For years, the New Orleans Saints held back Drew Brees from winning his second Super Bowl. Now, could the biggest obstacle for the 2019 Saints actually be their star quarterback?

They enter the season with a 15.7% chance of winning the Super Bowl per ESPN's Football Power Index, odds that are second in the NFL behind the Chiefs, but those numbers surely rely upon the future Hall of Fame quarterback continuing to play at his accepted level of greatness.

In 2018, Brees was great ... until he wasn't. Through the first 11 games, he put together what might be the best stretch of play we've ever seen from an NFL quarterback, with the Saints starting 10-1 while averaging more than 37 points per game. Over his final four games of the regular season and in the Saints' two postseason contests, though -- sandwiched around a Teddy Bridgewater start in Week 17 -- Brees looked ordinary:

Overnight, he went from playing like an MVP to putting up numbers in the range of quarterbacks such as Derek Carr and Case Keenum. And while it was lost in the shuffle of the infamous missed pass-interference call against Nickell Robey-Coleman late in the fourth quarter of the NFC title game, Brees didn't play all that well in that loss to the Rams, averaging just 6.2 yards per attempt and turning a pair of short fields into just three points before throwing an interception on the first possession of overtime.

Do the Saints have something to be concerned about with their star quarterback? At 40 years old, Brees is in barely charted territory; the only other passers in league history to make it through a full season as a starter in their age-40 campaign are Tom Brady and Brett Favre. The good news for New Orleans is that both played well and made deep playoff runs, although Saints fans won't need to be reminded of how Favre's 2009 season ended.

Normally, I like to go through the research before detailing what I've found, but I'm going to stick the findings up here first this time. The best answer to the Brees question I just posed is "maybe." I don't think what happened at the end of 2018 necessarily indicates that Brees is going to be at the Carr level again in 2019. Simultaneously, though, there's no guarantee we will see the old Brees, because we just don't have a big enough sample of Hall of Fame quarterbacks getting to this point of their careers in the modern era to know what triggers their decline. Brees could fall off and become a middling quarterback for reasons totally unrelated to what we saw last season. So could Brady. Father Time arrives without warning.

Perhaps most important for Saints fans, even given that I expect New Orleans to decline in 2019, I still think the Saints are a viable Super Bowl contender if the Brees from those final few weeks shows up, given their strength elsewhere on the roster. Figuring out which edition of Brees will show up for the majority of 2019 is virtually impossible, but in going through a series of questions related to that bigger question, I think we can get to some interesting stuff about aging, quarterback play and what might happen with the Saints this season.

What happened at the end of 2019?

After watching those games again and looking over the numbers, there's not a clear, obvious cause for what happened to Brees after Thanksgiving. The best explanation is a confluence of factors coming together to drive the decline, including:

Regression toward the mean. In short, it might not be realistically possible for any quarterback to play for a full season as well as Brees had over the first 11 games of 2018. He wasn't due to struggle -- that's the gambler's fallacy -- but his true talent baseline isn't completing more than 76% of his passes while throwing nearly 15 touchdown passes for every interception, because nobody is that good.

As an example, take Brees' competition at the time for the MVP award, eventual winner Patrick Mahomes. Through the first 12 weeks of the season, Mahomes posted a Total QBR of 82.6, a passer rating of 117.9 and averaged just over 10.0 adjusted yards per attempt, or AY/A. (I'll be using AY/A, which counts touchdowns as worth 20 additional yards and interceptions as a 45-yard penalty, as an improvement upon passer rating in this piece. You can read more about that statistic here.)

After Week 12, Mahomes declined, albeit far more gently than his competition in New Orleans. From Week 13 on, Mahomes posted a Total QBR of 72.9, a passer rating of 103.6 and averaged 8.7 AY/A. It's just hard for any quarterback to be as good as Brees (or Mahomes) was during the first three months of the season, so it would have been natural to expect some drop-off from the Saints star over December and January. Obviously, Brees declined more than we would have expected given his history, so that's not enough of an answer on its own.

The road trip. Brees has exhibited a pronounced home/road split during his career; the Purdue product has averaged 8.1 AY/A and thrown 2.7 touchdowns for every interception at home, but Brees has fallen off to 7.1 AY/A while averaging 1.8 scores for every pick on the road. His decline started with three consecutive games on the road before a home game against the Steelers, in which he went 27-of-39 for 326 yards with a touchdown in his best outing of the final six. (He sat out the meaningless Week 17 game against the Panthers.)

Brees' road drop-off is typically attributed to weather after moving away from the climate-controlled confines of the Superdome, but I don't think weather played a big part here. His road trip included a game against the Cowboys in the Jerry Dome, a 71-degree afternoon in Tampa, Florida, and a 48-degree night game against the Panthers in Charlotte, North Carolina. His coldest game of the season was a 39-degree road game against the Bengals in early November, when the future Hall of Famer went a near-perfect 22-of-25 for 265 yards and three touchdowns. And when you include the two playoff games, Brees played exactly as many games at home as he did on the road during this six-game decline, so traveling doesn't tell the full story, either.

The deep passes. Brees was an absolute savage on deep throws during his dominant start, but that mastery seemed to fade as the season went along. The NFL defines deep passes as throws traveling 16 or more yards in the air, and during the first 11 games, Brees completed more than 60% of those throws, averaged nearly 17 yards per attempt and tossed nine touchdown passes against just one pick.

During his six-game end to the season, though, Brees completed just 41.3% of those passes, averaged 10.8 yards per attempt and threw two picks against one lone touchdown pass, which came on his first deep attempt of that stretch, to Keith Kirkwood. During one stretch against the Panthers and Steelers, Brees was just 1-of-9 for 19 yards on deep throws.

In watching Brees actually throw, though, I don't see a quarterback who was struggling to hit his receivers downfield. He was playing well and generally putting the ball in the right place at the right time. That 1-of-9 stretch included potentially six perfect or nearly perfect throws, with those passes either dropped, broken up by defenders who made great plays or caught.

The Saints coaches didn't appear to think that Brees was struggling to complete deep throws, either, given that the veteran quarterback was dialing them up about as frequently as he had been during that early run. He averaged 5.7 deep passes per game during the 11-game start to the season, and he averaged five deep throws per game through his next five starts. The sixth game was the loss to the Rams, which is a good representation of what Brees dealt with here.

The Rams played a fair amount of two-deep coverage and tried to force everything underneath, which is why Brees was 11-of-13 on throws to Alvin Kamara in the NFC Championship Game. Brees took four deep shots during that game. One was a perfectly placed pass to tight end Dan Arnold, only for the backup to drop a would-be touchdown on the final attempt of the opening drive. The second was a third-and-17 dig route to Ted Ginn Jr., who got open, only for Brees either to expect Ginn to settle on the route or throw behind his receiver. The third was the 43-yard pass to Ginn in the fourth quarter that set up the fateful non-call, and while that throw hung on Brees a bit, he put the pass where it needed to go. The fourth and final throw was his last pass of the season, the interception he threw while being hit in his motion (and arguably in the helmet) by Dante Fowler Jr.

As with the bigger picture, it's not as though Brees just forgot how to hit his receivers downfield. He naturally regressed toward the rest of the league and had some bad luck in a small sample.

A red zone decline. Again, the Saints were just running too hot to keep going at their pre-Thanksgiving rate. Brees & Co. averaged 5.7 points per red zone trip through the first 12 weeks of the season, which is ridiculous, even by their standards. For context, between the first 12 weeks of the season from 2013 to 2017, the Saints averaged just under an even 5.0 points per red zone trip. From Week 13 on, the Saints hit 4.9 points per red zone trip, which is right at the league average over that same time frame.

Brees' performance didn't slip much in the red zone over that stretch, either by the numbers or on tape, although he did put a slant to Michael Thomas low two plays before the infamous no-call against Robey-Coleman. They dropped one of his 72 red zone pass attempts during the first 12 weeks of the season, only to put three of Brees' 32 attempts on the ground afterward. Arnold dropped a would-be touchdown in the first quarter against the Rams, Ginn dropped a quick out in the same game, and Tommylee Lewis dropped a slant on a pick play inside the 5-yard line against the Cowboys which might have gone for a score. The Arnold drive ended in a field goal, and Smith's drop was on first-and-goal on a series which ended with Kamara getting stuffed at the goal line. Both games were decided by three points.

Pressure. The Saints lost star left tackle Terron Armstead in that 51-14 victory over the Bengals, but there's not a real clear correlation between Armstead's absence and Brees' drop-off. They scored 79 points in the first two games Armstead missed, wins against the Eagles and Falcons. Armstead did sit out the three-game road trip, but he was back on the field for Brees' three subsequent home starts to end the season.

Furthermore, Brees wasn't pressured at a significantly higher rate during his decline. He was actually pressured more frequently during his hot start (20.8%) than his subsequent falloff (19.5%). What changed was how successful opposing teams were when they did pressure Brees. Before Thanksgiving, teams turned just under 13% of their pressures against Brees into sacks. Afterward, that figure nearly doubled, to 25%.

The pressure also told with takeaways. After throwing two interceptions through 12 weeks, Brees threw five over the ensuing six games. Three were a direct result of pressure -- this short-armed checkdown to Kamara against the Cowboys, a rushed screen pass against the Buccaneers and the season-ending interception against the Rams.

The other two were different stories. One was a very good throw to Arnold against the Panthers, only for the ball to be dislodged by Thomas Davis' helmet and picked off by James Bradberry. The other was an underthrown bomb on the first play of the game against the Eagles on a play that should have resulted in a long Ginn touchdown, only for Cre'von LeBlanc to take away the pass.

What about that play against Atlanta?

In an attempt to draw a clear line between Brees' two seasons, some have pointed to the Thanksgiving night win over the Falcons and the tackle Brees attempted on his second interception of the season. After an ill-advised throw was picked off by Damontae Kazee, the Falcons safety attempted to return the pass for a pick-six. Kazee should have been (and eventually was) ruled down at the point of the catch, but without a whistle being blown, Brees attempted to take on blocking Falcons corner Desmond Trufant and was whipped sideways as Kazee ran past.

The timing on a macro level makes sense, given that Brees' second season started with the Cowboys game the following week. The problem with that theory, though, is that Brees threw that interception against the Falcons in the first quarter. Over the remainder of the game, he went 10-of-15 for 115 yards with three touchdowns and no picks. The Saints didn't need much from Brees in a game during which they ran for 150 yards and the Falcons turned the ball over four times, but when they called on their star quarterback, he was efficient and extremely effective.

Is it possible that Brees rode an adrenaline high (or a painkilling shot) in the game, woke up sore, and then wasn't the same player after that Thursday night game? I suppose, but Occam's razor suggests that Brees wouldn't have been his usual dominant self for three quarters after the hit before suddenly declining the following week against Dallas. I'm skeptical the hit played a significant role in his falloff.

Did Brees simply get tired?

Every quarterback wears down some as the season goes along, but I don't think there's a strong case for fatigue as the most likely cause of the decline. Brees doesn't have a notable track record of declining as the year goes along. Over his career, he has roughly posted the same AY/A in September (7.37 AY/A) as he has in December and January regular-season games (7.35 AY/A).

Could Brees' age make him more likely to gas out late in the season? Perhaps. I went back through 2001 and split the league's quarterbacks into four groups by their age as of Sept. 1 in each season to see if older quarterbacks were more likely to decline toward the end of the season than their younger counterparts. Brees fell into the camp of quarterbacks who were 36 and older, and the rest were split up with 31-35-year-olds, 26-30-year-olds and the 25-and-younger crowd.

When you compare their AY/A in December and regular season-games from January to their performance in September, October and November, the elder statesmen do drop off by about 6.6%. It's not enough to explain the entirety of Brees' decline, given that his regular-season AY/A fell off by about 33% between the first three months of the season and his three-game stretch in December, but it's probably some small part of the equation.

With that being said, it's hardly as if the Saints overworked Brees. The same quarterback who once routinely topped 450 passes between September and November -- topping out at 492 in 2012 -- threw only 384 passes through his first 12 games between September and November this time around.

So, has Brees lost it?

He could come back as a mediocre quarterback in 2019, but I don't think last season's falloff is enough to prove anything. For one, he wasn't even that bad; he wasn't that far off statistically from league average over his six-game stint. Brees looked better than his numbers on tape. All of the infrastructure surrounding him returns besides Mark Ingram, and the Saints actually upgraded offensive weapons by adding Jared Cook.

Brees has also come back from worse stints in the past. I put together a rolling five-game average of AY/A for each of his games with 10 or more pass attempts and then constructed a weighted league average to measure how Brees compared to the rest of the NFL's quarterbacks. I didn't include postseason games, but in the final five games of 2018, Brees was 7.5% below league average.

Even given that he was as high as 44.1% above league average earlier in the season, 7.5% below average just isn't remarkably bad. Brees was 3.7% below average over a five-game stint in December 2016 after throwing three interceptions in back-to-back games and nine over a five-game stretch. He showed no ill effects the following season. In October 2015, Brees dropped to 7.2% below average and then proceeded to throw 28 touchdowns against nine picks the rest of the way. Randomness and the aforementioned factors in a small sample are more likely to have driven Brees' decline than an inescapable, terminal plunge. My most likely outcome for him is that he's about halfway between league average and the lofty heights of his 11-game start to 2018.

With that being said, though, there's a chance Brees does drop to replacement level in 2019. Every quarterback in this age bracket has some chance of turning into a pumpkin in any given season without warning. Even if I don't think those final few games represent a meaningful indicator, history tells us Brees has some chance of falling out of contention in 2019.

What happens to Hall of Fame quarterbacks as they approach/pass 40?

Generally, they're great until the bottom suddenly falls out. We don't have many modern quarterbacks who really compare to Brees, so we don't have a huge sample of what happens at the end of their careers. From what we've seen, things generally don't end well. When you look at these past two generations of star quarterbacks, their careers seem to end one of three ways:

• Serious injuries: Troy Aikman (retired after age-34 season), Steve Young (38)

• Significant, sudden decline: Brett Favre (41), Jim Kelly (36), Dan Marino (38), Peyton Manning (39)

• Above-average play before leaving on own terms: John Elway (38), Joe Montana (38)

The one guy who doesn't really fit into any one group is Warren Moon, who was a Pro Bowler for the Seahawks in his age-41 season before being benched for Jon Kitna the following year and spending two seasons with the Chiefs as a backup. Aikman could also fit in the serious decline group, given that he had been right at or above league average for his entire career before taking a major step backward and posting a 78 AY/A+ in that final year before his concussion issues.

When my colleague Mike Triplett asked the Saints earlier this month whether they were concerned about Brees' disappointing end to 2018, they responded in the way you might expect: by expressing their confidence that Brees would continue to be great. Quarterbacks coach Joe Lombardi even referenced Nicholas Taleb's book "Fooled by Randomness" while suggesting what happened to Brees was more due to random chance than any sort of pattern. I mostly agree.

Where I disagree with Lombardi, though, is in what he said to Triplett next. "Listen, we didn't just score 48 points and then all of a sudden, one day later, age caught up to him," Lombardi said. "That's ridiculous, it's absurd."

I don't think it's quite as unlikely as Lombardi is suggesting. At the very least, we have several recent examples of quarterbacks playing at a world-class level before suddenly, inexplicably declining, including a strong comp for what happened to Brees last season:

• Favre threw 33 touchdown passes against seven picks during his age-40 season in 2009 and put an exclamation point on his season with a four-touchdown, 148.7-passer rating game against the Giants. He threw for four more touchdowns in a 34-3 blowout of the Cowboys in the divisional round before he famously threw a pair of picks in the dramatic NFC title game loss to the Saints. The following season, Favre threw just 11 touchdown passes against 19 interceptions and subsequently retired.

• Kelly was above league average at 35, including a 237-yard, four-touchdown game against the Rams in December 1995. The Bills made the playoffs and scored 37 points in a wild-card game win, although he wasn't spectacular. The following season, a 36-year-old Kelly threw 14 touchdown passes against 19 picks, his sack rate spiked to 8.9%, and the future Hall of Famer subsequently retired.

• Marino was an above-average passer at age 37, including a 355-yard, four-touchdown performance in Week 16 of the 1998 season against the Broncos. He had a couple of big games left in him, including a 393-yard, two-touchdown performance against the Colts in Week 4 of the following year, but Marino threw 12 touchdown passes against 17 picks that year, struggled to stay healthy, and retired after a 62-7 playoff loss to the Jaguars.

• The most obvious comparison might be Manning, who was an MVP candidate during his age-38 season in 2014. Manning threw for 257 yards and four touchdowns in a 39-36 win over the Dolphins. He was averaging 323.5 passing yards per game with 34 touchdowns against nine interceptions through that Dolphins win; over the final five games of the regular season, the former Colts star averaged 233.8 yards per game with five touchdowns against six picks. Manning struggled in a playoff loss to his old employers and threw for nine touchdowns against just 17 picks in a 2015 season where he was benched for Brock Osweiler.

Aikman and Young both retired as a result of injuries. The only guys who rode off on their horse into the sunset while playing well were Elway, who left after winning his second consecutive Super Bowl, and Montana, who had come back after missing the better part of two full seasons with injuries. Brees is now older than both were.

In looking through the end of each player's career, there's no sort of trend or indicator I can find that represented proof of their coming decline. If I do the same rolling AY/A average for each of them as I did for Brees, they resided in similarly lofty heights to Brees before declining rapidly in their final season.

As a Broncos player, Manning only dropped below league average over a five-game sample by Week 17 of the 2014 season, when his AY/A was 5.1% below the mark. Despite the fact that he was 34.2% above league average over a five-game stretch as recently as Week 9, Manning never made it back above league average. Does that mean Brees will follow in Manning's footsteps this upcoming season? It's possible, but one dominant quarterback's career path isn't proof that Brees will follow suit.

The instructive way to think about quarterback aging, I think, is like this: As a Hall of Fame-caliber passer enters his late 30s, there's some percentage chance that he'll suddenly lose it and become a replacement-level quarterback. Those chances grow greater with each advancing year of aging. The risk increases for passers with injury histories and for quarterbacks whose infrastructure collapses around them. Neither of those categories fit Brees, but as a 40-year-old quarterback, there's just not enough of a sample size to say much about his chances of declining.

Is there a 10% chance Brees drastically declines and is barely playable in 2019? I buy that. How about 20%? Maybe. Or 50%? That seems too high. It seems crazy to think that the bottom could suddenly fall out for a quarterback as good as Brees (or, for that matter, Brady), but the same was true for these other Hall of Fame-caliber quarterbacks too. The thing about falling off a cliff is that you don't see it coming.

Can the Saints win even if Brees does crater?

They can, and I don't have to go back very far for an example of how to pull it off. I brought up Manning's 2014-15 decline as the most plausible comparison for a Brees drop-off, but you also probably remember what happened as Manning went from an MVP candidate in 2014 to Osweiler-grade in 2015: The Broncos won Super Bowl 50! They had to bench Manning for Osweiler and then sub Manning back in for part of Week 17 and the playoffs, but the Broncos rode home-field advantage to the Super Bowl and then handily beat the Panthers in Santa Clara 24-10.

Manning was a passenger on that trip. He was barely passable even during the postseason, completing 55.4% of his passes while averaging 5.8 AY/A. He didn't even fill the classic game-manager role by holding on to the football, as the future Hall of Famer turned the ball over three times in three games. He finished 11th among the league's 12 quarterbacks that postseason in Total QBR, with a mark of 28.4.

The Broncos won that Super Bowl because of what was around their quarterback, namely, the defense. Denver's defense forced seven takeaways in three games and absolutely beat up opposing quarterbacks. Von Miller & Co. racked up 14 sacks and a staggering 33 knockdowns over three games, including 17 knockdowns of Tom Brady in the AFC Championship Game. Injuries helped, too; the Steelers were without Antonio Brown and Le'Veon Bell and had a gimpy Ben Roethlisberger after a brutal wild-card game against the Bengals, and the Patriots were without starting left tackle Nate Solder for most of that season.

It's not difficult to imagine the Saints winning games on the strength of their defense. The 2015 Broncos led the league in defensive DVOA, coming in 25.8% below league average. (With defensive DVOA, below average is good.) The Saints were only 11th in the league last season, with a DVOA 2.9% below league average, but they improved mightily after replacing Ken Crawley with Eli Apple. Over the second half of the season and into the playoffs, New Orleans boasted the fifth-best defense in football on a per-game basis by win probability added. Given its young talent on that side of the ball, it's not out of the question that this defense sticks as one of the top five defenses in football in 2019.

Of course, the Saints' offense also should be better than that of the 2015 Broncos, even if Brees does fall off. In Teddy Bridgewater, the Saints should have a better backup than the Broncos did with Osweiler. Their running game is also likely to be better than that Broncos team was at running the ball if Alvin Kamara stays healthy, though they'll need to adjust after losing center Max Unger to retirement. That Broncos team ranked 25th in offensive DVOA; a Saints team with Bridgewater under center for half of the season should theoretically be able to top that mark.

"I don't know" isn't the most satisfying answer anyone can give for what Brees will do in 2019, but it's the most realistic response to the question. Anyone who tells you they know for sure how Brees will play is lying. Given how he has played at a high level for the vast majority of his career, my best guess is that he'll continue to maintain a Pro Bowl-caliber level of play, minus some moderate penalty for aging. After what Brees showed before and after Thanksgiving last season, though, the error bars for any projection are far larger than they would be for most veteran quarterbacks. Everything from an MVP award to a benching-worthy campaign is possible for him this season.