Cascade Lab

Game scores0 of 0 set

Type a final score for either side (away @ home). An untouched side stays at its Vegas implied number. Pins beat scores; percentiles come last — same hierarchy as OWS.

How the cascade works

Every game has a pool of simulated outcomes (default 10,000) where each player's points are correlated through team points, game script and target share. When you pin a player, set a score or pull a percentile, the board re-weights those sims toward the ones that match what you asserted (pins keep an effective sample of ~500 sims, scores ~700, percentiles ~250) and re-reads every player's median (Proj), 80th percentile (Ceiling) and 90th percentile. Nothing is "adjusted" by hand — the whole game moves together.

set by youre-cascaded by your inputsavailable inputΔ = change vs base projection

Build settings

Lineups Min unique Min salary Max salary Ceiling blend % Over-max exposure Restarts / lineup Projection source

Randomness (per lineup draw)

Team stacks

Team stack % Game stack % (of stacks) Force 2 flex % (of stacks) Forced no-stack %

Allow DST vs your QBAllow DST vs flex players

Ownership

Total own min %Total own max %

Players under N% owned (min – max per lineup)

Max exposure by position

Player groups

A group says: on Pct% of lineups that contain the 🔒 locked players (or all lineups if none are locked), use between Min and Max players from the group.

Player pool

hide 0 proj

Lineups

Exposure

Bucket mix

Paste lineups to sim (optional)

Contest

Entries Entry fee $ Rake % Places paid % Top-heaviness Min cash ×fee

World

Sims Field lineups Build to sim

What the numbers mean

Each lineup is scored in every slate sim and ranked against a field of lineups generated from projected ownership (with realistic stacking and salary use). The contest's payout curve turns rank into dollars. ROI is expected return per entry; Cash the share of sims it finishes in the money; Top 0.1% and Win use a tail model so large fields aren't limited by the sample. Numbers are relative, not gospel — use them to compare rosters.

Lineups by ROI

Score distribution

This week's slate

One click pulls everything from public feeds — no One Week Season involved: DraftKings (players, salaries, injury status, matchups), ESPN (totals & spreads), Sleeper (stat-line projections & injury news). Player IDs are DraftKings' own, so your pins, stars and groups survive a refresh.

Week Slate

Manual uploads (fallback if a feed is down)

DraftKings salary CSV
DKSalaries.csv from any contest Projections / ownership CSV (any source)
name + projection; optional team, pos, ceiling, ownership, salary

Estimate ownership when no source is loaded

Projection columns understood: name / player · team · pos · proj / projection / fpts / points / dk · ceil / ceiling · own / ownership / proj own · salary. Bink-style exports (Id,Name,Position,Team,ProjPts,ProjOwn) work as-is.

Poker Grinders roster sync

Sleeper league IDManager

Vegas lines

Home line as the book writes it (−3.5 = home favored by 3.5). Pulled from ESPN; edit and rebuild if you disagree.

Game

Total

Home line

Implied

Kickoff

Model knobs

Sims per game Seed Shared passing variance Target competition (κ) Game pace variance Tail shrink above ceiling Total σ (pts) Margin σ (pts)

Higher shared passing variance ties a QB more tightly to his receivers (OWS behaves like ~0.55–0.6 here). Lower κ means receivers compete harder for targets. Rebuild after changing.

Poker Grinders scoring

Pulled from your Sleeper league (Superflex, half-PPR, TE +0.5). When the slate comes from the weekly pull, league points are scored exactly from Sleeper's stat lines; the built-in Week 1 sample converts DK points through a positional template instead.

Save & share

Everything autosaves in this browser. Export a snapshot to move it between devices or keep a week's work.

Import snapshot

Player pool

Lineups

Lineups by ROI

This week's slate

Poker Grinders roster sync

Vegas lines

Model knobs

Poker Grinders scoring

Save & share

Working…