Week 1 is almost here! We’re just a few days away from getting to watch actual, honest-to-goodness football instead of just a flood of non-news about Ezekiel Elliott. If you heeded the advice of the FantasyPros Podcast (and play on a site that will let you do this), there’s a chance you didn’t draft a defense at all and spent that roster spot on someone like Devin Singletary. Now it’s time to set your week 1 lineup, and you might have someone like N’Keal Harry or Darwin Thompson burning a hole in your roster. I’m here to tell you what defense to pick up, with an eye toward streaming the entire year.

Dominate your NFL football pools with TeamRankings (FREE TRIAL) »

First, here are the guiding principles behind this strategy:

1. During draft season, we’re really bad at knowing which defenses will put up good season-long numbers. Last year’s consensus #1 defense, Jacksonville, finished 14th with an average of just 6.7 fantasy points per game. Of the top-five defenses in 2018 ADP, only one of them (the Rams) finished in the top five, and three of them finished 14th or worse. Combined, those top-five defenses averaged only 7.3 fantasy points per week. By contrast, if you used my highest-ranked player with ownership under 50%, you would have scored 8.9 points on average.
2. Who you’re playing against is much more important than who the defense is. The way we score defenses in fantasy doesn’t do a very good job of aligning the best real-world defense with high fantasy scoring. It turns out that the offenses ability (or inability) to score points and their tendency to turn the ball over are way more predictive than the defenses ability to prevent points or force turnovers. Last year, if your fantasy defense was “whoever’s playing the Bills this week” would have averaged 11.5 points per game. In fact, teams opposing the Bills, Cardinals, Jaguars, and 49ers would have ranked second through fifth among fantasy defenses last year.
3. Vegas is the best predictor of scoring. If you aren’t familiar, the two most common ways of betting on NFL games are against the spread picks and over/unders. For against the spread picks, sportsbooks will publish an expected margin of victory, and you will bet on which side of that line the result will land. For example, Vegas has the Jets at -3 over the Bills, meaning the expected outcome is the Jets winning by 3. If you bet on the Jets, you win if they win by more than 3. If you bet on the Bills, you win if they win, or if they lose by less than 3. Over/under betting is where books will publish an expected total score for both teams and you bet on if they will combine to score more or less than that number. The O/U for the Jets-Bills game is 40, so that is the expected total score. We can combine the total and spread to get an expected point total for each team. In this example, The Jets are expected to score 21.5 points and the Bills are expected to score 18.5. It turns out that these are the absolute best points projections out there because the books have millions of dollars riding on their accuracy. If you can consistently predict scoring better than Vegas lines, you should already be retired.
4. Turnovers and sacks matter a lot, so you should target bad quarterbacks. If your defense allows 14 points, that’s pretty good in today’s high-flying NFL. In fantasy though, that’s only worth 1 point. Most of the time, your defense will get most of its fantasy points from sacks and turnovers. Because about half of all fumbles occur on sacks, a whopping 80% of turnovers happen when the offense is trying to pass the ball. This means that facing a bad QBs who’s likely to throw a lot is exactly where you want to be.

Rankings Time

Without further ado, below are my projections for every team. The ownership shown is from Yahoo. If you have questions, feel free to hit me up on twitter.

Some Thoughts on the Top Picks

1. BAL @ MIA: We’re pretty bad at knowing which teams will be good before the season starts. Last year we thought the Broncos would at least be above average, but by mid season we were targeting them with fantasy defenses. Still, we have to go with our best guess given the information that we have, and right now we think Baltimore is a good defense and we think Miami is going to suck. There’s obvious risk here. We’ve seen Ryan Fitzpatrick play like Tom Brady when the stars align, but the more likely outcome is that he throws a bunch of interceptions. Miami’s projected point total of 15.5 is rock-bottom; we only see projections below 16 a few times a year. I would start worse defenses than Baltimore against Miami, and if you got them at a reasonable draft price, they’re a great choice Week 1.
2. NYJ vs BUF: When two terrible teams face-off, it’s common for both sides to be viable fantasy options. This almost always happens when the Vegas total is 40 or lower and the spread is within three. This is one such case, and the home team is usually the better choice. The fact that the Jets are 14% owned is criminal, and as a result, I’m starting them in almost every one of my leagues.
3. SEA vs CIN: This is the first Vegas line that surprised me – 9.5 points seems like a lot for what feels like two middle of the road teams. I guess the books think the Bengals aren’t much without AJ Green, and I’m inclined to take their advice. The Bengals don’t project to turn the ball over or get sacked quite as much as other top-tier opponents, but a projected point total under 20 is always worth targeting, and they’re at 17.
4. CLE vs TEN: The Titans are another team that could take a step forward, but knowing what we know now, my expectation is that they’ll be a tasty D/ST target again. And if my Madden campaign is going to be believed, the Browns are going to win the Super Bowl.
5. PHI vs WAS: Philly is tied with Seattle as the heaviest favorites of the week at -9.5. Case Keenum is bad at football and made Denver an excellent target last year. Look for him to do the same thing in Washington this year.
6. CHI vs GB: If you have the Bears defense, you probably drafted them pretty early for a defense and don’t really care what I have to say. There will be a couple of weeks this season where you might want to bench them for a streamer (@LAR Week 10, vs KC Week 15), but this is not one of those weeks.
7. DAL vs NYG: It doesn’t matter how generational your running back is, a team with an awful quarterback is an awful team. Expect everyone the Giants visit on the road to be a reasonable starter this year.
8. DEN @ OAK: The Raiders look to still be in rebuilding mode. They have some shiny new receivers, but it’s still Derek Carr throwing them the ball and Jon Gruden telling them what to do. Even though they’re at home and slightly favored, Carr’s interception projection more than makes up for that in fantasy.
9. NO vs HOU: This game could be a shootout, but Houston’s point projection is pretty average. They are, however, the most sack-welcoming team in the league. Houston allowed an incredible 3.8 sacks per game last year, which was firmly worst in the league. The next worst team allowed 3.3, and the league average was 2.5.
10. DET @ ARI: There are a lot of unknowns here because we don’t know how good Kyler Murray is. He could be this year’s Baker Mayfield, but he could also be the next Johnny Manziel. The Cardinals were awful last year with Josh Rosen, and it’s not like the supporting cast has gotten that much better. I’m perfectly willing to be against an inexperienced QB, but the risk is there.
11. KC @ JAC: Jacksonville is another team that was awful last year but upgraded their QB. Nick Foles is a known quantity though – like Ryan Fitzpatrick, he’s sometimes awesome and sometimes horrible. With the league’s best offense on the other side, there’s a likely scenario where Foles takes a lot of ill-advised downfield shots to try to stay in the game, creating a lot of opportunity for turnovers in the process.

Get free start/sit and waiver wire advice for your fantasy team

Subscribe: Apple Podcasts | Google Play | Spotify | Stitcher | SoundCloud | TuneIn | RSS

Jacob Herlin is a Senior Data Analyst for FantasyPros. For more from Jacob, follow him @jacoblawherlin.