NFL is the most popular — and most beatable — DFS sport, because the people who win don't pick "good players." They build correlated, leverage-aware lineups around how a game will actually unfold. Here's the complete CourtEdge method, from bankroll to the final roster lock.
The average player opens the optimizer, sorts by projection, and rosters the nine highest-projected players that fit the cap. That lineup loses. Not because the projections are wrong — because a roster of disconnected studs ignores the two forces that actually decide tournaments: correlation (which players score together) and ownership leverage (being right in a way the field isn't).
NFL is uniquely suited to both. Scoring is bunched — when a QB throws for 350 and 3 TDs, his receivers feast with him; when a game becomes a shootout, both passing offenses erupt together. The winners harness those relationships on purpose. The losers roster a great QB and a great defense facing that same QB and wonder why they never cash.
1. Bankroll & contest selection (do this first)
Lineup construction gets the attention, but bankroll management is what separates players who last from players who reload. Two rules carry most of the weight:
- Cap your weekly exposure. Risk a fixed, small slice of your roll each week — many winning players keep it to 10–20% — so a cold run can't end your season.
- Match contests to your edge. Cash games (50-50s, double-ups, head-to-heads) pay ~2× to the top ~half — lower variance, the place to grow a roll. Tournaments (GPPs) pay a tiny fraction enormous sums — high variance, where the leverage skills below pay off. Most players should run more cash than they think and treat GPPs as the smaller, swing portion.
You can be the best lineup builder alive and still go broke with bad game selection and sizing. Win the boring game first.
2. The building block: the stack
A stack is rostering players whose fantasy points rise together. In NFL the foundational stack is QB + his pass-catcher(s). When the QB throws a touchdown, someone caught it — so a QB and his WR1 double-count the same scoring play. That positive correlation is free leverage: one good game powers multiple roster spots.
The standard shapes:
- QB + 1 (single stack): QB + his WR1 or TE. The minimum correlation; fine for cash.
- QB + 2 (double stack): QB + two pass-catchers. More ceiling; the GPP staple.
- Onslaught stack: QB + 2 of his pass-catchers in a projected shootout. Maximum ceiling, higher variance.
3. The bring-back and the game stack
A stack alone bets that one offense explodes. But the biggest fantasy games are shootouts where both teams trade scores into the fourth quarter. To capture that, add a bring-back: a pass-catcher from the opposing team in your QB's game.
Why it works: if your QB is throwing all game, his team is probably either ahead in a track meet or chasing points — both scripts mean the other team is also throwing and scoring. A QB + 2 of his receivers + 1 opposing receiver is a game stack, and it's the single most reliable tournament-winning shape in NFL DFS. You're no longer betting on one offense; you're betting the game goes off.
The canonical GPP core
QB + WR1 (same team) + TE/WR2 (same team) + WR from the opposing team (bring-back) → fill RB and a value piece around it. That five-player game stack is the backbone of most large-field winners.
4. Correlation by position — what scores together
| Pairing | Correlation | Use it? |
|---|
| QB + own WR/TE | Strong positive | Yes — core of every stack |
| QB + opposing WR (bring-back) | Positive (shootout) | Yes — GPP staple |
| RB + own defense | Positive (game flow: lead → run → D plays with a lead) | Cash-friendly |
| QB + own RB | Slightly negative (pass vs run share) | Usually avoid stacking |
| QB + opposing defense | Strong negative | Never — they cancel |
| RB + opposing RB | Negative (one game script) | Avoid |
The two cardinal sins: QB against your own defense (your D wins when his QB fails — they fight each other) and two RBs in the same game (only one script can dominate). Build with the positives, avoid the negatives, and your roster stops working against itself.
5. Ownership & leverage
In a large GPP, projection isn't enough — thousands of opponents have the same projections. Your edge is being right in a way the field isn't. That means tracking projected ownership and deliberately leaning toward correct-but-underowned players.
- Pivot off chalk. When a popular value play is 40% owned, the cheaper or trickier option at the same position can be 8% owned with a similar ceiling. If your pivot hits, you leap the 40% who didn't have it.
- Leverage the stack, not just players. The most powerful leverage is an underowned game — a stack of a game the field has written off as low-total. If it shoots out, almost nobody has it.
- Don't punt leverage in cash. Leverage is a GPP tool. In cash, take the chalk if it's the best play — you're beating the median, and ownership doesn't matter when ~half the field cashes.
Leverage = your read ÷ the field's ownership. Hunt the games and players where that ratio is high — that's where tournaments are won.
6. Cash vs GPP builds
| | Cash (50-50 / DH) | GPP (tournament) |
|---|
| Goal | Beat the median (~top 50%) | Beat the field (top 1%) |
| Stack | Single stack OK; floor-first | Double stack + bring-back |
| Players | High-floor, high-volume, safe roles | High-ceiling, underowned, boom plays |
| Ownership | Ignore it — take the best plays | Lean underowned; pivot the chalk |
| Lineups | 1–3 | Many, differentiated |
7. NFL Showdown — a one-line reminder
Single-game NFL slates run on the Captain (1.5×) mechanic, and the rules flip slightly: ceiling over floor in the Captain slot, fade the obvious Captain in GPPs, and game-stack with a bring-back. The Captain decision is so important it has its own CourtEdge guide — but the short version: the winning Showdown Captain is almost never the chalk.
The pre-lock checklist
- Sized correctly? (≤ your weekly cap, right cash/GPP split.)
- Is there a QB stack? (At least QB + 1, ideally + 2 for GPP.)
- Is there a bring-back in your main game?
- Any negative correlation to remove? (QB vs own D, double-RB games.)
- For GPP — where's your leverage? (Underowned game or pivot.)
- Value plays in real roles, not just cheap?
- Did you check inactives / weather before lock?
The whole guide in three sentences
Win the boring game first — size your bankroll and pick the right contests, leaning more cash than you think. Then stop drafting disconnected studs and start building a bet on a game: a QB stack with a bring-back, free of negative correlation, that captures the shootout outcome. In tournaments, express that bet through underowned games and players so that when your read hits, you've left the field behind.