We use team statistics to create models that predict outcomes of NFL games. Our models generate either a win probability for the home team or a point spread, with each model generating a unique prediction. Our models consider many different metrics, such as turnovers, first downs, points scored, pace of play, offensive and defensive efficiency, as well as a few metrics that we have created ourselves. Beyond these statistics, some models consider factors such as “did a team make the playoffs last year?” and “how far is the away team traveling?” See this article for a more detailed breakdown of each model. The article below focuses on the four divisional round matchups for the coming weekend, but a breakdown of predictions for the entire 2016 NFL Playoffs can be found here, and our predictions from round 1 are here, where our average probabilities correctly predicted all four winners.
Round 2 Matchups
Seahawks @ Falcons (-4.5)
Our numbers have this game as a toss-up, with the average of our six models coming to 50.67% (where > 50% predicts that Atlanta will win and < 50% that Seattle will win). Those models that favor Seattle cling to their previous playoff success as well as their road strength of schedule and run defense from this season. Other models highlight Atlanta’s elite passing offense, advantages in points scored and rushing yards, and the fact that they are coming off of a bye (an advantage for all home teams this week). In the past two years, Atlanta is an unimpressive 6-7 as a home favorite, along with 3-10 against the spread (ATS). Their problem in those games was their defense. The Falcons gave up an average of 28 points per game at home this season, with each of their 8 home games going over the Vegas total. Seattle, however, has struggled scoring away from home (only 15.9 PPG this year), so it is certainly possible that the Seahawks simply will not be able to keep up with the Falcons. Seattle has played as an underdog only 7 times in the past 4 years, going 3-3-1 in those games (5-1-1 ATS). Ultimately, this is a game where contradictory information appears to be telling us that anything could happen.
The table below gives our predicted spread for this matchup, as well as win probabilities for the home team from six different models. The Model 1 value of 0.08 means that this model gives Atlanta an 8% chance to win (and Seattle a 92% chance)
| Home | Away | Our Spread | Mod 1 | Mod 2 | Mod 3 | Mod 4 | Mod 5 | Mod 6 | Avg |
| ATL | SEA | ATL by 10 | 0.08 | 0.58 | 0.53 | 0.35 | 0.64 | 0.86 | 0.51 |
Prediction: ATL 30, SEA 27
Texans @ Patriots (-16)
Houston, we have a problem. I wish we could offer some reasons why the Texans will shock the world here, but our models agree with Vegas (where New England is favored by 16 points). We expect the Patriots to obliterate the Texans this weekend, just as we expect Vince Wilfork to do to five Big Macs following his post-game presser. In this game, our numbers are driven by New England’s massive advantages in passing efficiency, points scored, red zone offense and defense, and turnovers. The Patriots have had an extra week to rest at home, while the Texans will travel to Foxborough, where they lost 27-0 earlier this year (before Tom Brady was back). While the weather currently projects to be clear on Saturday night, it should be noted that the Texans (whose home stadium is indoors) went 1-5 in outdoor games this year. As if we needed more reason to believe in the Patriots, they are 19-1 (14-5-1 ATS) as home favorites outside of their division (since 2013), while the Texans are 2-13 (3-12 ATS) as road underdogs outside of their division (since 2014).
| Home | Away | Our Spread | Mod 1 | Mod 2 | Mod 3 | Mod 4 | Mod 5 | Mod 6 | Avg |
| NE | HOU | NE by 17 | 0.85 | 0.73 | 0.72 | 0.54 | 0.75 | 0.93 | 0.75 |
Prediction: NE 31, HOU 17
Steelers @ Chiefs (-1.5)
This is another game with disagreement among our models, with the average win probability for Kansas City coming to 0.43 (giving the Steelers a win probability of 0.57). Pittsburgh has been a different team on the road in recent years, averaging only 22 PPG (compared to 30 PPG at home). Ben Roethlisberger’s splits follow a similar pattern: 2.9 TDs and 8.9 YDS/ATT at home, versus 1.0 TDs and 7.5 YDS/ATT on the road. However, as we pointed out last week, Le’Veon Bell has been gashing defenses all season long, and the Wild Card round was no exception as he put up 167 rushing yards. This does not bode well for the Chiefs as they’ve allowed opponents 4.4 yards per carry and 112 yards per game over their last 6 contests. However, Kansas City forced an average of 2.6 turnovers in home games this season, by far the best mark in the NFL. As a result, Kansas City has gotten a league-high 8 TDs from their special teams and defense. If Pittsburgh can take care of the ball and limit those big plays, the Chiefs may not be able to put up many points against a Steelers defense that has allowed only 16.6 PPG in their last 8 games. That being said, Kansas City is 15-4 (9-10 ATS) in home games outside of their division the last four years, while Pittsburgh is 11-10 in road games outside of their division over the same span (going 11-9-1 ATS). Here are a few interesting situational factors to pay attention to in this one:
- This will be Pittsburgh’s first road game in a month
- Andy Reid is 16-2 coming off of a bye in his career (including a dismantling of the Raiders in Oakland this season)
- Ben Roethlisberger is dealing with a foot “injury”
| Home | Away | Our Spread | Mod 1 | Mod 2 | Mod 3 | Mod 4 | Mod 5 | Mod 6 | Avg |
| KC | PIT | KC by 6 | 0.08 | 0.25 | 0.48 | 0.32 | 0.52 | 0.93 | 0.43 |
Prediction: PIT 21, KC 20
Packers @ Cowboys (-4.5)
I am still mentally recovering from picking the Packers to win last week. Thankfully, I do not have to do that again this weekend. Five of our six models have the Cowboys winning (with an average probability of 0.73), and our point spread model predicts the Cowboys should easily cover the spread. While the Packers may have a slightly better passing offense than Dallas, they certainly don’t blow them out of the water in any metrics. On the flip side, Green Bay doesn’t have a prayer of matching Dallas in terms of rushing yards, first downs from rushing, red zone defense, or punts forced (meaning they aren’t as good at ending opponent drives). The Packers are giving up 30.6 PPG in their last four road games, they will be without Jordy Nelson, and they’re only 2-7 (3-6 ATS) as a road underdog outside of their division (since 2013). Dallas is 9-2 (5-6 ATS) as a home favorite outside of their division going back to 2013, along with 6-1 (4-3 ATS) this season.
| Home | Away | Our Spread | Mod 1 | Mod 2 | Mod 3 | Mod 4 | Mod 5 | Mod 6 | Avg |
| DAL | GB | DAL by 12 | 0.46 | 0.85 | 0.81 | 0.62 | 0.76 | 0.86 | 0.73 |
Prediction: DAL 31, GB 24
Round 1 Recap
In round 1, all of the favorites won and covered. Our average win probabilities (taking the average from all six models) went 4-0, while Models 1 and 6 missed one game each. Our spreads went 1-3 against Vegas, but one of those had Oakland favored (we advised this could be misleading due to the Derek Carr injury).
| MODEL | RD 1 | RD 2 | RD 3 | RD 4 | TOT |
|---|---|---|---|---|---|
| Model 1 | 3-1 | – | – | – | 3-1 |
| Model 2 | 4-0 | – | – | – | 4-0 |
| Model 3 | 4-0 | – | – | – | 4-0 |
| Model 4 | 4-0 | – | – | – | 4-0 |
| Model 5 | 4-0 | – | – | – | 4-0 |
| Model 6 | 3-1 | – | – | – | 3-1 |
| Avg. | 4-0 | – | – | – | 4-0 |
| Spread (vs. Vegas) | 1-3 | – | – | – | 1-3 |
| Spread (Winner Correct) | 3-1 | 3-1 |