Deshaun Watson Knew He Had Problems. He Fixed Them In Week 5.

Deshaun Watson entered Week 5 trying to rebound from yet another sack-filled performance. Over the past two seasons, he’s been dumped more times than any other passer. But rather than point fingers at his team’s porous offensive line, he resolved to address his role in the problem.

“I held the football too long, and I fumbled the ball and things like that,” Watson said after the Houston Texans’ 16-10 home loss to the Carolina Panthers on Sept. 29. “The O-line did a hell of a job. That’s me. I’ve got to check the ball down, take what they give us.” He also said that he failed to “scan and pick up the blitz.”

On Sunday, Watson’s improvement those areas was so dramatic that it led to one of the greatest passing performances in NFL history. In the Texans’ 53-32 victory over the Atlanta Falcons, Watson’s passer rating was perfect over 33 pass attempts, of which he completed 28 for 426 yards and five touchdowns. That was tied for the most attempts with a perfect passer rating since the 1970 AFL-NFL merger. And his 98.2 raw QBR (not adjusted for opponent) was the 14th-best ever for a quarterback with at least 30 pass attempts.

Most importantly, the Falcons didn’t sack Watson even one time. It was just the second sackless game in Watson’s career. (In his other 26 games, he’s been dumped 99 times, or nearly four per game.) In the two games in his career without a sack, Watson threw 10 touchdown passes on 53 attempts with no picks. The strength of the opposing defense matters in contextualizing how well Watson performed, obviously, but when he can avoid these negative plays, Watson looks more like Patrick Mahomes and less like Luke Falk.

The first way Watson improved was simply in his time to attempt a pass. From the start of 2018 though Week 4 of this season, when Watson was sacked 80 times in 20 games, he attempted 53 percent of his passes less than 2.5 seconds from the snap, compared with the NFL average of 57 percent.⁴ But in Week 5, Watson reinvented himself as a much more decisive and quick-trigger QB: He threw 73 percent of his 33 total attempts less than 2.5 seconds after the snap.

Importantly, he did not cut his time to throw by throwing shorter passes. Watson was his typical downfield self, averaging 8.73 air yards per attempt compared with 8.67 in the prior period. (The NFL average since 2018 is 7.89.)

But what about his efforts to improve his blitz recognition? We measured this by examining how often a quarterback faces an unblocked rusher. Though this can happen because of miscommunication among the offensive linemen, Watson suggested that, as quarterback, he’s the one responsible: He has to make pass-blocking adjustments to account for pass-rushers and also identify any potential rusher not assigned a blocker at the snap.

The Falcons still beat the Texans’ protection with at least one unblocked rusher at a slightly higher rate than Watson had faced previously: They broke through nine times in 36 dropbacks, or a rate of 25 percent, compared with 21 percent of Watson’s dropbacks from 2018 through Week 4.

But Watson recognized the danger and beat the blitz with well-timed passes. On the nine plays where he faced unblocked rushers, he was 9-for-9 for 136 yards and a touchdown, posted a near-perfect 99.7 QBR and had a passer rating of 155.8. Not only was Watson not sacked, but he didn’t even need to run — which meant that he could avoid exposing himself to the punishment of being tackled. And keeping Watson pain-free has been a challenge for the Texans: He was beaten up so savagely last year that he couldn’t fly in for a road game. (He had to take a bus.)

Sacks aren’t merely a threat to the quarterback’s health — they also suffocate an offense. I reported last year in The Wall Street Journal that on a series of downs without a sack, the chance of getting another first down (or touchdown) is 74.7 percent. A first-down sack lowers that probability to 36.7 percent, and a sack on second down drops it to 22.1 percent.

And that’s with an average quarterback. Watson is a superior QB who can seem unstoppable when he avoids getting dumped in his own backfield. On Sunday, he’ll probably need to keep his uniform relatively clean again at Kansas City to win a showdown with Mahomes. And now, it’s the Chiefs who are facing protection woes, with Mahomes getting dumped more times in Week 5 against the Colts (four) than previously all season (three). Perhaps unsurprisingly, the result for Mahomes was the lowest point total (13) of his career.

Looking Ahead: Week 6

Best matchup: No. 4 Philadelphia at No. 10 Minnesota (-1.5), 1 p.m. ET Sunday
Matchup quality: 87th percentile⁵
Matchup evenness: 82nd percentile

A couple of weird trends collide in the best game of Week 6. As we alluded to in Monday’s NFL chat, the Vikings only ever seem to beat the teams they’re supposed to; so far this season, Minnesota is 3-0 as a favorite in our Elo model and 0-2 as an underdog. (It’s part of a long-running issue under QB Kirk Cousins.) Meanwhile, you never know what to expect from the Eagles, who have won games as underdogs — like when they beat the Packers in Green Bay two weeks ago — and lost as heavy favorites — like when they fell to the Lions at home the week before that. So something’s got to give in this matchup. Elo says the Vikings have the edge (55 percent chance to win) at home … but the Eagles are also a good opponent, the likes of which Minnesota has struggled against in the Cousins Era. Since last season, the Vikings are 1-6-1 against teams with an Elo rating over 1550 going into the matchup. Then again, that one victory? It was against the Eagles in Week 5 of last season. See, these teams are weird! But we’ll learn more about each of them Sunday.

See also: No. 7 San Francisco at No. 3 L.A. Rams (91st/56th); No. 6 Seattle at No. 18 Cleveland (62nd/77th)

Biggest playoff implications: No. 7 San Francisco at No. 3 L.A. Rams (-3.5), 4:05 p.m. ET Sunday
Potential shift in playoff odds: 28.9 total percentage points

It’s been musical favorites in the NFC West recently. Los Angeles entered last week in the driver’s seat, but the Rams were also playing the week’s most important game — against the Seahawks — and Seattle won, briefly making it favorites to win the division. Then, San Francisco’s Monday night demolition of Cleveland made the Niners favorites, which is where things currently stand. But that could shift again after Sunday’s game between the 49ers and Rams. According to our model, both teams’ playoff odds will shift by double-digit percentage points no matter what. If the Rams win, they’d gain 13 points, and the Niners would lose 11 points, though each team would still be more likely to make the playoffs than not. But if San Francisco wins, look out: The 49ers would gain 17 points of postseason probability, giving them a commanding 85 percent chance of making the playoffs. And the Rams would see a massive 20-point decline in their playoff probability, making them only 37 percent likely to return to the playoffs and defend their NFC title. Our model gives the Rams a 62 percent chance of winning Sunday — and thereby making that apocalyptic scenario moot — but if they don’t, the consequences could be big.

See also: No. 16 Detroit at No. 5 Green Bay (26.4); No. 4 Philadelphia at No. 10 Minnesota (25.0)

Best QB duel: No. 1 Patrick Mahomes (KC) vs. No. 4 Deshaun Watson (HOU)

See also: No. 7 Aaron Rodgers (GB) vs. No. 11 Matthew Stafford (DET); No. 2 Matt Ryan (ATL) vs. No. 21 Kyler Murray (ARI)

FiveThirtyEight vs. the Readers

As a weekly tradition here at FiveThirtyEight, we look at how our Elo model did against everybody who made picks in our forecasting game. (If you entered, you can find yourself on our leaderboard here. I am currently in 1,080th place!) These are the games in which Elo made its best — and worst — predictions against the field last week:

After back-to-back losses against the field in Weeks 3 and 4, Elo got its revenge in Week 5. The readers had a few strong picks — most notably in rightly downplaying Tennessee’s favorite status in the Music City Miracle rematch (Buffalo won) and setting Cincy as only the slimmest of favorites against Arizona (and the Bengals did indeed disappoint). But that was offset by the readers’ overconfidence in Chase Daniel and the Bears, Baker Mayfield and the Browns, Gardner Minshew and the Jags and Jared Goff and the Rams. Altogether, the average player lost to Elo by a margin of 19.8 points in Week 5, the widest margin either side has had in any week so far this season.

Still, congrats are in order to Mark Nitsche, who led all readers in Week 5 with 229.0 points, and to Jamie Porter, whose total of 440.9 points leads the full-season contest. Thanks to everyone who played — and if you haven’t, be sure to get in on the action! You can make picks now and try your luck against Elo, even if you missed Week 5.

Check out our latest NFL predictions.