Planning an OSRS Grind: Kills, Hours, and When to Stop
Most grinds go wrong at the planning stage, not the grinding stage. People commit to an item without converting the drop rate into hours, hit the “expected” kill count, and discover the maths never promised anything there.
Step 1: convert the drop rate to hours, not kills
Kill counts are abstract. Hours are the thing you actually spend. Take the confidence levels from the OSRS dry calculator and divide by your realistic kills per hour.
For a 1/512 drop at 25 kills/hour:
| Confidence | Kills | Hours |
|---|---|---|
| 50% | 355 | 14 h |
| 63.2% (“expected”) | 512 | 20 h |
| 90% | 1,178 | 47 h |
| 99% | 2,356 | 94 h |
Use your rate, not a max-efficiency number from a guide. If you die, bank, restock, or take breaks, your effective rate is lower than the theoretical one — often by 20–30%. A grind planned at 30 kills/hour that actually runs at 22 is 36% longer than budgeted before any bad luck.
Step 2: pick a confidence level you can commit to
This is the decision people skip. “I’ll grind until I get it” is not a plan, because there is no kill count that guarantees the drop — the curve approaches 100% and never reaches it.
Pick a number in advance:
- 50% — a coin flip. Reasonable for a cheap, repeatable grind you enjoy.
- 90% — a serious commitment. About 3.3× the hours of the 50% mark.
- 99% — usually not worth targeting. Double the 90% hours for nine percentage points.
The 90-to-99 step is where planning breaks down. On the 1/512 example that is 47 extra hours to move from “very likely” to “almost certain”, and the item is no more valuable for having taken longer.
Step 3: sanity-check against the real item
Long grinds scale brutally. Twisted bow at 1/3,428:
| Confidence | Kills |
|---|---|
| 50% | 2,376 |
| 90% | 7,892 |
| 99% | 15,784 |
At 3 raids per hour that 90% figure is over 2,600 hours. Seeing the number before starting is the entire point — it converts “I’ll get there eventually” into a decision about how you want to spend a year.
Step 4: set a stopping rule in advance
Write down, before the first kill, what would make you stop. Good stopping rules are specific:
- A kill count. “I stop at 2× the drop rate and reassess.” Concrete and checkable.
- An hours budget. “40 hours, then I move to something else.”
- A goal change. “If the item’s price drops below X, this is no longer worth the time.”
Bad stopping rules are emotional: “I’ll stop when it feels unfair.” That threshold moves every session, and it always moves in the direction of continuing.
The reason to decide early is that a partially completed grind creates a pull to continue that has nothing to do with the odds. Kills already done have no effect on the kills remaining — the rate is unchanged, and the time already spent is spent either way. Deciding in advance is how you avoid re-deciding under that pull.
Step 5: recheck your position, not your luck
Every so often, divide your current KC by the drop rate denominator:
- Under 2× — you are in the normal range. Nothing has gone wrong.
- 2× to 3× — mildly unlucky, roughly the unluckiest 5–13%.
- Over 4× — genuinely rare, top 2%.
The how unlucky am I guide has the full ladder. What this check is for is calibration: most people who feel hopelessly dry are at 1.2×, and seeing that plainly is more useful than any amount of reassurance.
Things that change the plan
Rates that vary with difficulty. Some drops scale with invocation or challenge settings. A higher rate may cost more time per kill — compare hours to the target confidence, not kills.
Multiple uniques from one table. If you want several items, the chance of having all of them is the product of the individual chances, which falls fast. Three items each at 80% is 51%, not 80%.
Tables rolled more than once per kill. Check the roll count on the wiki; more rolls means better effective odds per kill than 1/n suggests.
Group content. Rates and roll mechanics can differ from solo. Confirm before planning around a solo figure.
The honest summary
The formulas cannot tell you when your drop arrives. What they can do is tell you, before you commit, what you are actually signing up for — and give you a number to check against later instead of a feeling. Both formulas behind these figures are set out in the drop probability formula reference.
References & Sources
- [1] NIST/SEMATECH e-Handbook - Binomial Distribution (opens in new tab)
Supports: Independent-trial model behind the confidence-level kill counts used for grind planning.
Verified
- [2] Old School RuneScape Wiki - Drop Rate (opens in new tab)
Supports: Current per-item drop rates, including rates that vary with difficulty or invocation settings.
Verified
- [3] NIST/SEMATECH e-Handbook - Gallery of Distributions (opens in new tab)
Supports: Reference properties of the discrete distributions used to model repeated independent attempts.
Verified