Circa Grandissimo 2026 · $100,000 entry · · · exact per-entry inventories · field behaviour fitted to this contest's own picks
Distinct over the legs that still matter (you and at least one opponent alive with ≥2% probability). Later legs are placeholders you will re-plan.
Run the analysis to see this week's decision table, the highest-EV paths and the projected field.
Everything the simulator consumes is editable here. Blank cells use the embedded PoolGenius (and SurvivorAtlas, when loaded) grids and the fitted field model; filled cells override them and are highlighted. Edits are kept in this browser and can be exported as a file.
Per game: the two providers, their blend at the current slider, and your number (win % of the first-named team). A decided game is fixed and locked. Effective is what the number in force is worth that far out with the future-week discount on: each simulated season draws the true line around it (team strength drifts ≈0.75 pts/√week, plus ≈2.4 pts of game noise — calibrated on ~1,300 look-ahead vs closing lines), which regresses far-out favourites toward 50% and makes a team's future games move together. Hand-edited games get the same treatment (your week-16 number is a forecast too).
Two modes (“Your edits”, top bar). Default — this week locked, later weeks reference: as below. Lock all my edits: every leg you edited is played exactly as entered — your shares become whole opponents among whoever is alive in each simulated season, no regression — and every win % you typed is treated as the truth (no future-week discount on that game).
Current leg — locked: your read is the field, exactly: shares are converted to whole opponents (largest remainder) and every simulated world has exactly those picks, assigned only to opponents for whom the team is legal. Later legs — reference: the field plays only the teams you list; your percentages are targets in expectation, and with the future-week discount on, a read h weeks out keeps weight 1/(1+(h/half-life)²) and the rest follows the model's own split among your listed teams (never other teams). An opponent who has already used every team you listed takes its best legal favourite (counted as a forced pick). Legs you leave blank follow the fitted field model under the pick floor — no slop. “Model (rest)” deliberately hands a share back to the fitted model.
One leg per line, 4: MIN 50, BAL 34, SEA 12, BUF 4; add MODEL 10 to send 10% to the fitted model. Applying replaces those legs' entries.
One field model is one opinion. The ensemble values this week's candidate paths against many sampled fields — each a joint draw of forecast error, provider blend, fitted-behaviour parameters (with their estimation covariance), field concentration, rationality λ and jitter around your ownership numbers — and reports, per first pick, the mean EV, its spread, how often it is the best pick and its mean regret. Runs in parallel across — background threads.
Run the analysis to see the fitted field model.
Tick “Alt-model sweep” and run to see how the week's ranking moves under alternative fitted field specifications and a ±25% concentration shock. The Field ensemble tab generalises this to many sampled fields.
One entry per line: ENTRY | Team, Team, Team. Remove eliminated entries; add each week's picks as they publish; then set “Current leg” above. Forecasts are the dated PoolGenius and provider grids embedded in this page (see provenance); refresh them by regenerating the page from new captures.