One sleek builder for every slate โ MLB, NFL, CFB, NBA, NHL, WNBA, LoL, PGA, NASCAR & MMA. Load a slate, pick your contest type, build, and export a DraftKings upload file. Premium members: same email as /picks.
๐ฅ
Load your slate
Drop a projections CSV here (columns: Name, Position, Team, Salary, Projected, Ceiling, Floor, ProjOwn, DFS_ID). The sport is detected automatically.
Checking for today's published slateโฆ
1 What are you playing?
Loading tonightโs DraftKings contestsโฆ
Not in the list? Enter it by hand โ field size, fee, and the payout ladder off the DK contest page.
Pick a contest above, or enter one below.
Enter a contest by hand, or paste the exact payouts
Fill these in off the DraftKings contest page and the simulation is calibrated to that exact
contest instead of a generic tournament shape. The payout ladder is the one that matters most โ
with it every prize is looked up by finishing rank rather than modeled, and modeled ROI uses those posted payouts.
Field size and entry fee are the minimum. Add 1st prize and cash % and the curve is modeled;
paste the ladder below and it is exact.
Tournament default: 90%. The other 10% may use RB, WR or TE when that build grades better.
Off by default. Projected ownership still informs simulations and rankings.
under%
above%
How to build a set that can win a GPP โ 6 steps
Pick your contest first. Choose it from the DK list, or hit โ๏ธ Type in a contest myself and fill in field size, entry fee, and the payout ladder off the contest's Payouts tab. Everything downstream โ ROI, win%, which lineups get selected โ is calibrated to that contest. Skip this and the numbers are a generic shape.
Load tonight's core. Hit โญ Add core plays. Three players lock in from the playbook โ the anchor bat, the leverage bat and the arm. Three locks on a ten-man roster is deliberate: it leaves the optimizer room to work.
Set the stack shapes. In Fine-tune, hit GPP mix, or type your own. The numbers are the percent of your lineups using each shape. 5-3 means five hitters from one team plus three from another. Large-field MLB GPPs are won by five-man stacks โ if that row is blank you are leaving the shape to chance.
Anchor a team. In the team row, put a number next to the team from tonight's playbook โ again a percent of your lineups, not a hitter count. 40 on Cleveland means 40% of your set is built around Cleveland.
Build, then read the exposure table. It is sorted by position. Anything sitting far above the field with a big green Edge is where your leverage is; anything far above the field that is also chalk is where your risk is. Type a Cap % on any player to pull him back, or a Min % to require him in at least that share of the set.
Sanity-check the ROI spread. A healthy 150-lineup set has lineups on both sides of zero. If every lineup shows a positive ROI, something is wrong and you should tell me โ that is not a set, that is a bug.
Two traps. Locking two hitters from the same team you then ask for a five-man stack of shrinks the legal space hard โ a nine-bat lineup only has so many distinct five-man combinations, and you will get fewer five-stacks, not more. And Min unique above 2 on a 150-max will cut your set short; leave it at 1โ2 unless you know why you are raising it.
Every percentage is editable. Shapes account for all eight hitters; pitchers never count. The requested construction is enforced during generation.
5-1-1-1
5-2-1
4-2-1-1
4-3-1
4-2-2
4-4
5-3
Total must equal 100%.
This row sets which team, not how many hitters. CLE 40 = 40% of your lineups are anchored on Cleveland. It does not mean a 4-man or 5-man stack โ that comes from the shape row above. Set both: CLE 40 + 5-3 25 gives you Cleveland five-stacks.
Measured on the settled $150K Bat Flip, 9,803 entries: 64% of the top 100 rostered a pitcher from a team they also stacked 3+, against 11% at the field median. The stack shapes were identical top and bottom — the difference was the arm. It is high variance, not free money: the bottom 1,000 ran it at 21.8%, because when the game goes the other way the arm and the bats die together.
Locks force a player into every lineup and fades remove him from all of them. These sit in between, which is where most real rules live: at least 2 of these four, at most 1 of these three, or if I have Thomas then I need Copper or Bonner with him. Every rule is checked while lineups are generated, so a lineup that breaks one is never built.
Research exports can exceed the contest entry limit.
Players
0
Teams
0
Salary cap
$50,000
Built
โ
Player pool click Lock/Fade to force or remove a player