What is a DFS cash game?

A DFS cash game is any contest with a flat payout structure that pays the top ~44% of entrants — most commonly 50/50s (top half doubles their money at ~1.8x buy-in after rake) and double-ups (top half doubles). Contrast with tournaments/GPPs where 20% of entrants cash but payouts skew heavily toward the top 1%. Cash games reward floor; tournaments reward ceiling.

How do you build a DFS cash-game lineup?

Optimize on floor projections, not mean or ceiling. Every rostered player should clear ~2.5 fantasy points per $1,000 of salary in the projected floor (10th percentile) — that's the value threshold cash-game math demands. Skip GPP-style leverage picks; take the highest-projected safe player at each position and let the correlation naturally follow.

Should I stack in DFS cash games?

Generally no. Stacking amplifies both the ceiling AND the floor variance — you win harder when the stack hits and lose harder when it busts. Cash-game payouts are flat, so the floor variance costs you money the ceiling can't compensate for. The exception is naked-QB stacks (QB + one bring-back position) in NFL, which slightly raise floor without the full variance cost.

Why does contest selection matter more for cash games than GPPs?

Because cash-game ROI is bounded — you can't win 100x on a 50/50. The only way to compound cash income is volume × edge, and rake percentage per contest eats directly into your edge. A 5% rake versus 10% rake means the difference between long-term profitability and long-term breakeven at the same skill level.

How do cash and GPP lineups differ on the same slate?

They should share maybe 3-4 players out of 8-9. Cash builds emphasize highest-floor value plays; GPP builds emphasize highest-ceiling leverage plays. If your cash and GPP builds are 80%+ identical, one of the two is misconstructed — probably the GPP build (too safe to differentiate) or the cash build (too much stacking).

DFS Cash Game Strategy: How to Beat 50/50s and Double-Ups Consistently

By DFS Simulator TeamPublished July 29, 20269 min read

DFS cash games are the income surface — flat payouts, top ~44% cash, ROI stabilizes fast enough that skill compounds weekly instead of quarterly. The construction discipline is completely different from GPP construction: floor matters, ceiling doesn't, stacking mostly hurts, and rake percentage per contest can be worth more than a projection edge.

Most DFS content covers GPP construction because tournament payouts are what generate the eye-catching screenshots. But for the median winning DFS player, cash games are where income actually compounds. This piece walks through the construction rules that separate winning cash-game portfolios from breakeven ones — and why the same lineup that wins a GPP is often a losing cash-game entry.

The math of flat payouts

In a 50/50 double-up contest, entrants pay a fixed buy-in and the top half of the field receives (roughly) 1.8x their buy-in after rake. Every entrant in the top half receives the same payout — the 1st-place finisher and the 44th- percentile finisher both cash for 1.8x. No prize ladder, no top-heavy payout curve.

The consequence: a lineup that finishes 1st and a lineup that finishes 44th earn the same money. Ceiling adds zero value in cash games; only "did I clear the cash line?" matters. Every construction decision should optimize the probability of clearing the line, not the size of the outcome above it.

Floor optimization

The concrete construction rule: every player on your cash- game roster should clear ~2.5 fantasy points per $1,000 of salary in the projected FLOOR (10th percentile of simulated outcomes). That's the value bar the flat-payout math demands. Above the floor, sort by mean projection — take the safest projected value at each position and let correlation happen naturally rather than engineering it.

Correlated Monte Carlo simulation makes this workflow possible without hand-computation. See our simulation-explained piece for how the floor column is derived; the DFS Simulator surfaces the floor projection alongside mean and ceiling so cash and GPP construction can share the same simulation run with different sort keys.

Why stacking hurts cash

Stacking amplifies both tails of the distribution: when the stack hits, you win big; when it busts, you lose big. In GPPs, the tail asymmetry (top 1% wins massive multiples) makes stacking profitable because you're paying for upside. In cash games with flat payouts, tail asymmetry gives you nothing — winning by 20 points pays the same as winning by 2. What matters is the probability of clearing the line at all, and stacking widens the distribution in both directions, which HURTS that probability.

The exception is a naked-QB stack in NFL (QB + one bring-back receiver from a different game). This slightly raises the QB slot's floor without amplifying variance across multiple positions. Not a stacking play — a game-flow hedge.

Contest selection: rake is edge

Cash-game ROI is bounded above by the payout multiplier (usually 1.8x) and eaten from below by rake. A 5%-rake contest requires ~52.5% win rate to break even; a 10%-rake contest requires ~55.5%. That 3-percentage-point difference is enormous — it exceeds the edge most sharp cash players actually have over the median field.

Sharp cash players filter their contest universe by rake first, contest size second, and only then by projected edge. Multi-entry cash contests (where you can enter 3 lineups into the same 50/50) rake heavier than single-entry; small-field cash (10-entry double-ups) rake less than large-field. See our contest-selection piece for the full framework.

Volume compounds edge

A player with a real 55% cash-game win rate on 20 contests per week over a 40-week season plays 800 contests. At $10 buy-ins and 1.8x payouts, the expected profit is approximately $800 (55% × 800 × $10 × 0.8 payout minus 45% × 800 × $10 loss). That's meaningful, boring, replicable income.

The same 55% player entering 3 contests per week for the same season nets ~$120 — the edge is identical but the volume dilutes to noise. Cash-game bankrolls are built by entering EVERY slate you have edge on, not by picking your best spots. GPP-style contest selection (only enter the highest-EV nights) inverts the correct approach for cash.

Building parallel cash + GPP portfolios

Run one simulation of the slate with your baseline projections and stacking rules disabled.
For cash: sort by floor. Pick the 6-9 highest-floor value plays (2.5+ pts/$1k floor) that fit the salary cap. Enter across your cash-game buy-in tier at high volume.
Re-run with stacking rules enabled. Sort by ceiling. Pick the top 5-20 diverse builds. Enter in your GPP contests at lower volume.
Cash lineups and GPP lineups should share maybe 3 players. If they share 7+, one of the two is misconstructed.

DFS bankroll management — cash-game volume math connects directly to bankroll sizing
DFS contest selection — which cash contests to concentrate volume in
How many DFS lineups to enter — the GPP counterpart discipline
Stacking strategy — the correlation math that DOES help GPPs
All DFS sports