A DFS lineup optimizer sounds like a shortcut to winning. It is closer to a calculator. Given a list of players, their salaries and a projected score for each, it finds the combination that scores most on paper while fitting under the cap. It does that faster than any person could, and it never forgets a rule. What it cannot do is know whether the projections are right, or whether thousands of other people asked the same calculator the same question.
This page explains how a DFS lineup optimizer works, using a small example you can check by hand, then covers the rules that make optimizers useful in tournaments and what separates a free DFS lineup optimizer from a paid one.
How a DFS lineup optimizer works
Every optimizer solves the same problem. Pick one player for each roster slot so that:
- the total salary is at or under the cap,
- each slot gets an eligible player, and nobody appears twice,
- any extra rules you set are respected,
- and the total projected score is as high as possible.
With nine slots and a few hundred players on a Sunday slate, the number of valid lineups runs into the billions. Optimizers do not check them one at a time. Most frame the problem as an integer program, a type of mathematical model that lets a solver rule out huge numbers of combinations at once and still guarantee the best answer. That is why a good tool returns the top lineup in a second or two.
A worked example you can check
Shrink the problem to a three-slot lineup, one quarterback and two running backs, with a $20,000 cap and seven players to choose from.
| Player | Salary | Projection | Points per $1,000 |
|---|---|---|---|
| QB A | $7,000 | 22.0 | 3.14 |
| QB B | $6,200 | 19.5 | 3.15 |
| QB C | $5,400 | 17.0 | 3.15 |
| RB D | $8,200 | 20.5 | 2.50 |
| RB E | $7,000 | 17.5 | 2.50 |
| RB F | $5,600 | 14.0 | 2.50 |
| RB G | $4,500 | 12.0 | 2.67 |
There are 18 possible lineups; 14 fit under the cap. The best is QB A with RB D and RB G: $19,700 and 54.5 projected points. Second is QB B with RB D and RB F at exactly $20,000 and 54.0 points. Third is QB A with RB E and RB F, $19,600 and 53.5.
Two things stand out. The top three lineups are separated by one point, a gap smaller than a single reception on DraftKings. And the "best" lineup includes the running back with the best points per dollar, RB G, not the one with the highest projection. Optimizers do not pick the best players; they pick the best combination for the money. Change RB G's projection from 12.0 to 11.0, a perfectly ordinary disagreement between two projection sources, and the top lineup changes.
How this example was calculated
All 18 combinations of one quarterback and two running backs were enumerated and scored; the 14 under the $20,000 cap were ranked by total projection. Players, salaries and projections are invented for illustration and do not describe real athletes. The brute-force method gives the same answer an integer-programming solver would, which is the point of the example: with seven players you can verify it with a pencil. On a full slate the same logic runs over hundreds of players, which is where a solver becomes necessary.
Projections and salaries: how it works in practice
The calculation is the easy part; how it works in practice depends on what you feed it. Salaries come from the site and are fixed for the slate. Projections are where optimizers differ. A projection is an estimate of a player's average fantasy score for this game, built from role, matchup, recent usage, game total, weather and injury news. It is converted to each site's scoring, which matters: the same receiver projects higher on DraftKings than on FanDuel because receptions are worth twice as much there. The DraftKings vs FanDuel comparison covers the scoring tables.
Because the optimizer maximises projections, any error in them is amplified. If one source overrates a cheap running back by two points, the optimizer will put him in every lineup it can. Most experienced users review the projections before they review the output.
Rules that make an optimizer useful in tournaments
In cash games, the single top projected lineup is often a reasonable answer. In tournaments it is a problem, because many entrants use similar projections and arrive at similar lineups. Optimizers add controls to steer away from that.
- Stacks. Require a quarterback to be paired with one or two of his pass catchers, and optionally a player from the opposing team. The strategy guide explains why correlated players belong together in tournaments.
- Exposure limits. When building twenty or 150 lineups, cap the share any one player can appear in.
- Locks and excludes. Force a player in or keep him out, usually based on news or a strong opinion the projections do not reflect.
- Minimum uniqueness. Require each new lineup to differ from the previous ones by a set number of players.
- Randomness. Jitter projections slightly on each run so the output varies.
- Ownership rules. Some tools cap the combined projected ownership of a lineup, pushing it away from the most popular builds.
The worked example above shows why these rules matter. When the top lineups are a point apart, the difference between them is noise. Constraints let you choose among nearly equal options on purpose rather than by accident.
Free DFS lineup optimizer vs paid tools
Most DFS research sites offer a free DFS lineup optimizer, usually with limits: a few lineups at a time, a subset of sports, basic constraints only, or projections that update less often. Paid tiers lift those limits and add the things that cost money to produce.
| Feature | Free versions | Paid versions |
|---|---|---|
| Optimisation maths | Same | Same |
| Lineups per run | Often capped | Up to contest maximums |
| Projections | Basic or user-supplied | Proprietary, updated with news |
| Ownership estimates | Rare | Common |
| Stacking and exposure controls | Basic | Detailed |
| Late swap re-optimisation | Sometimes | Usually |
| Bulk upload to the site | Sometimes | Usually |
The honest summary: the solver is cheap, the data is not, and how it works is the same at every price. If you build your own projections, a free tool gives you nearly everything. If you rely on the tool's projections, you are paying for the projections, and you share them with every other subscriber.
Where optimizers go wrong
- Stale news. A player ruled out after the projections were last updated still looks like the best value.
- Wrong site scoring. Using a DraftKings projection for a FanDuel contest overrates every high-catch player.
- False precision. A 0.3-point edge between two lineups is not a reason to prefer one.
- Everyone has the same tool. In large tournaments the top projected lineup is rarely unique.
Knowing how a DFS lineup optimizer works is what lets you use one without being used by it. An optimizer is best treated as a fast, tireless assistant that checks your arithmetic. Decisions about which players you believe in, which games you expect to shoot out and how different you want to be from the field still belong to you. For how those decisions fit together, start with the strategy guide; for the format basics, the beginner guide.
Questions readers ask
What is a DFS lineup optimizer?
A program that takes player projections and salaries and returns the lineup, or set of lineups, with the highest projected total that fits the site's salary cap and roster rules, plus any constraints you add.
How does a DFS lineup optimizer work?
It treats lineup building as a constrained maximisation problem: maximise total projected points, subject to the salary cap, the number of players per position and your own rules. Most tools solve it with integer programming, which finds the best combination without checking every possible lineup one by one.
Is there a free DFS lineup optimizer?
Yes. Several DFS research sites offer free versions, usually with limits on the number of lineups, sports or custom rules. The optimisation itself is not the expensive part; the projections and ownership data that feed it are.
Are paid optimizers better than free ones?
Paid tools usually differ in data rather than maths: their own projections, ownership estimates, late news updates, and controls for stacks, exposure and randomness. If you supply your own projections, a free optimizer can produce the same lineup a paid one would.
Do optimizers win DFS tournaments?
They build lineups; they do not guarantee results. A lineup that is optimal on public projections is often popular, which limits its value in large tournaments. Players who use optimizers well add rules that steer the tool away from the most common builds.
What is exposure in a lineup optimizer?
Exposure is the share of your lineups that include a given player. When you build many lineups, an exposure cap such as 30% stops the optimizer from putting its favourite player into every single one.
What is randomness in a DFS optimizer?
A setting that nudges projections up or down by a random amount on each run, so the optimizer produces varied lineups instead of the same top lineup again and again. It is a crude way to reflect that projections are uncertain.
Can an optimizer handle late swap?
Many can re-optimise only the unlocked slots after the early games start, keeping the locked players fixed. Whether that is available depends on the tool and on whether the contest allows late swap.
Is using a lineup optimizer allowed on DraftKings and FanDuel?
Building lineups with outside tools is common and generally permitted; both sites restrict automated scripts that interact with the site itself and set limits on entries. Check the operator's terms for its current rules on third-party tools.






