Catch Rate Calculator
Catch rate gets asked in two unrelated worlds. In the field, on the water and in sales logs it is a success percentage: catches out of attempts, with a pace per hour and an honest sense of how much that percentage can be trusted. In Pokémon it is a capture probability set by game code: HP, species, ball and status combined into one number per throw. Most calculators serve one world and pretend the other does not exist.
This one serves both, in two tabs. The general tab turns attempts and catches into a rate per attempt, per-hour figures, a projection for future attempts with at-least-one odds, and a 95% confidence interval. The Pokémon tab applies the Generation III and IV capture mechanics exactly: modified catch rate, shake probability, expected balls and the balls needed for 95% cumulative odds.
The general mode assumes each attempt is independent with the same underlying probability, and says so in its results. The Pokémon mode implements Generation III and IV mechanics; later generations changed the formula details.
Quick Answer
Key takeaways
- A percentage without its attempt count is half a result. 15% from 120 attempts carries a confidence interval of 8.61% to 21.39%; 15% from 12 attempts would be far wider, and the general tab shows the interval every time.
- Per-hour pace and per-attempt rate answer different questions. A faster method with a lower success rate can still land more per hour, which is why the general tab reports both.
- In Pokémon mode, HP is the lever you control mid-battle: halving a rate-45 target HP doubles its modified rate from 15.00 to 30.00 with the same ball.
- Balls for 95% odds is the stocking number, not the expected number. Expect 8.50 balls for the example target, but one run in twenty needs more than 24.
How to use the two tabs
Stay on the general tab for anything measured in attempts: casts, shots, calls, traps checked, passes thrown. Enter attempts and catches, add the hours if you want pace figures, and add a future attempt count if you are planning the next session: the tool returns expected catches over those attempts and the probability of at least one success, which is the figure that answers "is it worth going at all". The projection maths, one minus the failure probability raised to the number of attempts, is the same compounding the compound probability calculator works through for repeated events generally.
Switch to the Pokémon tab before a capture that matters. Enter the target max HP and current HP, its species catch rate (look it up in your Pokédex resources: 255 for the likes of Caterpie, 45 for Gible, 3 for most legendaries), then your ball and its status. The results give the modified catch rate a, the per-throw probability from the shake formula, the expected number of balls, and the number that reaches 95% cumulative odds, which is the number to stock. Where the effort behind a fishing catch rate is itself the question, catch relative to hours or hooks fished, the fisheries-standard version is the catch per unit effort calculator.
The general catch rate formulas
Rate per attempt. Catches / attempts. 18 / 120 = 0.1500, reported as 15.00%. Everything else in the tab builds on this p.
Pace. Catches / hours and attempts / hours. 18 / 6.5 = 2.769 catches per hour; 120 / 6.5 = 18.462 attempts per hour. Pace rewards tempo; rate rewards accuracy.
Projection. Expected catches in n future attempts = p x n (0.15 x 10 = 1.50). Probability of at least one = 1 - (1 - p)^n: 1 - 0.85^10 = 80.31%.
Uncertainty. The 95% confidence interval uses the normal approximation, p plus or minus 1.96 x the square root of (p x (1 - p) / attempts): 15.00% plus or minus 6.39 points, giving 8.61% to 21.39% at 120 attempts, following the standard proportion-interval method set out in the NIST Engineering Statistics Handbook guidance on proportions.
Worked example: 18 from 120 over an afternoon
| Output | Value |
|---|---|
| Catch rate per attempt | 0.1500 (15.00%) |
| Catches per hour / attempts per hour (6.5 hrs) | 2.769 / 18.462 |
| Expected catches in 10 future attempts | 1.50 |
| Probability of at least one catch in 10 attempts | 80.31% |
| 95% confidence interval | 8.61% to 21.39% |
The interval is the row most tools hide. It says the afternoon was consistent with a true long-run rate anywhere between about 9% and 21%, which changes how confidently you should change tactics on the strength of it. Double the sample to 240 attempts at the same rate and the interval narrows to roughly 10.5% to 19.5%: precision is bought with attempts, not with decimals.
The Pokémon capture formula (Generation III and IV)
The games first build a modified catch rate, a, from the target state, then convert it into a shake probability. The mechanics documented in Bulbapedia documentation of catch rate mechanics are:
| Scenario (species rate 45) | a | Per throw | Expected balls | Balls for 95% |
|---|---|---|---|---|
| Full HP (100/100), Poké Ball, no status | 15.00 | 5.88% | 17.00 | 50 |
| Half HP (50/100), Poké Ball, no status | 30.00 | 11.76% | 8.50 | 24 |
| Half HP, Ultra Ball, asleep | 120.00 | 47.06% | 2.13 | 5 |
Read the table as a shopping list written by arithmetic. Weakening the target doubles your odds before you spend anything; the Ultra Ball and sleep together then quadruple them again. Against that, walking into a full-HP legendary attempt with standard balls is a 50-ball plan for 95% confidence, a number most players only learn by running out. Generation V onward reworked parts of this system (critical captures, new ball bonuses, a changed status table), so treat this tab as the classic III and IV ruleset it states, not a universal simulator.
What competing catch rate calculators miss
Miss 1: a bare percentage. The general tools that rank for this keyword, Brave Calculator and Savvy Calculator among them, return catches divided by attempts and stop. No pace per hour, no projection, no interval. A 15% rate from 12 attempts and from 1,200 attempts look identical on those pages and mean completely different things.
Miss 2: one world only. Pokémon-specific tools go deep on balls and generations but cannot answer a fisherman, a receiver or a sales rep; general tools cannot price a single Poké Ball throw. The search term belongs to both audiences, and splitting them across two thin pages serves neither. This page keeps both engines in one card, each with its own worked example.
Miss 3: expected balls without the confidence number. Pokémon tools that do show an expectation rarely show the 95% figure beside it. The expectation is an average over many hunts; the 95% number is what a single hunt should pack for, and on the example target they differ by a factor of three (8.50 against 24).
Frequently Asked Questions
Successful catches divided by total attempts, shown both as a proportion and a percentage: 18 catches from 120 attempts is 0.1500, or 15.00%. The general tab also derives catches per hour and attempts per hour when you enter the time, a projection for a future number of attempts, and a 95% confidence interval around the percentage.
No. Catch rate is a share of attempts that succeeded. Catch per unit effort divides the catch by the work behind it, such as hours fished or hooks set, and is the fisheries science standard for comparing abundance and method efficiency. A caster can have a high catch rate per cast and a poor CPUE if each cast takes an hour of preparation; the two measures answer different questions.
It is the range of true catch probabilities consistent with your data, using the normal approximation p plus or minus 1.96 times the square root of (p times (1 minus p) divided by attempts). A 15% rate from 120 attempts gives 8.61% to 21.39%: the true long-run rate is plausibly anywhere in that band. More attempts narrow the band; at 240 attempts the same rate gives roughly 10.5% to 19.5%.
Solve 1 minus (1 minus p) to the power n equals 0.95 for n: n equals log(0.05) divided by log(1 minus p), rounded up. At a 15% catch rate that is log(0.05) divided by log(0.85), about 18.43, so 19 attempts. The Pokémon mode reports this automatically as the balls needed for 95% cumulative odds.
For Generations III and IV, the game computes a modified catch rate a from HP, the species catch rate, the ball bonus and the status bonus. If a is 255 or more, capture is guaranteed; otherwise the per-throw probability is approximately (b divided by 65535) to the fourth power, with b equal to 1048560 divided by the fourth root of (16711680 divided by a), which works out roughly to a divided by 255. Later generations changed the details, so treat this as the classic formula.
In Generations III and IV, the Ultra Ball times-2 bonus and the sleep or freeze times-2 bonus are the strongest single multipliers in this tool. Combined with low HP they multiply the base rate dramatically: a Gible (rate 45) at half HP, paralyzed, caught with an Ultra Ball has a modified rate of 90, about a 35.29% chance per throw, against 11.76% for the same target at half HP with a standard ball and no status.
Because the projection and interval maths treat every attempt as a fresh roll with the same probability. Real attempts drift: fatigue, weather, learning and changing conditions all move the true rate during a session. The figures are still the best single summary of the session, but a rate built from your first ten attempts of the day should not be projected onto your last ten without that caveat.
Add This Calculator to Your Website
References
- Bulbapedia, Catch Rate: the documented Generation III and IV capture mechanics, including the modified catch rate and shake probability formulas used in the Pokémon mode. Bulbapedia: Catch Rate. Accessed October 2026.
- The Cave of Dragonflies, Generation III Capture Mechanics: an independent derivation of the third-generation capture routine and its probability conversion. Dragonfly Cave: Gen III Capturing. Accessed October 2026.
- The Cave of Dragonflies, Generation IV Capture Mechanics: the fourth-generation variant of the capture routine. Dragonfly Cave: Gen IV Capturing. Accessed October 2026.
- National Institute of Standards and Technology, Engineering Statistics Handbook, Estimation for a Proportion: the normal-approximation confidence interval method used for the general mode interval. NIST Handbook: Estimation for a Proportion. Accessed October 2026.
