have gone for them. I knew Atlanta could never feel comfortable until it was really over, because the Patriots are arguably the "hardest kill" in the league right now. It wasn't always like that. From 2001 to 2010, the Patriots were 1-22 (.043) when trailing by at least 17 points at any time in the game. The lone win was a 21-point comeback in Chicago in 2002. Since 2011, the Patriots are 4-4 in such games, and they also gave the 2012 49ers and 2015 Eagles fights at home before finally falling by seven points.

Patriots: Harder to Kill (Games Trailing by 17+ Points at Any Point)
Seasons Games Record PF PA 4QC/GWD Record Win or Close
2001-2010 23 1-22 (.043) 14.5 30.0 1-3 (.250) 4-of-23 (17.4%)
2011-2016 8 4-4 (.500) 29.5 30.9 3-2 (.600) 6-of-8 (75.0%)

Maybe if the Patriots are involved, the win probability model should cater to them. Believe it or not, but this Super Bowl had the largest deficit of any of New England's seven trips under Brady and Belichick, yet the final six-point margin of victory was the largest too.

Whether you love or hate the Patriots, they do always bring the drama and create instant memories with their Super Bowls this century. It took 51 Super Bowls for a monster comeback and an overtime game, and those things alone will always make this one of the most memorable Super Bowls in our lifetimes.

Season Summary

Fourth-quarter comeback wins: 73
Game-winning drives: 86
Games with 4QC/GWD opportunity: 156/267 (58.4 percent)
10+ point comeback wins (any point in the game): 38 (and one tie)

Historic data on fourth-quarter comebacks and game-winning drives can be found at Pro Football Reference. Screen caps come from NFL Game Pass.


218 comments, Last at 14 Feb 2017, 2:53pm

1 Re: Clutch Encounters: Super Bowl LI

"We have seen some great rally attempts from huge deficits in recent Super Bowls, including the 2010 Steelers (who fell behind 21-3, then pulled within a field goal before losing to Green Bay 28-25) and 2012 49ers (who trailed 28-6 after the second-half kickoff, but had a chance to tie the game on a two-point conversion in the fourth quarter before ultimately losing to Baltimore 31-29)."

The GB-Pitt game ended 31-25, and the BAL-SF game ended 34-31. Plus, not only did the 49ers have the chance to tie the game on a 2-point conversion, they had a chance to WIN the game in the last two minutes when they had 1st and goal from the 7-yard line. According to PFR, the 49ers got all the way up to 56.1% win probability.

3 Re: Clutch Encounters: Super Bowl LI

If I recall correctly, the four comeback wins for NE are back-to-back rallies from 17-0 and 21-0 against Miami and Buffalo, respectively, in 2011, the 24-0 comeback against Denver in 2013 and last night.

If so, then this is not correct:

yet the final six-point margin of victory was the largest too.

I'm pretty sure they ended up killing the Bills 49-21 in that game.

5 Re: Clutch Encounters: Super Bowl LI

"Your WP model is wrong because it didn't give >50% WP to this event which, when viewed in real time, was considered a surprise" and "This event happened, and your WP model gave it a non-zero % chance. Your WP model is therefore wrong" are my least favorite arguments and have been since election night.

Note - I may be rewording those arguments, that's basically what they say.

Note - I'm not really talking about FO writers or commentator.

13 Re: Clutch Encounters: Super Bowl LI

I think the criticisms are a bit more sophisticated than that. The biggest reason is that the number of very large comebacks has increased dramatically in recent years.

But as for the game yesterday, the question is whether a 0.2% WP is really appropriate when the Pats were down 25. To simplify things what they needed was the following sequence: TD, stop, FG, stop, TD, stop, TD, + OT (~coin flip) and at least 4 extra points somehow. If we assign 0.5 for each probability (just to be simple), we're looking at roughly 0.5^8, which is about 0.4%. So, yes, that was a reasonable WP at the time.

But...ESPN is saying the probability was still 0.4% after the Pats cut the lead to 16. That seems a bit low. I'd probably say it was about 1% by then. It was starting to become clear that the Pats offense could move the ball efficiently.

By the time they got the ball back trailing only 8, I thought they had at least a 66% chance of tying the game to force OT, and thus a 33% chance of winning. And when they won the coin toss, I thought their WP was well over 90%.

But these kinds of estimates are based on seat-of-the-pants thinking. The problem is no model is really able to capture all of the exact values of variables at any point in time, so the best we can do is make estimates for what these values might be, based on historical estimates. In pure games of chance like poker and roulette, these historical estimates are 100% accurate (to be a bit more honest, probabilities in real games of chance are generated mathematically and then care is taken to ensure that the real world duplicates these probabilities as nearly as possible). In something like football, that kind of accuracy is impossible.

I do agree with your basic point, that the public at large just doesn't have a feel for what probability means. The most recent example of this is the November election, when the media treated a 70% likelihood as a 99.9% likelihood.

Sorry, I've rambled a bit so I'll leave it at that.

29 Re: Clutch Encounters: Super Bowl LI

"Well, all right, last desperate million-to-one chances always work, right, no problem, but... well, it's pretty wossname, specific. I mean, isn't it?"

"You tell me," said Nobby.

"What if it's just a thousand-to-one chance?" said Colon agonizedly.


"Anyone ever heard of a thousand-to-one shot coming up?"

Carrot looked up. "Don't be daft, Sergeant," he said. "No one ever saw a thousand-to-one chance come up. The odds against it are-" his lips moved - "millions to one."

7 Re: Clutch Encounters: Super Bowl LI

Is my impression of overtime correct: when they realized the 1958 Championship was going to be tied, they improvised sudden death, and when it came time to establish the OT rules, everyone was like, that was cool, let's keep doing it that way. I probably entirely mangled the origin story, but I'm not sure.

27 Re: Clutch Encounters: Super Bowl LI

I took a look at coverage of the 1947 Steelers-Eagles Eastern Division playoff, and the newspapers made a point of mentioning before the game that if it was tied after four quarters, there would be sudden-death overtime. (It wasn't an issue. Philadelphia won 21-0.)

64 Re: Clutch Encounters: Super Bowl LI

I'd been able to block out everything about that game except the missed kick, but I seem to recall that play now. Thanks joe ;p

Man, that was a terrible season. The Jim Zorn era began so promisingly (6-2 in the first half of 2008--with Jason Campbell MVP talk!)...and everything after that was so, so bad.

9 Re: Clutch Encounters: Super Bowl LI

I see nothing wrong with the current OT rules. I could understand the argument when a team could receive the kickoff, throw two passes and win the game with a 45 yard field goal. But playing defense is part of the game, and every bit as important as offense. If you kick off to start overtime, and allow the opponent to drive the length of the field and score 6, you deserve the loss, regardless of whether or not your offense got on the field.

48 Re: Clutch Encounters: Super Bowl LI

According to this article on The Ringer:

The team that receives first wins 54.8% of the time. By way of comparison, the college rule gives the second team an advantage that plays out at nearly an identical rate.

212 Re: Clutch Encounters: Super Bowl LI

Technically, does a safety win the game outright? Or is the rule wording "the first team to lead after each team has had possession or a TD has been scored"?

If a team gets a safety, does the game immediately end, or does the free kick occur? If it does, I see that as soon as the receiving team takes possession, the game would end, but it IS possible to onside a free kick after a safety, or just to recover the free kick yourself. Could a safety-committing team recover its own free kick and get a chance to drive for a FG or a TD to win outright?

35 Re: Clutch Encounters: Super Bowl LI

The current overtime rules are hugely better than they were before. But they are still non-sensical.

No one would suggest using any kind of sudden death in a baseball game even though playing defense is part of that game too. You could make a rule that if the first team up at bat scores only one run, then the second team gets a chance to bat, but if the first team scores two runs then the second team doesn't get a chance. That might make sudden death slightly more fair, but it still doesn't make much sense.

The game isn't just about one offense playing against one defense.

112 Re: Clutch Encounters: Super Bowl LI

Sorry, both teams get 60 minutes to beat the other team. If you don't manage to do that in regulation, you simply don't deserve more than a coin flip. So overtime rules simply don't have to be perfectly fair. 55:45 is good enough.

I liked the old OT format better.

119 Re: Clutch Encounters: Super Bowl LI

Well, if the goal is to end the game without giving anybody more than they deserve, just have a coin toss to determine the winner, when they are tied at triple zeros. The goal, of course, is to end the game without playing a full fith quarter, in an entertaining fashion. Next to a coin toss, the next most unentertaining way to end the game, is, it seems to me, the sequnce of coin toss, touchback, 35 yard drive, field goal, so I like the present system better, and I'm ok with it. I might like to see both teams' offenses and defenses perform, however.

129 Re: Clutch Encounters: Super Bowl LI

As you might or might not know, the outcome of the old format was about the same as it is now - less than 55% win pct for the winner of the coin toss. That's fair enough. It's not like they didn't have enough opportunities to solve this terrible, terrible problem beforehand.

And in case you did not notice: the new system did not make the OT any more fair even though "it looks like it should". And the same holds true for all other proposed mechanisms. Guarantee the other team the possession, and then it favors the second team. make it more complicated or make it more artificial, but you will never get a 50-50 chance. It it doesn't have to be that way. It's ignoring effects that come with the changes you are proposing.

BTW I am also in for getting rid of instant replay completely. It may help to reduce some officiating errors (the Patriots challenged exactly one play this season, so instant replay doesn't seem to be that important), but it also introduces other effects that affect the game in the exact opposite direction (hell, it might force the ref to artificially call a close situation a score just because IR might "help him out" and scoring plays are subject to review). Can't measure that, but thinking that a change doesn't introduce something you did not think of in your perfect wet dream theory is just ignorance. It might sound great in theory, but it doesn't work in practice.

132 Re: Clutch Encounters: Super Bowl LI

I did know that. I still think the sequence of touchback, 35 yard drive, field goal is the most boring way to end a tie game after regulation. That doesn't happen any longer, and that's good. You are the one bringing the concept of "fair" into this. I haven't thought that was a useful way by which to interpret human affairs, in any realm, since I was about 12 years old.

I have absolutely no idea what ideas you are trying to transmit with the last paragraph, other than you don't want instant replay. OK, you don't want instant replay.

152 Re: Clutch Encounters: Super Bowl LI

At the risk of taking this down a whole new road, your point about 1 challenge by the Patriots is interesting. From 2000 to 2012, Belichick averaged 7.2 challenges per year with at least 5 challenges every year since 2001. Then in 2013 he suddenly reduced his challenges. He went 1, 3, 2, 1 (for an average of 1.75). He only won 2 of those 7 challenges, which was much lower than his previous career average.

I have to guess that he made a conscious decision that he would only challenge plays that had a significant shift in win probability.

155 Re: Clutch Encounters: Super Bowl LI

Since 2013 hasn't pretty much everything that's worth challenging been reviewed by the booth? All scoring plays, turnovers and everything in the 2-min warning.

About all that's left is the occasional out of bounds, spot of the ball or Julian Edelman catch.

157 Re: Clutch Encounters: Super Bowl LI

According to

League average challenges:

2016 - 4.0 per game
2015 - 6.3
2014 - 5.2
2013 - 5.8

I can't find anything before 2013. So not sure if there was a decrease in 2013, but Belichick has been way below league average since 2013.

159 Re: Clutch Encounters: Super Bowl LI

Yeah, pretty much.

It's generally not worth challenging turnovers that weren't ruled as such on the field, because the whistle was usually blown to early, or it's tough to see.

Combine that with the fact that the refs usually err on the side of a reviewable play for the endzone/turnovers, there's not a ton left.

You're very often risking a Timeout for a couple yards of field position - some games thats worth it, but most often its not.

162 Re: Clutch Encounters: Super Bowl LI

In my eyes, Bill Belichick's whole approach has changed the past years quite significantly.

He appears much less of a control freak to me now*, and seems to give both his coaches and his players quite a bit more leeway. Might be influenced by the train wreck of team they had in the 2009 and 2010 seasons, but the way he handled and has talked about the past two SB squads speaks volumes. Players seem to be a bit more lose, a bit less afraid of making mistakes.

And I guess Belichick only challenges two kinds of plays
a) very important stuff
b) when a player made a great play and got robbed by the refs

The second is much more of a trust and team building emphasis than it is to gain an additonal inch because the refs (who are human too and are allowed to make mistakes as long as their calls are consistent) made a mistake.

*=he still controls most of the important stuff, but years ago he seemed to micromanage absolutely everything. I think he delegates much more stuff now and is way more comfortable doing that than a couple of seasons ago.

PS: regarding the challenges: There simply weren't any obvious blown calls against the Patriots this season

175 Re: Clutch Encounters: Super Bowl LI

There's nothing intrinsically wrong with a (regular season) tie; we've all seen games where neither team deserved to lose (or win). The problem is teams playing for the tie instead of trying to win in the last few minutes, which makes what should be exciting football boring.

52 Re: Clutch Encounters: Super Bowl LI

Given that it is impossible for a defense to score in baseball, your analogy isn't quite on target. It also should be noted that any system that ensures a response possession grants a strategic advantage to the latter team in a way unlike being the home team in baseball. For instance, the college set up sees the second team win at a rate similar to the first in the NFL.

About the only way around this is to add a full period, but this has issues as well. Anything shorter than 10 minutes can be run off by a good drive and wouldn't fix the "first" problem. A full quarter seems excessive, especially after the 60 minutes beforehand.

I really don't think there is a way to completely fix the problem. As is, the advantage of the first team has been reduced to roughly a 55/45 split, which might just be the best we can do.

163 Re: Clutch Encounters: Super Bowl LI

Given the extremely low incidence of defensive scoring, the baseball analogy isn't that far off.

I prefer the college system because both offenses and both defenses are involved. But I agree it still doesn't have both teams facing the exact same situation which is the crux of the matter for me.

That could potentially be resolved in the college system by eliminating field goals from OT thereby eliminating any strategic decisions. Give each offense the ball at the 50 and see who can score a touchdown, no punts, no field goals. Defense scores, it's game over. I'm not sure that's perfect either, but it would make a lot more sense to me. And I think it'd be pretty cool.

177 Re: Clutch Encounters: Super Bowl LI

This intrigues me, but it still gives the second team a strategic edge by knowing how much they need to score. Maybe if you offset that with it becoming sudden death after each team has had two possessions?

To clarify, if team B stops team A's first drive and then scores on theirs, that would satisfy the criteria, correct?

195 Re: Clutch Encounters: Super Bowl LI

No, they would still have to regain possession while having the lead. Team B's score would only give them the lead, so they'd still have to make a defensive stop and regain the ball. Or perhaps try an onside kick, if they have zero confidence in their defense, I suppose.

It seems to me that this would lead to lots of interesting decisions in overtime games. However, it would also lead to the occasional extremely long game.

50 Re: Clutch Encounters: Super Bowl LI

Baseball is a completely different game than football, with no clock, for example, so I don't see why one would draw an analogy between the two. An 'innings' is a discrete portion of the game, it's easy to add one or more, and each lasts an average of about 20 minutes, even in today's game, so it's not a lot more to add.

But the more important distinction is the wear-and-tear of the game. You can play a 20-inning baseball game, though you might run out of pitchers. You can't play 9 quarters of football, not if you want player to remain upright.

62 Re: Clutch Encounters: Super Bowl LI

My personal preference, for overtime in the playoffs, is to play full quarters with the same rules as the 4th quarter with each team getting 3 TO's until a winner occurs at the end of a quarter. I would also say that we don't start with a kickoff but play continues as it does going from 1st to 2nd quarter (or 3rd to 4th). This seems to me to best eliminate the problems with current OT. But I do see the downside, and personally would not want to play 8 quarters of football in a row, and that others would weigh the merits differently.

67 Re: Clutch Encounters: Super Bowl LI

That still gives a significant advantage to the team that gets the ball first - they're most likely going to get to run more drives than the other team, and are going to be able to put the other team under time pressure if their offense is good.

What you really need is to add TWO quarters at a time, that are structured like halves (so each team kicks off at the beginning of one) - but thats a little bit ridiculous.

You can do a lot - but I don't think you're going to get much better than the 55/45 we're at now. Overtime isn't supposed to perfectly determine a winner - its supposed to quickly and reasonably determine a winner in a game where both teams seem to be closely matched.

71 Re: Clutch Encounters: Super Bowl LI

His suggestion was that the team that gets the ball first in overtime is the team who has the ball last in the fourth quarter. This allows teams to strategize possession at the end of the fourth quarter.

For instance, on Atlanta's final possession, they could have operated differently knowing they'd retain possession at the end of regulation. So their goal is just to gain as much yardage as possible to maximize their field position.

I've heard the suggestion before, and I'm a fan. I think it could even return the game to straight sudden death, since the luck of a coin flip is eliminated.

85 Re: Clutch Encounters: Super Bowl LI

Ha! Sorry! That's my tag. It's on all of my comments.

And I'm being facetious on my previous comment - that's what they do to break a tie in soccer. A practice that soccer players hate.

- Anything is possible when you have no idea what you are talking about.

91 Re: Clutch Encounters: Super Bowl LI

Another possibility is not adding a full quarter, but playing, say, 6-8 more minutes. The team that gets the ball first has an advantage in that they could conceivably eat up the full 8 minutes and win it with a score, but if they can't, then the second team has the same opportunity in addition to the advantage of information (what do I have to beat?). In other words having the ball first or both both have certain advantages to them.

Then if you're tied after 8, you add 6, then 4, then 2, 2, 2 until the tie is broken. And of course, teams alternate starting the additional periods with the ball.

94 Re: Clutch Encounters: Super Bowl LI

Yeah, the only advantage of that system is that the second team still has an equal chance in the event of a quick touchdown. I suspect it would swing win% wildly toward the initial team, though, given that you could easily eat up 6-8 minutes with a properly managed FG drive.

158 Re: Clutch Encounters: Super Bowl LI

Your pedantry caused you to miss the point.

In 2016, 26 of 32 teams averaged 2:30 or more per drive. That isn't scoring drives, mind you, it's the average of all drives. I can't find the average scoring drive, but to be conservative I think we could tack a minute on to that. This means, in your hypothetical 6 minute period, an average scoring drive by an average offensive team would eat up 60% of the time remaining. If you factor in a better than average offense against a defense that has already played a full game, the bump to 4:00-4:30 means 70-75% of the period could be eroded with relatively normal performance.

In either case, the first team is afforded the full compliment of offensive plays while a high proportion of second teams will have limited time with which to mount their response. Ergo, you haven't solved the issue of first team advantage your suggestion attempts to do.

None of this is an issue if the first team doesn't score, of course, but that isn't a problem for the current system as well.

EDIT TO ADD: I'll grant you that the second team has an offsetting schematic advantage by virtue of knowing how many points they need. I suppose it could be possible in theory to construct a perfect blend of factors that funnels down a large scale 50/50 (I'm not convinced of this, but I'll play along), but it will still be subject to imbalances within particular games. For instance, does knowing you need a TD make it "fair" that you only have 2 minutes to work with after the opponent ate up 6:00 of the 8:00 period?

164 Re: Clutch Encounters: Super Bowl LI

The balances sure aren't perfect, but they're complex enough to create a solid appearance. Suppose the first team takes five minutes to score a FG. Now the second team only has three minutes, but they know a TD would win it. Less time, better information. And if the period ends in a tie, then the second team would have the advantage of starting the six minute period. Double jeopardy.

Mostly I like it because the end of the 4th quarter, with all the strategizing, is much more exciting than OT as it currently stands. This would give us another "2 minute warning", with all it entails in a close game. Maybe more than one.

197 Re: Clutch Encounters: Super Bowl LI

Scenario: Eags at Cowboiys
Eags 17, Cwboys 14
0:35 left in regulation
Cowboys have ball and one timeout.
1st and 10 at Eags 35.

under current rules- Cowboys probably just tyring to get into better field goal range hopinh to get game t o overtime. IF score touchdown, great, but wodul probably be happy to settle for successful field goal. Will hope to win OT coin tosss, score Touchdown and win or stop Eags if they win coin toss and then kick a field goald or td to win.

under my proposal-----Cowboys are nto AS happy to just settle for field goal to tie score 17-17 beucuae whether clock reads 0:00 or 0:01 or 0:05 or so, Cowboys have to kcik off to Eags. No overtime coin toss. once regulation clock hits 0:00, there is commercial break and play picks up right where thigns left off at end of regulation and is true sudden death

198 Re: Clutch Encounters: Super Bowl LI

not sure about seeing conservatism under my plan.

consider what we see a lot nwo-- example- gamne is tied 20-20 and team has ball at own 35 with 18 second sand plays conservative and game just goes to OT. it is a waste of 18 seconds usually. we saw this in super bwol on play d. lewis got injured.

under my plan we see regular football plays instea dof werid wish upon a star plays.

200 Re: Clutch Encounters: Super Bowl LI

I'm not sure why you can't see this, Joe, it seems patently obvious to me. Under the current system, a team that gets the ball with 1:10 left in a tie game will set off on an exciting attempt to move the ball into scoring territory. Under your proposal, they would run their normal offense, content in the knowledge that they will retain possession at the same spot after the intermission.

Or what about having the ball near midfield with 0:18 left in a tie game? Why would you ever attempt a long FG in that situation? If you miss it, the other team opens OT with great field position. Instead, it would be more beneficial to run the clock down so you have your full offensive arsenal at your disposal in overtime.

201 Re: Clutch Encounters: Super Bowl LI

there r relaly issues with whatever system is in place. am just tired of reading and heariong complaints about who won coint oss and "no fairsies our team never got ball in overtime" and also get annoyed a little when a tema sits on ball for final thirty seconds to set up for ovetime (waste of time literally and figuratively)).

do wonder if my plan wodul distort end of regulation situaitons too much. do realize my idea might be too crpapy.

only way we would not see any complaining regarding fairness is if go to coleleg system. instead of starting at opponent 25 maybe start at 50 or own 25 so we don't get too many 3+ OT sessions.

202 Re: Clutch Encounters: Super Bowl LI

It won't happen (sadly) but I'd love to have bidding for possession and starting field position. No luck involved -- who gets the ball and where they start will be completely up to the coaches' estimate of the capabilities of their team.

I think the simplest would be to award possession to the team willing to start closest to their own goalline.

If not that, then flip a coin to designate the coach who will place the ball. The other coach will then decide based on that placement whether he wants to go on offense or defense. So Coach A could say something like "the ball starts on the offense's 15". Coach B could then say "we'll be offense."

213 Re: Clutch Encounters: Super Bowl LI

It always seemed to me that the problem with sudden death wasn't sudden death, but the fact that the first possessor was determined by something arbitrary like a coin flip.

My preferred method has always been go back to sudden death, but have a predefined rule of who gets the ball, and hence the advantage, in overtime. Most coaches these days who win the opening toss defer to get the ball first after halftime. So award the ball first in overtime, should it occur, to the team that got it in the 1st quarter. A coach would have to balance the advantage of deferring against the disadvantage of not having first OT possession, should it occur. At the end of regulation, one team is favored should the game go to OT, but both teams know who, which would control their respective strategies.

Or you could just have home team gets the ball first in OT, and formalize home team advantage more, like in baseball or hockey.

215 Re: Clutch Encounters: Super Bowl LI

I don't think you can have "deferring is an advantage" as an outright rule. Just because BB usually does it, there is a suggestion it's advantageous but that perhaps is just because of how he chooses to coach football. He didn't defer against Pittsburgh in AFCCG.

But I think the premise is correct of having something standardised and the simple thing would be to say that the team which losses the coin toss at the beginning of the game, gets the choices in overtime i.e. they're the automatic winner of coin toss in OT.

In fact that seems super-fair as essentially each team "won" a coin toss.

96 Re: Clutch Encounters: Super Bowl LI

They can play 9 periods of hockey (it has happened twice in the NHL), and those are 20 minutes, not 15. I don't want to argue over whether one sport is tougher than the other, but if an NFL game had to go on that long, whether because neither team managed to score under the current format or because they changed the rules, the players would find a way to do it.

124 Re: Clutch Encounters: Super Bowl LI

Of course they would, but you are comparing standard vs. non-standard situations. Even in the the current format, games could potentially go on for several periods after regulation if no one scores. The question isn't about whether teams would "figure it out" in an unusual circumstance, it is about the best way to manage typical circumstances.

126 Re: Clutch Encounters: Super Bowl LI

By the third overtime, the hockey players are running on fumes, even with the intermissions to resurface the ice. There's not a lot of contact. It's mostly get to the puck, get it into the next zone, and if you can take a shot in the offensive end, do so. It's exciting, given that it's the playoffs, but it's not great hockey.

10 Re: Clutch Encounters: Super Bowl LI

Watching that game last night, I cringed when I saw Matt Ryan drop back to pass on 3rd and 1 with 8:31 because I'm reminded of Atlanta's comeback in 1998 against the Vikings, which turned on Chuck Smith beating Todd Steussie, getting the strip sack, leading to a Falcons TD and 6 point deficit at halftime rather than a 13 point lead.

Once Dont'a Hightower forced that fumble, I knew Atlanta was in serious trouble and their refusal to run the ball after Julio's catch was their version of the Gary Anderson missed field gpal, meaning when that happened it was game over for the Vikings or in this case, game over for the Falcons.

12 Re: Clutch Encounters: Super Bowl LI

"Had he let things go to the two-minute warning, he would have seen that the catch was good and that there was no need to blow his final timeout. "

I don't know why people keep repeating this nonsense. The patriots were lined up and set - if he wanted to challenge he couldn't wait, because they were going to run another play before 2:00

14 Re: Clutch Encounters: Super Bowl LI

Yes, Quinn had to make the decision without having time to properly review the replay. Of all the mistakes made yesterday, this one seems very forgivable.

And really, the lost timeout didn't hurt them. Well, not as much as many other things did, most notably the general cluelessness of their offensive coordinator.

16 Re: Clutch Encounters: Super Bowl LI

The other thought here, unless I am much mistaken, is that Quinn cannot legally throw the flag after the warning, because all reviews go to the booth within two minutes. Hence, to assured of a review, he had no choice but to throw the flag. Yes, the booth should certainly have reviewed the play if it had the chance, but there is no guarantee...

17 Re: Clutch Encounters: Super Bowl LI

I agree, completely forgivable decision to toss the challenge flag there. Pats are lined up, 2m warning is approaching so he can't know he'll get it reviewed, finally, it's basically the game there and it was such a ridiculous sequence that I can forgive a coach for thinking, even hoping, that the replay will show the tippy-tip of the football's nose hitting the ground before Edelman really got it in his hands.

49 Re: Clutch Encounters: Super Bowl LI

Once it became clear that the game would come down to a two-point conversion (and given the number of snaps Atlanta's d had been left out there, that should have been the supposition once New England had begun to move it on the final drive), I thought it was paramount to save a timeout for that play, but, really, on both conversions, New England looked vastly better prepared

69 Re: Clutch Encounters: Super Bowl LI

The point is that there was risk to "just waiting until the two minute warming" - the risk that the Patriots do get the play off (because they were clearly trying), and you don't get to challenge.

It was pretty clear on replay that Edelman caught the ball, but on the field it looked like exactly the sort of play you have to challenge.

24 Re: Clutch Encounters: Super Bowl LI

"Suddenly, we had a one-score game at 28-20 with 5:56 left, and the comeback started to feel inevitable at this point."

I'm curious to know what the author means by this. Is he saying that a team down 8 points at their own 9 yard line with 3:30 to go, facing the best offense in the league, should be expected to win in that situation? Or is it just that he's seen Brady pull off so many comebacks, that he expected him to do it again?

Because I don't see anything inevitable about a 91 yard TD drive in the final 3 minutes, followed by a successful 2 point try, followed by a 75 yard TD drive in overtime.

28 Re: Clutch Encounters: Super Bowl LI

He meant that it felt like everything was falling apart for Atlanta - their defense was exhausted, their offense was making mistakes, and both morale and momentum were in the Patriots favor. That last might not be something that stats can analyze, but it's very much a real factor, and a huge comeback in progress can really have an impact on how players play.

45 Re: Clutch Encounters: Super Bowl LI

So your position is that the comeback was easy, because the defense was tired? Do you feel the same way about Warner leading two late TD drives to tie against the Patriots, or Manning's comeback in the 2006 AFC championship? Because I can assure you, the Patriots defense was exhausted in both of those cases, yet I've never heard anyone talk about it in the past.

Does it really hurt you that much to just face reality, and acknowledge that Brady did an amazing job of leading his team down the field late in the game? Because it sounds like you're just making excuses.

53 Re: Clutch Encounters: Super Bowl LI

Um, does it hurt you so much to acknnowledge that defensive exhaustion affects defensive performance? Of course I feel that same way about the examples your provide (assuming the snap counts are what you imply) because I, trying to avoid being an idiot, acknowledge the effect of fatigue on atheletic performance. Why do you think the college ot system has so many times resulted in fnal point totals going up by 50% or more? It isn't solely because of where they place the ball. Defenses get exhausted, especially in the passing era of today, because of the nature of what is required of defenses, compared to offenses. Pass rushing is far more tiring than pass blocking, and the need for every player on defense to run to the ball until the whistle blows, means defensive efficiency drops extremely quickly in a way that offensive efficiency does not. These are facts, as much as it hurts your feelings. Are you insane?

Now that we have the invective, that you appear to desire, out of the way, this really is an area rich for quantitative analysis; 4th quarter defensive efficiency by leading teams, adjusted for snap count. How many times have we seen a defense enjoying a lead of 10 points or more, enter the 4th quarter having played more than 60 snaps, and what has their defensive efficiency been in the 4th quarter, compared to the rest of the game, especially when their offensive teammates aren't staying on the field? What happens to defensive efficiency as the snap counts passes 70, 75, 80?

You gotta stop taking this stuff personally.

61 Re: Clutch Encounters: Super Bowl LI

I went ahead and examined the snap counts.

Super Bowl XXXVI is a particularly terrible comparison. With 1:18 to go in the third and having run 45 plays, the Rams scored their first TD in 12 plays, taking 6:47 off the clock. They forced a three-and-out and got the ball back with 7:44 left now having run 57 plays, but they were forced to punt on 4th and 20 from the NE 49 with 4:01 left. Again New England went three-and-out and, the Rams now having run 66 plays, scored the tying TD in three plays. Meanwhile, New England's 69th play in this game immediately followed Ryan's fumble, when it was still a 16-point deficit. Not to mention the disparate quality of the 2001 Patriots and 2016 Falcons defenses.

The 2006 AFCCG is a better comp. When it was 34-31 and the Colts started their go-ahead drive, they'd run 75 plays. New England's 75th play in this game was the first play of the tying drive, so that's almost an exact parallel. Of course, the overtime factor is significant and the gap in quality of the '06 NE and '16 ATL defenses is again fairly sizable.

72 Re: Clutch Encounters: Super Bowl LI

To pile on here about the defense being exhausted, there's an interview anectdote (on 5th quarter?) with Devin McCourty about the end of the game.

The gist is that when the Patriots won the toss in overtime, Patricia started rifling through his playbook, but a bunch of defensive players grabbed him and told him the game was over.

Some of that is probably just regular old bravado, and some of it is probably just faith in Brady, but I don't think very many people expected Atlanta to be able to stop NE at that point.

That overtime drive had 9 plays - 7 successes, 5 10+ yard plays.

1,10: 6 yard pass
2,4: 14 yard pass
1,10: 18 yard pass
1,10: -3 yard pass
2,13: 15 yard pass
1,10: 10 yard run
1,10: 13 yard PI
1,2: incomplete
2,2: TD.

No third downs, only two unsuccessful plays (and one was in the context of having 4 shots at the endzone from the 2 yard line). They got rolled over.

77 Re: Clutch Encounters: Super Bowl LI

I'm not taking stuff personally. I just find it hard to believe that pulling off a comeback from where the Patriots were when they faced 3rd and 10 at their own 9 suddenly becomes easy, just because the other team is fatigued. The point about the Rams comeback is that the Patriots D was clearly gassed on the tying TD drive, whatever the snap counts say. That doesn't take away from what Warner did or Manning did in the examples I cited.

It just seems like you're trying to discredit an amazing comeback, and I don't see any reason to do that.

78 Re: Clutch Encounters: Super Bowl LI

Nobody is trying to discredit anything.

They're making the point that the Falcons defense being gassed was clearly a significant factor - and it was.

That doesn't make it EASY to drive 90 yards, but it does make it EASIER. It was still a fantastic comeback.

But I was a lot more sure they were going to win when they got the ball back at 3:30 than I would have been if Atlanta's defense hadn't already run 80+ plays.

83 Re: Clutch Encounters: Super Bowl LI

Find me instances where a top tier offense went up against a mid-tier defense, which was leading by 10 points or more, that had been on the field for 71 plays halfway through the 4th quarter, that managed to have successful stops, compared to how many time it failed to have successful stops. Then we can get some sense of the probabilities. My admittedly anecdotal experience has been that even top tier defenses fail with a high rate in such circumstances, which leads me to think that a mid tier defenses fail at a higher rate.

87 Re: Clutch Encounters: Super Bowl LI

I agree that the Patriots were more likely than otherwise to succeed because the Falcons were gassed. But down 8 91 yards away is still pretty dire. Perhaps we have different thresholds for what a "pretty high probablility" is, but it strikes me that you're understimating the difficulty of the task New England was facing.

Since we don't have any data, it's hard to know for sure. If you want to undertake a study of the scenarios you have mentioned, I will be happy to look at the results. But I see no reason to change my position until we see the numbers.

90 Re: Clutch Encounters: Super Bowl LI

91 yards makes it more dire.

Down 8 makes it less so though - because moving the ball is a whole lot easier when punting on 4th down isn't an option - the defense can't just defend the first down line on 3rd - because giving up a shorter play that gets the other team close still hurts you. It just changes the math a ton - and makes the threshhold for success on an individual play much lower.

100 Re: Clutch Encounters: Super Bowl LI

I will add that it wasn't one thing or the other that made the comeback seem inevitable, but all of it: the Pats offense was clicking, the Atlanta defense had collapsed, and to top it off, the Atlanta defense was pretty terrible to begin with and the Pats offense very good. If this had been Denver no one would have taken the comeback for granted, no matter if they were on offense or defense and no matter how many snaps they'd played.

I thought Atlanta's best chance would be to score again after the Pats if they left enough time on the clock and it could've worked. But stopping the NE offense seemed pretty improbable at that point.

103 Re: Clutch Encounters: Super Bowl LI

That's why I wonder why coaches, in those situations, don't send an offensive player out there to rush the passer. Julio wasn't doing anything or another WR could have probably done better than Beasley, who had been out there for hours trying to get by a much bigger man.

106 Re: Clutch Encounters: Super Bowl LI

Like bringing in your ace starting pitcher to get three outs in the 8th inning of World Series game 6 or 7? Yeah, I understand the thinking, but pitching is pitching, and I dont know if a guy who never works on pass rushing would be any better than an exhausted guy who does. It would be unthinkable to do it in anything but the 4th quarter of the Super Bowl; the last thing you want is your All-Pro receiver handfighting with offensive tackles.

109 Re: Clutch Encounters: Super Bowl LI

Could you put Jones or whoever on the edge and just tell him run around the tackle? (again break glass situation only) Kind of like when they put receivers back on 'hail Mary' plays?

I think Will you've long pointed out that Bill Walsh always said 4th quarter pass rush was key (and forgive me if I've confused you with another poster)

"If your defensive linemen tire in the fourth quarter, you've had it," Walsh said. "You can't win with a great offense and soft spots in the defense."

120 Re: Clutch Encounters: Super Bowl LI

I don't know much (or, really, anything) about blocking or pass rushing, but I do know a little about grappling and mixed martial arts - enough to know that generally, anybody who knows what they're doing will utterly crush somebody who doesn't, no matter their athleticism.

I suspect a similar rule applies for line play - Jones might be able to beat a 5'10" RB though sheer athleticism, but he'd in turn get tossed like a ragdoll by a TE. And that ignores the possibility of getting a penalty because he doesn't know the proper techniques.

131 Re: Clutch Encounters: Super Bowl LI

You would be much, much, better off with a benchwarming defensive player. I don't know the Falcons deep reserves well enough to name somebody, but a guy like Worrilow, who at least practices rushing the passer and is fresh, or another guy who just plays kick coverage or something, as opposed to a wide receiver. An light-ish offensive lineman would even be a better choice, as there's probably one on the roster who played some defense in college.

142 Re: Clutch Encounters: Super Bowl LI

One of the most underrated absences for the Super Bowl, as it turned out, was Adrian Clayborn going on IR after the Divisional Round with a torn biceps. He's not going to be mistaken for a Hall of Fame player anytime soon, but they could really have done with him absorbing his usual 50-ish snaps at defensive end. Dwight Freeney alone played around twice as many snaps as he usually does (56, he was usually around 28 in the regular season) at two weeks shy of 37 years old.

Clayborn's replacement on the active roster, Joe Vellano, played 22.

82 Re: Clutch Encounters: Super Bowl LI

Ramirez, it may be helpful to ask Will to offer his gut feelings for how aided Brady was by defensive fatigue. I sincerely doubt he is saying that NE was likely to win prior to the tying drive, just that it is a factor not properly accounted for in the averages.

178 Re: Clutch Encounters: Super Bowl LI

I assume you're being facetious in order to point out Gost's struggles. But he still made 94% of his XP's this season. You don't really think they had a better than 94% chance of converting the two-pointer?

And, by the way, the league-wide rate was 94% this year, so Gost was exactly average and should not be getting the that crap he is about it.

I personally hate the longer extra point because misses are such a fluky way to win or lose a game. Not sure why everyone thought it would be a good idea to have more missed extra points. But now that we have more misses, as we apparently wanted, everyone is hammering kickers for missing extra points. WTF?

183 Re: Clutch Encounters: Super Bowl LI

I love the longer extra point. It adds interest and sets up some intriguing situations. For the NFL, it adds some value to their commercials - which also means that people are watching. That's got to be a good thing.

As for the "randomness" of missed kicks, as a Patriots fan I know all about that. Missed kicks cost us a chance at a Super Bowl appearance, and damned near cost us a Super Bowl win. But there's no doubt that the longer extra points added drama to those games.

No one complains about the "randomness" of fumbles or pass completions or field goals or anything else in football, except maybe coin tosses. Kicking extra points is a skill play, not a coin flip. I love that. And I love that it involves special teams and the kicker in the game more.

185 Re: Clutch Encounters: Super Bowl LI

Actually I would, yes. Make a TD worth 7 points. If you like the drama and additional strategy of the two-point conversion, you just change it's point value to +1 if you make it and -1 if you don't.

The problem with the XP kick isn't just it's success rate, it's that it's not really a competitive play. The defense has only a miniscule chance of affecting the play so it's just not interesting.

And a miss is a fluke that just randomly robs one team of a point. It's basically saying that a TD is usually worth 7 points, but every once in a while for no reason whatsoever, it'll only be worth 6. How does that make sense?

190 Re: Clutch Encounters: Super Bowl LI

In fact, the easiest way to improve the game would be, it seems to me, to narrow the uprights by about 2-4 yards. It would change incentives by a rather large amount, and those changed incentives would make for lot more significant plays between offense and defense.

196 Re: Clutch Encounters: Super Bowl LI

That has merit, given how tremendously (and continuously) FG-kicking has improved, especially for long-distance kicks. The change in FG success may be as great (in magnitude, not importance) as the change in the passing game. I don't recall the numbers, but the 50+ success rate is significantly higher than 10 years ago. I just reread an account of the Colts-Giants OT game (because its game flow was surprisingly similar to SB 51 - meh then hair-on-fire), and the article noted that Pat Summerall's 49-yarder to force a playoff was the NFL's longest FG that season. It would not surprise me if the current kickers had better success on 50+ than the in-line guys had for all FG attempts.

57 Re: Clutch Encounters: Super Bowl LI

I think any estimate of win probability here incorporates the fact that Brady was the one leading the offense.

You could do probability conditioned on any generic offense against any generic defense needing to score a certain number of points in a given amount of time. Or you could estimate the probability of the Patriots offense, led by Tom Brady, facing that Falcons' defense, scoring the same number of points in that amount of time. The latter probability would be much higher.

Most of the WPs you'll see on the Internet incorporate information about the teams involved. They don't start the game at 50-50. It would be nice to also incorporate snap count information to get an even better estimate, but as always, the more variables you put into a model, the harder it becomes to collect the data, and the greater the risk of overfitting.

36 Re: Clutch Encounters: Super Bowl LI

For any team to win, the other team has to "leave doors open" for them. Since Brady has more wins than any other QB ever, I guess you could say it's true by definition. Otherwise, I'm not sure what such a statement even means. It smells mightily of trying to diminish Brady's accomplishments.

51 Re: Clutch Encounters: Super Bowl LI

When you supplement that statement with snide comments like:

I see another precarious New England lead hanging in the balance in the final minute, and while no Malcolm Butler interception this time, a stop in the red zone happens again. Because you know who willed it to happen.

that closed Scott's personal SB preview, his viewpoint becomes patently obvious.

63 Re: Clutch Encounters: Super Bowl LI

I do believe in clutch performance, except with a few caveats:

1. Clutch is primarily the absence (or reduction) of psychological pressure-induced performance decline (call it a "choke" in shorthand"), rather than any "elevation" in performance.
2. To the extent that performance is elevated, it is very likely what soldiers in combat often report as an adrenaline rush where everything in their surroundings seems to slow down and they are able to react faster and more effectively than usual. It is extremely rare, doesn't last long, and impossible to replicate consciously.
3. It is extremely difficult - and maybe impossible in many cases - to distinguish between "choking" versus any of a thousand other variables at play over the course of a game.
4. Professional athletes have already been self-selected for "clutch" by the time they get to the pros.
5. It is far more likely to lose clutch (due to a high-profile incident playing on their minds, sometimes off-the-field; see Tiger Woods) than to gain it.
6. Randomness happens. A LOT.

66 Re: Clutch Encounters: Super Bowl LI

Oh, I think golf is the game which most obviously demonstrates that choking is a real thing, and I think it exists in all sports, even in those team sprts where it is less easily observed and measured. I also think everybody chokes, just some less frequently than others, and the more teammates you have, the more likely it is that your choking will remain concealed.

149 Re: Clutch Encounters: Super Bowl LI

How about those players who are poor in practice but good in games? They clearly elevate their game under pressure. So why couldn't they elevate it even more if there is additional pressure? Also, if #4 were true, players would also self-select themselves for not choking.

173 Re: Clutch Encounters: Super Bowl LI

When I was in high school, the symphonic band always played much, much better during a performance than we did during practice. It wasn't stepping up under pressure, it was just so much easier to focus when you can see the audience and feed off their energy.

It's not really down to clutch-ness there.

199 Re: Clutch Encounters: Super Bowl LI

From what I've read home field advantage is mostly due to referees unconscious bias.
" ... refs have less information to help them resist the normal subconscious urge to draw on external cues for assistance in making borderline calls. In psychology terms, this process is called cue learning. ... The most common cue in sports is crowd noise, and because crowd noise almost always supports the home team, the way the fans sway the referees is the No. 1 driver of home-field advantage in sports. And one notable experiment suggests that how loud a crowd is helps refs decide whether an interaction should be penalized. A pair of German researchers showed actual referees old video clips of possible soccer infractions, with crowd noise played at high or low volume. Refs looking at the exact same interactions were more likely to hand out a yellow card when they heard a lot of crowd noise than when the volume was low."