Look, using an NFL model and lineup optimizer can sometimes be overwhelming. I am doing my best to make it easy on you by adding things like my custom templates with rules and making in-depth tutorials.
I’ll walk you through how to use FantasyLabs’ NFL DFS optimizer to make 1 or 300 lineups for your NFL DFS lineups.
What the FantasyLabs Optimizer Does
Our optimizer is a tool designed to generate anywhere from one to hundreds of optimal lineups in a matter of seconds. Instead of a user manually tinkering with combinations to fit players under a strict salary cap, the optimizer mathematically calculates the highest-scoring player combinations.
However, the optimizer is strictly a tool. Not an “auto-win” button. Its main goal is to turn your personal research and read on a slate of games into actionable lineups. Without human decision-making and custom rules, an optimizer would simply generate the exact same default lineups for everyone using it. It provides the speed, flexibility, and reliability to build multi-entry combinations while adhering to the constraints you set.
Be sure to check out our Player Models, along with the suite of tools on FantasyLabs, such as the Lineup Optimizer, Lineup Builder, Trends tool, and Correlation Dashboard, to prepare for the abundance of NFL slates each week.
Oh yeah! And our SimLabs Lineup Generator gives you the ability to create advanced DFS lineups using the power of simulation.
You can check out our in-depth SimLabs guide here.
What Inputs Our Optimizer Uses
Our optimizer takes a massive amount of baseline data and runs it through the custom rules you dictate. The most important inputs include:
- Player Projections: The backbone of the optimizer. This includes baseline NFL point projections from award-winning NFL expert Sean Koerner, along with ceiling/floor projections, or custom backtested projections uploaded by the user. You can also create aggregate projections from various sources.
- Salary Constraints: Settings that dictate how much of the salary cap must be spent. For example, a common input is setting a minimum salary cap limit of 99%, because leaving too much money on the table statistically lowers your win equity.
- Correlation Rules: Inputs that force or restrict specific player pairings to capture game scripts. A standard input rule is telling the optimizer to always stack a quarterback with at least one (or two) of his pass-catchers, plus at least one player from the opposing team.
- Player Groups and Exposures: Inputs that manage your risk and cap how often players appear. You might set a rule like “at most two of Joe Burrow, Ja’Marr Chase, and Tee Higgins” to avoid over-investing in one offense or limit a volatile player’s total exposure to 15% across all generated lineups.
- Ownership Projections: Estimates of how popular a player will be. You can use these inputs to force the optimizer to include low-owned players in every lineup to ensure differentiation.
What a Good vs. Poor Lineup Looks Like
A Good Lineup:
- Highly Correlated: A winning tournament lineup heavily relies on positive correlation. If your quarterback has an incredible game, his top wide receiver likely does too. A good lineup capitalizes on these linked outcomes rather than just picking random players in a vacuum.
- Properly Differentiated: It perfectly mixes popular “chalk” plays with contrarian leverage picks.
- Maximizes Expected Value: It spends the salary cap efficiently and targets players who are great ceiling plays and values who have the potential to completely break a slate open.
A Poor Lineup:
- Blindly Generated: A lineup created by simply clicking “Optimize” with no custom rules applied. Because thousands of users have access to the same default projections, blindly optimized lineups are highly duplicated. Even if that lineup wins, the prize pool gets split among many different users.
- Contains Negative Correlations: A poor lineup ignores real-world game logic. An example includes playing a quarterback against his opposing defense.
- Overly “Chalky”: A lineup built exclusively with the most popular players on the slate. If even one heavily-owned player busts, the lineup is likely dead in tournaments. If it does succeed, it lacks the uniqueness required to actually pass other players at the top of the leaderboard.
