A good farming plan answers four questions: what exactly do I want, where can it come from, what counts as one chance, and how much time or gold am I willing to spend? Start with the item browser, then inspect the linked sources and the bestiary.
1. Confirm the exact item
A box is not its contents. A crafting material is not the finished equipment. Similar names can refer to different grades, classes, numbered boxes or temporary versions. Open the item details and compare the icon, requirements and description before planning a route.
Search the full name when possible. If a result lists no source, this handbook has not traced one for it yet. Quest, crafting, exchange, event and shop rewards are separate routes, and each one needs its own search.
2. Read what the percentage applies to
| Rate describes | One attempt means | What it does not mean |
|---|---|---|
| Mob drop | One eligible drop opportunity from that mob | Every mob with a similar name uses the same chance |
| Box contents | Opening the applicable box under its conditions | The chance to obtain the box itself |
| Stage reward | The specified reward opportunity | Every enemy in the stage can drop it |
| Crafting success | Completing that recipe successfully | Every successful result is the item you wanted |
| Result quantity | How many items that outcome gives | The chance of selecting that outcome |
A displayed unknown rate is not 0%. Do not convert an unlisted chance into an estimate. Stage variants, difficulty, eligibility and party distribution must match your run before you apply a published rate.
3. Expected attempts versus your actual chance
For a constant, independent chance p on each eligible attempt, the average number of attempts until the first success is 1 / p. Use p as a decimal: 2% is 0.02, so the average is 50 attempts.
The chance of at least one success after n attempts is 1 − (1 − p)^n. The number of attempts needed to reach a chosen chance q is ceil(log(1 − q) / log(1 − p)). These formulas assume no guarantee counter, no changing odds and no dependency between attempts.
| Illustrative chance per attempt | Average attempts | Attempts for at least 50% chance | Attempts for at least 90% chance | Attempts for at least 95% chance |
|---|---|---|---|---|
| 1% | 100 | 69 | 230 | 299 |
| 2% | 50 | 35 | 114 | 149 |
| 5% | 20 | 14 | 45 | 59 |
| 10% | 10 | 7 | 22 | 29 |
These are mathematical examples, not rates assigned to a particular Lunia item. At 1%, 100 attempts still leave about a 36.6% chance of no success. Even 299 attempts are not a guarantee.
4. Several opportunities in one run
If a run genuinely gives m independent, eligible opportunities, each with chance p, its chance of at least one drop is 1 − (1 − p)^m. Its expected number of drops is m × p. Those are different quantities because a run can produce more than one drop.
Example: four eligible enemies at 2% each give a 7.7632% chance of at least one drop per run, and an average of 0.08 drops per run. Do not simply label the run chance 8%. Do not count an enemy that is optional, absent in your difficulty or not killed on your route.
If the opportunities have different independent chances, use 1 − (1 − p1) × (1 − p2) × .... Do not add alternate versions of the same encounter as if they all happened together. A listed source is a place to investigate, not a promise that every listed mob appears on every run.
5. Boxes and multi-step rewards
For a box that drops with chance a, and a desired result that appears with conditional chance b when that box is opened, the chance through that path is a × b, provided the opening conditions match. For a box inside another box, multiply each applicable conditional chance along the path.
Example: a 10% box drop followed by a 5% desired content chance gives 0.5% per original opportunity, averaging 200 such opportunities. A 100% successful craft with several possible results likewise does not promise a specific result.
Check whether the outcome gives one item or several. An average material yield is useful for planning a recipe, but it is not the same as the chance to receive any material. Never combine mutually exclusive box results as independent rolls.
6. Turn attempts into a time budget
Measure the whole cycle: reaching the stage, loading, fighting, collecting rewards, returning, buying supplies and clearing your bag. A displayed clear time alone can be optimistic for your character and route.
At 2% per run and six minutes per complete run, average time to a first drop is 300 minutes. A 90% chance needs 114 runs, or 684 minutes at the same pace. The second number is not a guarantee either. Failed runs and pauses extend both estimates.
If a route offers chance p only on successful clears and you finish a fraction s of attempts, then s × p is an illustrative effective chance per started run, assuming the reward truly depends on completion. It does not apply to drops already collected before a failed clear.
7. Compare routes by what you keep
| Planning example | Route A | Route B |
|---|---|---|
| Chance per complete run | 5% | 2% |
| Full cycle time | 10 minutes | 3 minutes |
| Runs in one hour | 6 | 20 |
| Expected drops in one hour | 0.30 | 0.40 |
| Chance of at least one in that hour | 26.49% | 33.24% |
Route B wins this example despite its lower chance per run. In practice, also subtract key, potion, repair or other applicable costs, and consider useful side rewards. Compare only rewards you are eligible to keep under the party's distribution rules; a party drop is not automatically your personal drop.
For tradable items, compare the market price with your expected farming time and net income from an alternative activity. Asking prices are not proof of completed sales. Check fees, stack quantity, binding and expiry before buying or listing.
8. Keep a small farming record
Record the exact item, stage and difficulty, eligible opportunities, full cycle time, successes, quantities, costs and party conditions. Keep different difficulties and reward routes separate. Record runs with no drop too; remembering only lucky runs makes a route look better than it is.
A short sample is useful for estimating your clear speed and costs. It is poor evidence for a rare drop rate. Zero drops in ten runs does not disprove a 1% chance, and one lucky drop does not establish a 10% chance.
9. A practical stop rule
Before starting, choose a maximum number of runs, a time budget or a material budget. Stop when the limit is reached, the item drops, access expires or the route stops being enjoyable. Store partial crafting progress and resume another time. This protects both your gold and your attention.
10. Before the next run
- Confirm the exact item and the correct reward route.
- Match stage, difficulty, eligibility and box conditions.
- Separate known chances from unknown ones.
- Count actual opportunities, not all possible encounters.
- Include travel, failed attempts and supply costs in your plan.
- Check inventory room, entry items and party distribution.
- Read alchemy, enchanting or quests if the target belongs to a longer chain.