/cdn.vox-cdn.com/uploads/chorus_image/image/51410935/615060066.0.jpg)
Where does confidence come from? Not from certainty. Even when confidence is well-founded, it can be a bit of a bluff. Confidence has a vulnerability—it’s a leaping over the vulnerability, perhaps, but the vulnerability is built in. Confidence is trust.
I never at any point believed the Seattle Seahawks were going to lose to the Atlanta Falcons on Sunday, but it wasn’t a blind faith—no religion—or denial about what was happening even as the Falcons scored 21 unanswered in the third quarter. It was a calm that came from trusting the dynamics that had played out in the first half were going to continue expressing themselves over the long run. It was a gambler’s confidence in superior odds overcoming occasional bad beats. It was also the privilege of staking out an early lead.
As much as Pete Carroll stresses winning in the fourth quarter—and this game Seattle did win in the fourth quarter, 9-0—the Seahawks mostly won because they won 17-3 in the first two quarters.
For one quarter, the Seahawks defense couldn't stop anything. The other three quarters, they showed why they are an elite unit.
— Mike Bar (@SeahawkScout) October 17, 2016
Now, I might have started worrying if the Seahawks had ever fallen behind by more than one score. Maybe. That’s something they still haven’t done in 2016 by the way.
My initial confidence came from having watched the tape during the week, and believing Seattle had the advantages in certain key matchups: I didn’t think Julio Jones presented an unstoppable threat, and I didn’t expect Atlanta’s defense could thwart the Seahawks’ many rested weapons. That first half mainly replaced my blueprint with evidence. I was mildly surprised the Falcons’ run-stuffing unit looked more like it did against the Denver Broncos in its week 5 win than its other early-season performances. Even so, Christine Michael, Sr., had already broken through for a few long gains unlike C.J. Anderson the week before.
Jones was held to two catches for 24 yards, and the Atlanta offense only crossed over midfield once—on a drive that started from the Falcons’ best field position of the half and after 12 plays still netted a mere field goal. Meanwhile, Seattle went beyond the 40-yard line all but one drive, and more than doubled (5.9 to 2.8) Atlanta’s yards per play as it opened up the early gap.
A 14-point deficit at halftime is not an uncrossable horizon, obviously. But the point of a two-touchdown lead is not either to inflate your Elo rating, impress the power-pollsters, nor boost your Pythagorean differential. As the Falcons well know, the point of a large lead is to withstand a comeback by the other side.
:no_upscale()/cdn.vox-cdn.com/uploads/chorus_asset/file/7303215/IMG_5266.jpg)
According to ESPN’s win-probability graph (these charts vary, but ESPN’s rely on Brian Burke’s calculations), the Seahawks’ peak likelihood of winning prior to Atlanta’s final fourth-down play came when it was about 93 percent during the first drive in the third quarter after a Matt Ryan incomplete pass.
That may seem to you like an awfully high expectancy for any outcome with that margin and so much time left, especially while the trailing team held possession, and it’s often said when these kind of projections flip that the estimate was wrong. Our brains tend to round percentages to integer ratios. 93 percent is more than 9/10 so this kind of legitimate edge gets often overrated as if it were unbeatable. In reality it’s not much different from holding two pair against one pair with more to flop in Texas Hold’em. Anything can change with live cards.
Indeed, Ryan then completed three of his next four passes for two first downs and the 36-yard touchdown to Jones that nobody covered. After Seattle went three-and-out and Ryan marched the Falcons down for another touchdown, suddenly the game was tied a little more than halfway through the third quarter.
:no_upscale()/cdn.vox-cdn.com/uploads/chorus_asset/file/7303229/IMG_5265.jpg)
What’s important to understand about these two quick touchdowns, however, is that although they erased the scoreboard deficit they didn’t erase the advantage the Seahawks had established in the first half. Note that when Atlanta evened the score, Seattle’s win expectancy remained above 60 percent. That’s partly because the Seahawks got the ball back, and partly because of Seattle’s inherited advantage as the higher-FPI team playing at home (WP at kickoff was approximately 70 percent).
Harder to recognize is that this tie-game scenario didn’t undo all the possibilities accounted for when the model proposed the Seahawks as 93 percent favorites. It represented a much narrower wedge of those outcomes, for sure, but since we know something greater than 7 percent of the timelines available at halftime included the Falcons scoring more than 14 more points (the minimum then required for Atlanta to win) and that Seattle remained a favorite even when they did, then (approximately, but conservatively) 10 percent of the Seahawks’ win probability at its crest involved outcomes where the Falcons tied the game as early as they did, or later took the lead.
Far from contradicting the model that was highly confident in Seattle to win, the vulnerability to an Atlanta comeback was built into that confidence.
Of course, the positive slice of that 93 percent grew even smaller when the Falcons scored a third-straight touchdown and then Atlanta’s own win expectancy went as high as 86.5 percent after Steven Hauschka missed the 29-yard field goal that would have brought the Seahawks within four in the fourth quarter.
:no_upscale()/cdn.vox-cdn.com/uploads/chorus_asset/file/7303237/IMG_5264.jpg)
In a probabilistic sense, the fact Seattle had earlier led 17-3 did not matter anymore. The game state was the thing, and even though there was 11:30 left to play the Falcons now held the ball and the full-touchdown lead. But from a personal-confidence standpoint, what happened in periods one and two still mattered quite a lot.
The concerning part of the first half had been the Falcons’ defensive resilience as they withstood medium-range marches by Russell Wilson’s gang before finally succumbing to a nine-yard strike after the sudden change caused by Cliff Avril’s forced fumble. When Atlanta’s defense tightened again briefly to begin the second half and Jones unloaded for five catches and 115 yards in the third, these factors conspired with Seattle’s bizarre lapse in coverage on Levine Toilolo to produce the Falcons’ outburst in that quarter.
However, as Seahawks Twitter melted down in panic, the above-mentioned advantages were already starting to reassert themselves and prove Atlanta’s short-term results unsustainable.
Indeed, the drive that ended in the missed field goal was a 10-play, 64-yard endeavor that reached the 10-yard line and made Dan Quinn’s defense look like the version of itself that had allowed an average of 28 points per game. A drive earlier, Seattle could have slowed the Falcons’ roll with the game still tied but Pete Carroll opted to punt on fourth-and-one from the Atlanta 45. Carroll claims he has to keep his hormones in check in these situations, but the coach should have more confidence in the percentages. The offense was working.
It’s 4th-and-1 for the Seahawks from the Falcons’ 45. I would go for it.
— NYT 4th Down Bot (@NYT4thDownBot) October 16, 2016
Meanwhile, the Falcons went 3-3 on third down in the third quarter. It’s easy to credit this detail to “adjustments” but why couldn’t it just be variance? Or why couldn’t Seattle, likewise, readjust on the fly? They were (it turned out) the only third downs Atlanta successfully negotiated all day. The Falcons’ offense has been the NFL’s best on first and second down in 2016, but as I pointed out last Friday they were a disappointing 39 percent converting third downs—partly because of Jones’s inconsistency—while Seattle had been best in the league at stopping those. The final tally, 3-11, matched the Seahawks’ season rate almost perfectly.
I won’t deny there were narratives and trends that justified worrying about heading into the fourth quarter. You could point at a number of injuries or other swings of fortune. You could use a psychological model: whatever degree momentum is real—or whatever is the opposite of momentum; for the Seahawks Sunday, maybe frustration.
Confidence is the antidote to frustration. The controversial last play for the Falcons was so critical, it’s worth noting, because it was the play after yet another third-down stop. Gift from the refs or not, the reason it cost Atlanta the game was because Seattle, through an accumulation of its advantages, had put itself in that position.
The Seahawks are now 51-17 over their last 68 games, including playoffs, beginning with the 2012 win over Chicago. That’s also 29-5 at home (and 21-11 on the road, after the 1-5 start to Wilson’s career).
I trust this team. Do you?