Defenses are frustrating in fantasy. If you drafted Denver, the consensus #1 defense last year, and held on to them all season, you were pretty disappointed. They finished the season as the #17 defense in fantasy. That’s well outside of startable in a 10- or 12-team league. If you were somehow able to draft “whoever’s playing the Browns that week” as your defense, they would have finished #2 – only one fantasy point behind Jacksonville.

The reality is, it’s way more important who your defense is playing than how good they are. For that reason, a lot of people in the fantasy community advocate getting rid of defenses (and kickers) altogether. Still, that’s not entirely practical for everyone. The default settings on every fantasy site always include defenses, and good luck convincing your uncle that commissions your home league to drop them. “Defense wins championships,” he says.

As such, the strategy I and a lot of other people advocate is to stream them. Drop your defense and pick up whoever has a good matchup every week. If you have the bench space, plan ahead and pick up a team with a good matchup the following week, so you have less competition on the waiver wire. The purpose of this series of articles is to tell you which defenses you want to stream.

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

How Does it Work?
D/ST scoring comes from four places: points allowed, sacks, turnovers, and touchdowns. I include safeties in turnovers because (A) they are turnovers, and (B) they’re worth the same number of fantasy points. I project each of those stats separately and combine them into a total fantasy points projection.

Points allowed is simultaneously the hardest and the easiest category to predict. The state of the art projections for these are from Vegas lines. They’re about as good as it gets. Lots of people (myself included) have tried and failed to make more accurate points projections. The issue here is if you can reliably beat those projections by a non-negligible margin, you stand to make a lot of money betting on the games.

Nobody has more incentive to predict this well than the people working for sports books that wager millions of dollars on their predictions being unbiased. Points projections are hard because it’s basically impossible to do better than Vegas, but they’re also easy because nothing is stopping us as fantasy players from just using those projections directly, which is what I’ve done.

Going from a projection of points allowed to projected fantasy points is non-trivial. Tampa Bay is projected to score 20 points against New Orleans, which would be worth one fantasy point for the Saints D/ST. However, we have no illusion that this projection is a guarantee. There’s a range of possibilities for how many points Tampa Bay could score, and 20 is just the center of that range. To deal with this, I model points with a Poisson distribution, which allows me to calculate the probability of Tampa Bay scoring in each range (0-6 points, 7-13 points, etc.), and find that the expected value of fantasy points is 0.7, as a result.

The other three categories – sacks, turnovers, and touchdowns – are all projected based on stats over the prior year (for both the team in question and the opponent), whether the team is playing at home, and Vegas lines. While a point projection from Vegas is not a sacks projection, there is enough of a correlation between points allowed and sacks achieved that Vegas lines do have a role to play in that projection.

Just Give Me the Rankings Already
Hey, I know why you’re here. If you just want the projections, here you go:

 
The ownership data above is for Yahoo leagues. I included the model’s projections for the Eagles-Falcons game, even though this article is going up after it. Vegas overestimated the scoring in this game, with an O/U of 45.5 points compared to an actual total of 30. The Eagles D/ST was probably started in a lot of leagues, and if you did that it ended well, with 10 fantasy points thanks to four sacks and an interception, holding the Falcons to only 12 points. The Falcons, who were higher ranked (though still not recommended), scored three fantasy points on two sacks and 18 points allowed.
Tell Me About the Top Picks

• Baltimore, New Orleans, Jacksonville, and Pittsburgh are all strong defenses against, as far as we know right now, not-very-good offenses. It’s Week 1, so anything can happen — maybe Nathan Peterman proves something. Maybe Saquon Barkley elevates the Giants offense so much even Jacksonville can’t handle them. Anything is possible, but all four of these teams should be started in every league this week.
• Detroit might not have as good a defense as the four teams I just listed, but boy is that a juicy matchup against the rookie QB-led Jets at home.
• Minnesota against the Jimmy Garoppolo-led 49ers at home is also pretty nice. There’s a little more risk because everyone in the fantasy community expects Garoppolo to be good, but Minnesota’s defense is very strong. Vegas isn’t as high on Jimmy G, though — the 49ers got the same number of points as the Browns and the Buccaneers. If you have the Minnesota D/ST, you probably spent a bit of draft capital on them, and shouldn’t hesitate to start them right away.
• Denver hosting Seattle is another risky one. Denver’s defense is solid, but Russell Wilson is amazing. Vegas is giving this game surprisingly few points — just 42, the second-lowest of the week — for a game that could turn into a bit of a shootout if Denver fans get what they want from Case Keenum and Wilson starts throwing bombs to Doug Baldwin to catch up.
• The Chargers hosting Kansas City is a nice high-upside play, thanks to Patrick Mahomes. He has the honor of the highest turnover projection this week. If a time-traveler told me that one team scored 25 fantasy points off multiple pick-sixes, I would guess it was Baltimore. My second guess, though, would be the Chargers. That upside comes with downside, though. In addition to throwing picks, Mahomes is known for throwing the ball a mile. He also happens to have the fastest man in the NFL, Tyreek Hill, as his top receiver. There’s a world where Mahomes throws some bombs 10 yards deeper than anyone slower than Hill can get to.
• There are two notable low ranks (not including the Philadelphia): the Rams and Houston, both of which are owned in at least 90% of leagues.
  • The Rams aren’t concerning — Vegas seems to think this is going to be a shootout, but I have low expectations for the dynamic duo of Jon Gruden and Derek Carr. If you drafted the Rams, it’s probably not worth the bench spot to pick up a streamer, and you don’t want to drop them.
  • I cannot for the life of me figure out why Houston is owned in 90% of leagues. They were the fifth-worst fantasy defense last year, and the New England matchup is as bad as they come. If you drafted them based on having the second-best strength of schedule, that calculation is based on last year’s numbers in a flawed way. It’s primarily driven by two facts: (1) they have few terrible matchups after New England, in favor of several average ones, and (2) they play Indianapolis twice, who was the fourth-best matchup for defenses last year. That is not going to happen again.

Get a free analysis of your team with My Playbook

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.window.PLAYERCARDS_CONFIG={affiliateCode:”fpros_cards”};