Complete with your budget
—
Simulated total-goal completion
Average orbs needed
—
Estimated orbs to all goals
90% completion cost
—
90th percentile · simulated
Your orbs
100
2 per reroll right now
🎯 Skill wishlist
Tap a skill to add it; choose only its minimum rank.
Templates save locally on this device.
🧚 Current fairy
Edit slots or simulate a roll
Changes automatically recovered in this browser.
Add skill
Choose a skill and slide its rank. Use + on the next empty slot to enter the newly learned skill. Locks protect a slot from simulated rerolls.
🎲 Quick fairy simulator
Simulates one random outcome and applies it to the editable fairy. Undo reverses it.
Ready · results use the current conditional skill pool.
✎ Record an actual in-game reroll
Record actual reroll
2 orbs · exact result odds
🗺️ Your remaining quest
Each desired skill updates live as you edit the fairy.
🔮 Complete-fairy forecast
From your current skills to every selected minimum rank.
Simulation plan
Running simulations…
Completion probability by total orbs spentGold line = your budget
Median cost · 50% finish by
—
75% finish by
—
⚒️ Fairy enhancement simulator
Practice a real skill reroll using weighted probabilities. Never changes your actual fairy or orb balance.
Choose a skill slot to begin
Independent sandbox: rolls use exact conditional skill weights. This does not predict an actual in-game outcome.
🧪 Simulation research controls
Reproducible Monte Carlo datasets, downloadable raw costs, and policy selection.
Dataset pending.
📊 Orb cost distribution
How simulated completion costs are spread out (not just the average).
🧭 Budget sensitivity
Compare different orb budgets without changing your fairy.
🎯 Target difficulty
⚔️ Strategy duel
Which sequence of rerolls reaches your wishlist more cheaply?
Select a strategy to inspect its actual decision rules.
🎯 Odds Lab · Exact roll probabilities
Conditional on other learned skills
Selected reroll slot: —
Exact conditional odds update immediately when you change a skill, its rank, or which slot is selected. The learned family is excluded from other slots whether or not locked.
Model & sources
Exact odds use the published Radiant skill weights, renormalized after excluding skill families learned in other slots. Locks restrict the planner's choices; they do not change the game's roll distribution. Free skill acquisition at levels 30, 40, and 50 is modeled as a zero-orb draw. Costs follow the current number of learned skills (2/3/4/5). The long-run forecast is a Monte Carlo estimate under the named strategy, not a proof of the globally optimal policy. The mean and percentiles are shown as unconditional only if all sampled runs finish before the simulation safety cap. Results should be checked against the in-game Change Skill probabilities if the game changes.