The optimizer isn't wrong. It's answering a different question.
A lineup optimizer and a contest simulator solve different problems. Here's the difference in plain terms — plus what happened when we entered an optimizer's #1 lineup into a simulated contest.
The lineup optimizer
An optimizer asks:
It works with:
- Player projections
- Salaries
- Roster rules
- Exposure limits
- Stacking rules you set by hand
- Sometimes a "randomness" slider or ownership caps
Its core job is arranging players efficiently under the salary cap — and it's genuinely good at that. It's the right tool for cash games, and nothing beats it at building 150 legal lineups with your rules enforced.
The limitation: it treats a lineup as a static list of projections. Even across 150 lineups, it does not model:
- The full range of each player's possible games — including the ceiling games that win tournaments
- Correlation — the way a QB's big day is also his WR's big day
- How the rest of the tournament field will build
- How often your lineup will be duplicated, and the prize-splitting that comes with it
- The contest's actual payout structure
- Whether a lineup is profitable across thousands of possible Sundays
Adding randomness to projections creates lineup variety — but variety is not the same as modeling real NFL outcomes or contest economics. Your lineups end up different from each other, not from the field's.
The simulator
SimSlate asks:
To answer it, the simulator:
- Builds a full field of opposing lineups at realistic ownership
- Gives every player a realistic range of outcomes — not a single number, and not a bell curve
- Lets teammates spike together, so stacks pay off where they should
- Plays the slate thousands of times and ranks every lineup in every simulated contest
- Pays those ranks through the contest's real payout structure
- Splits prizes between duplicate lineups, like DraftKings does
- Reports each lineup's simulated ROI, win odds, and expected duplicates
The proof
We tested this on a real slate: we built the highest-projected legal lineup — exactly what an optimizer would hand you — and entered it into a simulated 10,000-entry contest.
The optimizer's lineup out-projected every one of the 10,000 lineups in the field — and still wasn't the best bet. About 100 field lineups earned a better simulated ROI, and the board's top build — a QB stack projecting 12 points fewer — earned nearly 2.5× as much. Of the 100 best-ROI lineups, 97 were QB stacks. The optimizer's wasn't.
The simplest distinction
An optimizer tells you which lineups you can build efficiently. A simulator tells you which lineups are actually good tournament investments.
An optimizer loves a lineup because it projects for the most points. The simulator may show that same lineup is over-owned, heavily duplicated, and poorly correlated — and find a slightly lower-projected lineup whose ceiling, stacks, low ownership, and duplication profile make it the better bet.
If you like hand-building, the Lineup Builder does both: build any lineup, then simulate it against a real field on the spot. The free plan runs a real 1,000-lineup field simulation, no card required; the full model lives on the methodology page.
What is a DFS lineup optimizer?
A tool that answers one question: what's the highest-projected legal lineup I can build? It works with projections, salaries, and roster rules, and arranges players efficiently under the cap.
Are lineup optimizers good for GPPs?
They're great at building legal lineups fast. But they don't model the tournament — outcome ranges, correlation, the field, duplicates, or payouts — and tournaments pay your rank against a crowd.
Why do optimizer lineups get massively duplicated?
Same projections in, same lineup out — and most of the field uses similar projections. Randomness sliders make your lineups differ from each other, not from everyone else's.
What does a contest simulator do that an optimizer can't?
It plays out the tournament: realistic outcome ranges, teammates spiking together, a full field of opponents, duplicates splitting prizes, and the real payout structure — then ranks lineups by simulated ROI.
Your optimizer found the biggest number. Now price it.
Run a real 1,000-lineup field simulation on the free plan — no card required.