How To Use NBA DFS Sims To Win On DraftKings & FanDuel

If you play NBA DFS, the hard part isn't naming the good players. It's deciding how much of each one to play across a whole pool of lineups, then knowing which of those lineups actually gives you the best shot to win money. That's the exact problem the NBA DFS Sims are built to solve, and it's how I attack large-field basketball tournaments on DraftKings 🎁 and FanDuel 🎁.

This is a hands-on walkthrough of how I use the Stokastic NBA Sims: building a lineup pool in the Contest Sims, running tens of thousands of simulated contests, and then trimming and shaping my player exposures with leverage, ROI boosts, and hard caps. If you haven't played since last NBA season, or you're brand new and have no clue what you're looking at, this is the orientation. Pull up today's board in the NBA DataHub and follow along.

In Summary

• Build A Pool That Mimics The Field First, Then Differentiate. In the Contest Sims you set the site, slate, pool size, and contest stakes, generate your lineups, then simulate them against each other.
• Let Simulated ROI Rank Your Lineups, not a single projected score. The Sims pay out a real GPP structure (set by "percentage to first") and sort every lineup by Sim ROI.
• Shape Exposures With Leverage, ROI Boosts, And Hard Caps on a separate page, not on the build page. Want exactly 50% of a player? You can set that.
• Process Over Results. The Sims tilt the odds in your favor over a season. They don't promise any single night.

Watch: How To Use The Stokastic NBA DFS Sims

The full walkthrough below builds a pool, runs the contest simulation, and shapes exposures step by step. Read on for the written version with the exact settings and a worked leverage example.

Watch on YouTube

Step 1: Build Your Pool In The Contest Sims (Match The Field, Don't Fight It)

Everything starts in the Contest Sims, where you build out the pool of lineups that ultimately gets simulated. Generating that pool takes 30 to 60 seconds, but four settings matter before you press the button:

A couple of notes on those rows. The Max package builds up to 5,000 lineups and Core builds up to 2,000, so always set the pool size as close to your contest's field as you can: a 1,000-entry single-entry contest wants a 1,000-lineup pool, while the marquee flagship GPPs want the full 5,000 (Max) or 2,000 (Core). The contest archetype matters because higher-stakes contests tend to be sharper, so the best projected players draw more ownership up top and less ownership in soft, low-stakes fields. Marquee is the setting you'll use most: think the $51 150-max on DraftKings or the $91 150-max on FanDuel.

Here's the mindset that trips up most players: this stage is about emulating the field, not building the lineups you personally want. You're modeling what everyone else will do so the simulation has a realistic opponent pool. You shape your exposures later, on a different page.

When you generate, you'll see projected ownership, your pool exposure, and the difference between them, usually within a couple percentage points. There's also a boost ownership control here, and it has exactly one use: when you think the model is wrong on a player's ownership. If you expect Luka Dončić to run hotter than his projected 50%, boost his ownership and regenerate so the simulated field reflects it. What you should not do here is boost a player because you personally want more of him. That's an exposure decision, and it lives on a different page (Step 3).

Step 2: Run The Contest Simulation (Let Sim ROI Do The Ranking)

Once your pool is built, head to the projections page. The first setting is percentage to first (first-place prize divided by the total prize pool), and you match it to your contest's payout shape. A standard DraftKings example: a $400,000 prize pool with $100,000 to first is 25% to first (100K ÷ 400K). When your contest's exact percentage isn't listed, pick the closest one. This matters because a top-heavy contest rewards a very different lineup than a flat one.

Below that, the Stokastic projections are already loaded: minutes, fantasy points per minute, projected fantasy points, points-per-dollar value, ownership, plus standard deviation and variance. There are three ways to run it:

• Use The Stokastic Projections As-Is. This is what most people do, and it's totally fine: just press run.
• Upload Your Own Projections via CSV, with three columns (player name, projection, ownership). Every column on the page then updates to your numbers.
• Hand-Edit A Player. Type a new projected total for one guy, or use the plus/minus arrows to nudge by a point, and the rewind arrow to undo.

One advanced touch on variance: during late swap, once a player's score is locked (his early game finished), you can enter his exact fantasy total and set his variance to zero, and every simulation will then treat that score as known instead of a range. Raising a player's variance widens his range of simulated outcomes; lowering it tightens them.

Press run contest simulation and here's what actually happens: the slate is simulated tens of thousands of times, and all the lineups you built play against each other inside that 25%-to-first payout structure across every one of those runs. Then they're sorted by simulated ROI. That's the whole point of this approach. We optimize for win probability across thousands of simulated slates, not for one projected total. In one run, the top lineup posted a 190.5% simulated ROI, and that ranking, not a single projected score, is how you should be sorting and selecting.

You'll also see each lineup's projected points, total ownership, how often it won a GPP, hit the top 10, and min-cashed, plus dupes. Dupes usually don't move the needle much on a big main slate, but on a two-game slate or a showdown they matter a lot: a lineup that wins but gets duped multiple times returns less than the same lineup coming in clean, and the simulation folds that straight into Sim ROI.

The key shift: you're not chasing the single highest-projected lineup. You're ranking thousands of builds by how they actually finish against a simulated field, with dupes and payout shape baked in. That ranking is the edge.

Step 3: A Worked Example, Trimming And Shaping Your Exposures

This is where the real edge lives. First, quick-favorite your lineups: there are one-click buttons for the top 20, 50, or 100, or you can type a custom number to match exactly how many you're entering (say 70). Want your builds a little more diverse? Set a number of uniques: at two uniques, each favorited lineup must differ from the last by at least two players, which is handy on showdown or two- and three-game slates where the player pool is thin.

Now open the exposures tab, where you can sort all players or filter by position. The column that matters is leverage = your exposure minus projected ownership. A real example from the walkthrough:

• Isaiah Hartenstein was projected for 45.6% ownership, and he was in 80% of the favorited lineups. That's +34.4% leverage (80% − 45.6%), the kind of overweight position on a productive player that wins GPPs when it hits.

Say you decide 80% is too much Hartenstein. You change exposure here, not on the build page, and there are three ways to do it:

1. ROI boost (loose). Apply a negative ROI boost (say −20%), re-favorite your top lineups, and his exposure slides down (in the walkthrough, from 80% to about 70%). Good when you don't have an exact target and just want "less of him." Delete the boost any time to snap back.
2. Exposure cap (exact maximum). Set Hartenstein to a hard 50% cap, hit set exposures and adjust favorites, and you land at exactly 50%. You can also set a global max exposure so no single player exceeds, say, 50%.
3. Exact minimum. To force a floor on an under-owned guy, combine the two. Brandon Podziemski sat at just 7% of lineups; to lock him at exactly 20%, boost his ROI aggressively (up to 100%) so he's over-rostered, then cap him at 20%. The boost pulls him up, the cap trims him to your number.

Leverage and ownership are the whole game at this stage. You're not just finding highly projected players. You're finding good leverage spots, where the field piles onto one or two guys and leaves similar-projection players under-owned. When you're set, head to the entry manager, upload your DraftKings or FanDuel entry CSV, and assign your favorited lineups (there's a separate walkthrough on the site for that step).

New to Stokastic? The NBA Sims build a full lineup pool, simulate it against a real tournament field tens of thousands of times, and rank every lineup by simulated ROI, with the leverage, ROI-boost, and exposure-cap controls shown above. Try it free for 7 days, and code NBASIMS10 takes 10% off your first month of the Stokastic NBA package.

start your free trial

This Is A Tournament (GPP) Workflow, Not A Cash-Game One

One important scope note: everything above (building a large pool, simulating it against a tournament field, setting "percentage to first," and leveraging ownership) is the GPP (tournament) process. The Contest Sims are built around tournament dynamics: a top-heavy payout, beating thousands of opponents, and getting leverage off the chalk.

Cash games are a different game. In a double-up or 50/50 you only need to beat roughly half the field, so the goal is the highest-floor lineup built straight off projections, not a leveraged, contrarian, simulated pool. Don't run your GPP pool into cash; build cash lineups around the projections and floor you'll find in the NBA DataHub instead.

Why This Beats Picking "The Best Players"

Picking the right players is only half of NBA DFS. The other half is roster construction and contest selection, and that's where the simulation earns its keep. Instead of one projected score, it shows you how often each lineup wins, hits the top 10, min-cashes, and gets duped, the full range of outcomes rather than a single number that hides the variance. Honestly, even if all you do is build a pool, run the contest sim, and make no manual adjustments at all, that process usually beats hand-building lineups by feel. The adjustments above are fully in your control, but they're a bonus, not a requirement. No process makes any single NBA slate predictable, which is exactly why the goal is win probability over a large sample, not a promise on any one night.

Try It On Your Next Slate

Want to run this on a real slate? Everything starts in the Stokastic NBA DataHub, with the night's projections, ownership, and slates in one place. Sample the workflow with our free DFS sims first. When you're ready for the full Contest Sims, ownership leverage, and exposure controls covered above, the Stokastic NBA package unlocks all of it. New users get a free 7-day trial, and code NBASIMS10 takes 10% off your first month: start your free trial

Prefer to tail a proven card instead of building your own? The value-first habit above still applies; you just outsource the legwork. DFSnDonuts Picks on Tails is a DFS-native expert (#8 on the MLB Hot Streak leaderboard as of July 16) whose plays come with the receipt every Tails expert carries: a full graded track record you can check before you tail. Start with the free expert picks feed, where free picks post every day.

Frequently Asked Questions

What are NBA DFS Sims? They're a DFS contest simulator that builds a large pool of lineups, then runs them against each other through tens of thousands of simulated contests (paying out a real GPP structure) and ranks every lineup by simulated ROI instead of a single projected score. Think of it as NBA DFS lineup-building that optimizes for win probability, not a single projection.

How many lineups should I build? Match your pool size to the contest field. For a 20–30,000-entry large-field GPP, the 5,000-lineup pool (Max package) is closest; Core builds up to 2,000. For a 1,000-entry contest, pick the size nearest that.

Do I have to use Stokastic's projections? No. You can run the Sims on Stokastic projections (most players do), upload your own via CSV with player name, projection, and ownership columns, or hand-edit individual players and rewind changes.

What's the difference between leverage and ownership? Projected ownership is how much of the field is expected to roster a player; leverage is your exposure minus that ownership. Positive leverage means you're overweight a low-owned player, the upside that wins tournaments.

Can I use the Sims for showdown? Yes. Single-game showdown contests are fully supported. Watch dupes more closely on showdown and two-game slates: the player pool is small, so identical lineups spike, and the Sims fold that dupe risk into Sim ROI. Setting a couple of uniques helps spread your builds out.

How do I set an exact exposure to a player? Use the exposures tab, not the build page. For an exact maximum, set an exposure cap (e.g., 50%) and adjust favorites. For an exact minimum, boost the player's ROI aggressively so he's over-rostered, then cap him at the number you want.

Will the Sims win me every contest? No. Nothing in DFS does, and NBA has real night-to-night variance. The Sims are designed to improve your win probability over a large sample, not to promise any single result.