Week #1’s predictions were, um, somewhat successful. I went 4-0 and beat the point spread on three games. The big miss was thinking the Giants-Packers game would be a nail-biter (Green Bay won by 25).
I doubt I’ll do that well again– these games are trickier to predict– but let’s see what we can do.
Seattle (10-5-1) @ Atlanta (11-5)
If you’ve been reading me for any length of time, you know that I’m data-driven. I’m relentless about paying attention to the evidence– ignoring intangibles like veteran savvy, momentum and emotional motivation . In this game, the home team (who usually win sin the playoffs) has a point differential (points scored minus points allowed) more than double the visitor (134-62). The Falcons have a more balanced offense (the Seahawks’ running game has been a train wreck most of the year), they have a huge edge in takeaway-giveaway ratio (+11 to +1), they had a week’s rest and have all their key players healthy (Seattle safety Kam Chancellor is out of the playoffs).
So naturally I’m picking the Seahawks.
There are two intangibles I pay attention to. The first is I’m skeptical of teams who run up the score. The Falcons scored 540 points this year– seventh on the list of most points in a season (since 1990). None of the six teams above them won the Super Bowl. The only teams of the six to make that list were the top two, and they both lost: The 2013 Broncos (606 points, first) lost 43-8, while the 2007 Patriots (589, second) lost 21-17.
Of the other teams in the top 10, the 1999 Rams won the Super Bowl (but they squeaked by a team coached by Jeff Fisher). So only three teams made the Super Bowl, and they went 1-2.
In the second 10, the 2009 Saints (509 points, 13th), 1994 49ers (505, 14th) and 1998 Broncos won, while the 2001 Rams (503, 15th), and 2015 Panthers (500, 17th) lost. That’s more teams and more wins 3-2. From 21-30, we have Washington in 1991 winning and nobody else even playing.
The lesson of this is not, I believe, “Defense wins championships”, so much as “An offense that rolls up the score in some games– but struggles in others– is likely to disappoint you.” The 2013 Broncos scored:a record-setting 606 points, but allowed 399, but they weren’t consistent about it. This table gives you the breakdown of points scored, their record, the total margin of victory and the record of the opponents. Most of their points came against teams who simply weren’t very good.
Outscoring opponents by 207 points is impressive, but 112 points of that margin came in the first two splits, against teams who went a collective 43-53 (an average record of 7-9).
Obviously the Broncos were a good team– a bad team doesn’t go 13-3. But they went 4-3 against teams who made the playoffs in thge regular season, and clearly were better at fantasy football than the real kind.
The Broncos are a representative sample of the clubs who finished very high in points scored. The Falcons are a substantially less impressive version of this breed to team:
Only one of the five teams they hung 40+ on had a winning record (Tampa, 9-7). The other four were New Orleans (7-9), Carolina (6-10), the Rams (4-12) and San Francisco (2-14).
Two of the teams they scored 30 against were playoff teams (Oakland and Green Bay), but the others were Arizona, San Diego and the Panthers and Saints again.
Facing four winning teams– Tampa and Denver (both 9-7), Kansas City (12-4) and Saturday’s opponent, they managed to beat only Denver. The best you can say is all three losses were by seven points or less, so they could have gone either way.
The final split is a loss to Philadelphia– which had lost two in a row, four of its last five and would go on to lose five more games.
So I don’t see the Falcons as a great team– or even a great offense, I see them as the team Terry Bradshaw once called “a great fantasy league team.”
They last made the playoffs in 2012, and a lot of the team has turned over, so you can’t assume they’ll play well in post-season– you have to treat them like a new team.
Then we get to the second intangible thing I look at: They never played well in the playoffs. They went 1-4 in fact. Three of the games were close, but two were stone beatings.
The only positive you can say about that period was that the one win was against Seattle.
The only reason I’m sweating this pick is that history has shown that I consistently overestimate the Seahawks when I make picks. They’re third in defense (292 points allowed– 114 fewer than Atlanta’s 406), but 18th in offense (354 points).
The Seahawks have had only six games where they ran the ball well. Last week they got 177 yards. If you want to believe the swill the announcers were spouting during the Lions game, they finally came into their own, thanks to Thomas Rawls recovering from the injury that kept him sidelines for eight weeks.
The hitch with that is that Rawls looked pretty good when he came back in week 11 (14 carries for 57 yards– 4.1 yards a carry), then had 38 yards on 12 carries (3.2). He had two weeks were he averaged more than 5.6 yards a rush… but against Carolina and Green Bay. He was under 2.0 a carry in the next three games (against the Rams, Cards and 49ers).
He got 161 yards on 27 carries against Detroit, but the Lions allowed opposing rushers 4.4 yards a carry. That would be an average of 106 yards. Rawls broke a 32-yarder and a 26-yarder– that’s the difference. If he goes back to his 2.0 average, then it’s Russell Wilson against “Matty Tank” (RB Devonta Freeman gaiend 40 yards on 12 carries when they played during the season.
The Seahawks can probably win that battle. They won 26-24 (in a game Rawls missed), with his backup getting hey won 26-24 without Rawls. Seattle scored 10 points in the last eight minutes; Matty Tank tossed a clutch interception between the two drives.
Atlanta is hot– they’ve won five of their last six– but they were 4-1 going into the loss to the Seahawks
The Falcons could win– and my repeated overestimation of Seattle makes me nervous. But the combination of an offense with inflated totals, the poor record against winning teams, the difference in experience and the history of underachieving makes me lean to the team with the neon green shoes
Prediction: Seattle 27, Atlanta 21
Houston (9-7) @ New England (14-2)
Well, that preview took longer than I wanted these to go. The good news:
- If I use the same framework for the Green Bay-Dallas game, it’ll be familiar
- This game won;t take long.
I can’t even construct a rationale for how Houston could win this game. To begin with, they’ve allowed more points (328) than they scored (279). You can say “They have a pretty good defense (it’s 11th– plus, 21 of those points scored on returns, when the defense wasn’t on the field)….
But New England (250 points) allowed fewer points than any other team. The Patriots allowed 6.8 yards per pass (eighth); the Texans average 5.9 (last). The Patriots allow 3.9 per rush (tenth); the Texans average 4.1 (tied for 18th).
New England has twice as many takeaways (23) as giveaways (11)– and four giveaways came in the first four games, when Tom Brady was suspended, and quarterbacks with no career games started played. Houston has a -7 ratio– they’ve lost the ball 24 times and taken it 17.
New England has the third-best offense (441 points), and it has been consistent:
- Two games with 40 points (both 41 points)
- Six with 30-35 points
- Six between 22-27 (two of those with the scrubs playing)
- Two games below 20 (a loss with a scrub and a 16-3 win in Denver, with 18-degree temperatures and gusting winds)
Houston coach Bill O’Brien has had a mess at quarterback all three seasons, and he has a really fine one going into this game. Tom Savage, their #4 pick in 2014 played in the last three games of the season, and led a come-from-behind win in game 14 and a win in game 15.
Savage started game 16, got a concussion after throwing only 8 passes– his replacement, Brock Osweiler lost the game. Houston won, but not due to anything Osweiler (14-25 for 168 yards) did. Had Derek Carr (not rookie Connor Cook, who threw three picks and went 18-45), the Rauders probably win.
So with the guy who’s been playing well healthy, O’Brien is starting Osweiler. This pretty much guarantees that people will be calling for the backup with every incompletion the starter throws.
Maybe the Texans canbeat Brady up– that’s how Denver beat New England in the playoffs a year ago. New England doesn’t have a great line; DE Jadaveon Clowney (7.5 sacks) and LBs Whitney Mercilus (6.0) and Benardrick McKinney (5.0) can put on some heat.
The problem is, even if they knocked him out, Jimmy Garoppolo and Jacoby Brissett might be better quarterbacks than anyone the Texans have. The Patriots beat Houston 27-0 with those guys playing; hard to see a closer game with Brady starting.
Prediction: New England 35, Houston 3
Green Bay (10-6) @ Dallas (13-3)
This game might not be any closer than the game involving the other Texas team… although there is more reason to doubt.
Dallas could be 14-2, but when they had the first seed locked up going into the final week, they benched RB Ezekiel Elliott and played QB Dak Prescott, WR DEz Bryant and TE Jason Whitten sparingly.
Dallas is fifth in the league in both points scored (421) and allowed (306). They project to 11-5, which is partly due to the no-show in game 16 and partly due to a 7-2 record in close games (indicating some good fortune). The offense was a little spotty:
The strength of schedule is a little misleading– the 30-39 includes the 1-15 Browns; the 20-point range includes the 2-14 49ers. Take those teams out, and the Cowboys went 11-3 against teams with .500 records or better.
Interestingly, two of the three losses were to New York Giants– the team Green Bay thrashed last week. The Giants shut down Prescott and won 20-19 and 20-7.If Packers coordinator Dom Capers can figure out how to emulate whatever Steve Spagnuolo did, they might have an upset.
Another reason for optimism: the Packers were 24th in defense (388 points), but were fourth in offense (432 points). Their offense was more consistent than Dallas’s:
The thing that makes the Packers look bad is the three-week stretch from weeks 9-11: a 31-26 loss to the Colts, a 47-25 beating by Tennessee and then a 42-24 loss to Washington. That’s three losses to weaker teams– two out-and-out beatings– with 75 points scored (below their average) and 120 allowed– a negative margin of 45 points.
If I could find an explanation for it, I’d be more willing to write the games off. The Packers turned the ball over six times (five in the games where they gave up 40+ points); both Tennessee (DeMarco Murray) and Washington (Robert Kelley) were way over 100 yards rushing. They just looked bad for a while– they had five losses in six games, counting those three games, a 33-32 loss to Atlanta and a 30-16 loss to Dallas– and went 9-1.
The big hope for Green Bay is the mismatch in post-season experience– especially at quarterback. The Packers have made the playoffs for the eighth consecutive season; Aaron Rodgers is a veteran who has had a bad post-season only once (2011). The Cowboys have been to the playoffs once in the previous six years; Dak Prescott is a rookie. I like Prescott very much, but some rookies fall flat in their first post-season; Elliott is unproven as well.
On the other hand, Green Bay can’t run. With James Starks and Eddie Lacy out, they’re using Seattle castoff Christine Michael and converted WR Ty Montgomery. And WR Jordy Nelson has three broken ribs and won’t play.
Dallas won by two touchdowns in the regular season; it’s hard to think the game will be much.
Prediction: Dallas 31, Green Bay 13
Pittsburgh (11-5) @ Kansas City (12-4)
This looked like it would be a very easy game to pick.I assumed the Steelers (399 points scored and 327 allowed; tenth in both rankings) would take the Chiefs (389 scored and 311 allowed). Andy Reid was 10-7 in the playoffs in his first ten seasons– and 1-4 in the last seven. QB Alex Smith is streaky. The numbers are also slightly skewed: Pittsburgh took the last game off against the Browns, but the Chiefs had to win to secure a first-round bye.
When the two teams played in week four, the Steelers won 43-14, with Ben Roethlisberger throwing five TD passes and Le’Veon Bell running for 144 yards on only 18 carries.
But after the Miami game, Roetrhlisberger limped into the postgame interview room wearing a boot and all bets are off. The Chiefs have had a lot of injuries to their front seven, but they can still rush the passer– and claim to LBs Justin Houston, Tambia Hali and SS Eric Berry healthy.
The Steelers have their best offensive line in eons… but if Roethlisberger can’t move, they can beat the daylights out of him. If it gets into a battle or running backs– Bell against Spencer Ware and Charcandrick West (Jamaal Charles is out yet again)– the Chiefs are very likely win that one. And there is no way that Landry Jones can step in and win.
But we don’t know how well Roethlisberger can play.
Roethlisberger is, in my opinion, the most tenacious quarterback of the era. He has played more games hurt– with more serious injuries– and done it at a higher level– than anyone else. Quite often he has played through self-inflicted wounds– finishing games where his team had a comfortable lead, playing meaningless late-season games when he should have been resting for the playoffs and/or throwing deep passes (and taking hits) when a handoff to a checkdown would be as useful.
But the guy is a moose. He can play effectively when he can barely walk on and off the field. He heals so rapidly that I sometimes wonder if he has adamantium in his skeleton. He’s gotten less foolhardy as he ages. When asked during the interview, he said the boot was just a precaution– that he wanted to make sure he didn’t get seriously hurt.
And he’s immeasurably better than Smith, whose close to being a deluxe “game manager.” During the regular season, Smith threw 489 passes. 15 were touchdowns; 8 interceptions. He produced 3,502 yards (7.2 per throw).
Roethlisberger threw only 20 more balls. He threw 14 more rouchdowns, only 5 more interceptions and gained 314 more yards (7.5 yards). Roethlisberger was sacked 11 times less, lost two fewer fumbles.
The only advantage Smith had was running the ball (five scores to Roethlisberger’s one). But the Chiefs didn’t have a stronger running attack– they gained 12 yards less, had a lower rushing average (4.2 to Pittsburgh’s 4.3) and lost 9 fumbles to the Steelers’ 3. They did score two more rushing touchdowns– but at a substantial cost.
But none of that tells us whether Roethlisberger will be healthy enough to outplay Smith. I’m guessing he won’t be. Roethlisberger is 34– two years older than Smith. he didn’t have a great game against Miami (he had two long TD passes to start the game, but then was intercepted twice). Had Matt Moore not fumbled twice and been intercepted once– and been unable to move the ball– Pittsburgh might not have won.
Kansis City is a much better team than Miami; it will make fewer mistakes and get Pittsburgh to make more. I think that will be the difference. This could be my biggest miss of the week– Roethisberger plays as well as he did in week four and the Steelers blow the Chiefs out. But I don’t think he can snap back again.
Prediction: Kansas City 24, Pittsburgh 13