Skip to main content

Defense/Special Teams Start/Sit: When to Stream and When to Stick

The DST position is the one spot on a fantasy roster where loyalty is almost always punished. Unlike running backs or receivers, where consistency and talent create a floor, defensive units are wildly dependent on opponent, game script, and game environment — making the streaming-vs-sticking question both more important and more answerable than it might appear at first glance.

Definition and scope

In fantasy football, a Defense/Special Teams unit (DST) is scored as a single roster slot that captures points from sacks, interceptions, fumble recoveries, defensive touchdowns, special teams touchdowns, safeties, and points-allowed tiers. The last category matters enormously: most standard leagues apply a sliding scale that rewards 0-point allowances with bonus points (typically +10 in ESPN default scoring) and penalizes high-scoring opponents with negative values. That negative scoring potential is what makes the DST position structurally different from every other slot on the roster.

"Streaming" a DST means dropping a held unit and adding a matchup-favorable one each week rather than committing to a single team for the season. "Sticking" means rostering a high-quality defense — think a San Francisco 49ers or Dallas Cowboys caliber unit — and starting them regardless of week-to-week matchup variation. The full start/sit decision framework addresses how these two philosophies interact across all positions, but DST is where the tension is sharpest.

How it works

The streaming model operates on a simple premise: opponent quality is more predictive of DST weekly scoring than the defensive unit's own talent level. The basis for this comes from how fantasy DST points are actually distributed. A dominant defense facing a high-powered offense in a high-total game will frequently underperform a mediocre defense facing a backup quarterback in a low-total game.

Three factors drive DST streaming decisions:

Common scenarios

The streaming model is clearest in two scenarios. The first: a manager holds a middling defense — say, the New York Giants — who draw a top-5 offense in a game with a 52-point total. That combination alone is a release signal. The second scenario is the inverse: a premium defense like the Baltimore Ravens faces a rookie quarterback making his third career start in a game with a total of 38.5. That matchup justifies starting even a team whose season-long stats look pedestrian.

The "stick" case is narrower than most managers assume. It applies when a rostered defense ranks in the top 5 of actual defensive DVOA and faces a below-average offense or when waiver wire options are depleted — which becomes more common after Week 10 in most leagues. Bye weeks compound this problem; the bye week management section addresses the downstream effects on roster construction.

Special teams touchdowns introduce meaningful variance. A unit with a returner who averages more than 26 yards per kick return (a threshold that separates the top tier from league average, per NFL official statistics) adds a floor-raising element that pure defensive metrics miss.

Decision boundaries

The clearest decision rule: if the held DST's opponent ranks in the top 8 of offensive DVOA for the current season and the game total exceeds 48, streaming is correct in nearly every scenario — regardless of how strong the rostered defense appears on paper.

A structured set of thresholds:

The streaming vs. starting your roster discussion explores this same framework applied to skill positions, where the calculus tilts differently because talent floors are more stable. For DST specifically, the position's dependence on matchup over talent is the defining structural reality — something the DST start/sit strategy page addresses with positional depth. Managers who treat their defense the way they treat a must-start wide receiver will, on average, leave points on the waiver wire. The homepage at fantasystartsit.com aggregates the weekly tools that make these decisions faster and more systematic.