NFL franchises use contextualized data to create competitive advantages. In order to realize an edge, teams need to employ the right data in the right way at the right time. This means distilling, interpreting and applying only the most influential data in a framework that accounts for their personnel, their opponents and their evolving game situations. My goal is to be your analytics department. Each week this season, I want to work for you by giving you a peek into which numbers flag in my models as the most impactful ... or the most misunderstood.
As always, let me know if your eye test is picking up on something interesting, or if there's a stat/trend you'd like me to take a deeper look at. You can hit me up on Twitter @cfrelund. As with any great analytics department, the more collaborative this is, the more value we can create.
With four weeks left in the regular season, the Saints are the only team to have officially earned entry into the postseason by clinching the NFC South title. The Ravens, Patriots, Chiefs, Packers, Seahawks and Niners -- who all reach the playoffs in over 96.6 percent of my model's simulations -- should also be planning on playing in January.
This season, wild-card berths and seeding projects to come down to the wire. In order to make sense of it all, I applied my vetted model to the rest of the regular-season games to find out which teams were most likely to make it -- and which games mattered most. I ran 6.4 million simulations (100,000 for each remaining game) using the most predictive metrics from the past 15 seasons, factoring in current team personnel, trends, play-calling and matchups.
These are sequential simulations, meaning I'm not just adding the 100,000 game simulations for each team together each week, but rather looking at four-game stretches and reflecting patterns that similar teams, personnel and matchups have followed in the past. (Think: number of road vs. home games, playing in the cold or wind for a dome team, weeks since bye, playing several high-pressure defenses in a row vs. spacing them out, etc.) My model assigns a confidence percentage to each team in each game. The sequential model also has a confidence level for the four-game stretch.
For all of the 2019 games so far, when my model projects a win confidence of 60 percent or greater, it has been correct 74.5 percent of the time (70 games correct, 24 incorrect). Of course, there is always a chance, in any given week, that a heavily favored team loses, or a major injury significantly changes a team -- all of which could and would shift the playoff outlooks.
However, given the best predictive information available right now, my model shows that the following eight games matter most in determining who will be playing in the postseason and what the seeding will be:
WEEK 14
Sunday, 1 p.m. ET (FOX)
This game could very well determine who gets a first-round bye and maybe even home-field advantage through the playoffs. Should the Niners upset the Saints in New Orleans (my model favors the Saints in only 52.2 percent of simulations), San Francisco increases its chances to land a first-round bye by almost 18 percentage points (from 55.2 percent with a loss to 72.9 percent with a win). The likelihood that the Saints miss out on a first-round bye, meanwhile, would increase by about 20 percentage points (New Orleans gets either the No. 1 or No. 2 seed in 60 percent of simulations in a loss, compared to 79.2 percent of simulations with a win). If the Saints win this game, they would be the most likely to earn the No. 1 seed in the NFC, thus securing home-field advantage.
Sunday, 4:25 p.m. ET (CBS)
Kansas City (leading the AFC West at 8-4) and New England (leading the AFC East at 10-2) have all but secured January games, and while the Chiefs are unlikely to earn a playoff bye, they could spoil the Patriots' chances to have home-field advantage throughout the playoffs this Sunday. If the Chiefs pull off the upset (my model projects this game as a near coin flip, with the home-team Patriots narrow winners in 50.8 percent of simulations), the Patriots project to earn the No. 1 seed in only 18 percent of simulations, falling to the No. 2 seed in 74 percent of simulations. If the Patriots win, their chances to earn the No. 1 seed will be 56 percent.
Sunday, 4:25 p.m. ET (CBS)
The loser of this game is essentially eliminated from the postseason. The 6-6 Raiders are two games behind the 8-4 Chiefs in the AFC West, meaning their best path to the playoffs is to nab a wild-card berth. The 7-5 Titans have two remaining matchups with the Texans (who lead the AFC South at 8-4) and host the Saints, meaning they have a tough road to the playoffs but do still have a chance to overtake Houston and win their division. (Right now, I have the Titans making the playoffs in 50.2 percent of simulations.) My model has the Titans winning in 57.6 percent of this week's simulations.
The outcome of this game has implications for other teams in the AFC chase, with a Titans win negatively impacting the Bills' and Steelers' odds to secure playoff berths and favorable seeds. Pittsburgh fans should certainly be rooting for the Raiders in this one, as an Oakland win increases the Steelers' playoff odds by about 7 percentage points, from 49 to 55.7 (so long as the Steelers also win in Arizona this week).
Sunday, 8:20 p.m. ET (NBC)
In the NFC, it's very likely that both wild-card teams will have at least 11 wins, with the No. 5 and No. 6 seeds going to the 10-2 Niners and 8-4 Vikings in the current playoff picture. If the 7-5 Rams are going to edge out the Vikings for the final wild-card spot, they'll need to beat the Seahawks at home this week. My model projects the Rams to win this contest in 51 percent of simulations. If Seattle wins, the 'Hawks will add about 5 percentage points to their chances of securing a first-round bye (50.2 percent to 55.4) -- and the Rams will basically be eliminated from the playoffs, with a 6.2 percent chance to make the postseason.
WEEK 15
Sunday, 8:20 p.m. ET (NBC)
The Bills currently make the playoffs in 79.2 percent of simulations, presuming probable losses in Week 14 (at home versus the Ravens) and Week 16 (to the Patriots in Foxborough). This sets up Week 15 as pivotal to their chances of securing a playoff berth and/or determining which seed Buffalo ends up with as a wild-card team. (With the Chiefs likely to end up as the No. 3 seed, the wild-card team that lands the No. 6 seed will be in line to travel to Kansas City.) Right now, my model favors the Bills in this game, with Buffalo beating the Steelers in 51.7 percent of simulations, though that number will change after this week's games.
WEEK 16
Sunday, 4:25 p.m. ET (FOX)
The NFC East will likely be won by a team that doesn't finish the season with double-digit wins. Ahead of Week 14, the Cowboys (who currently lead the division) project to reach the playoffs in 57.9 percent of simulations, while the Eagles reach the playoffs in 51.9 percent. Looking ahead, the 5-7 Eagles face the 2-10 Giants twice and 3-9 Redskins once, giving Philadelphia a strong chance to win three additional games. The 6-6 Cowboys, meanwhile, have the 6-6 Bears, 7-5 Rams and Redskins, with the odds favoring Dallas to win two of three. Thus, Dallas and Philly could both finish with 8-7 records outside of this matchup, making the Week 16 clash essentially a play-in game. Note that a Cowboys upset of the Rams in Week 15 wouldn't sap this game of too much juice; if Dallas entered at 8-6 and Philadelphia entered at 7-7, the Cowboys would still need to beat the Eagles to secure the division.
Monday, 8:15 p.m. ET (ESPN)
The 9-3 Packers' remaining schedule is as follows: host Washington and Chicago, then travel to Minnesota and finish the regular season in Detroit. The 8-4 Vikings will be at home in three of the four remaining weeks, hosting Detroit, Green Bay and Chicago (they also have to make a trip in Week 15 to Los Angeles to play the Chargers). This game has less to do with the Packers' chances and more to do with Minnesota as a wild-card team -- and, more specifically, which wild-card team. The Packers project to top the Vikings in Week 16 in fewer than 56 percent of simulations. If the Vikings do lose this matchup, they most likely set up a third meeting with the Packers in Green Bay in Round 1 of the playoffs, with the Packers on course to snagging the No. 3 seed and Minnesota falling to No. 6.
WEEK 17
Sunday, 4:25 p.m. ET (FOX)
With Seattle and San Francisco both currently sitting at 10-2, this game is very likely to determine who will have a playoff bye as the NFC West champions and who will be the No. 5 seed, most likely traveling to Dallas on Wild Card Weekend, rather than getting a chance to rest. The difference in terms of making it to the Super Bowl? Big. I ran my playoff models 100,000 times for each game in each round. Earning a bye and home-field advantage for at least one game is worth an increase of about 31 percent in Seattle's chances to reach Miami and about 26 percent in the Niners' chances (based on their probable matchups). So, basically, the team that wins in Week 17 is about 30 percent more likely -- thirty percent! -- to make the Super Bowl. That's massive.
Follow Cynthia Frelund on Twitter @cfrelund.
