After estimating sphere usage, calculate how many spheres, ammo or supplies to prepare.
capture rate
Capture rate calculator
Each Pal capture coefficient comes from extracted data. Sphere level, HP control, status control and bonuses use the visible site estimate model. The result is preparation guidance, not a capture guarantee.
Do not brute-force captures
The value of this page is estimating sphere cost and risk, not promising that the capture will succeed.
After input, the estimated probability and expected sphere count appear here.
Estimate model (site model, every input is traceable)
| Step | Formula / coefficient | Source |
|---|---|---|
| Base rate | 30 x CaptureRate (the Pal capture coefficient in extracted data) | PalDB extraction layer (data/*.json) |
| Sphere level | +5 percentage points per level. | site model v20260703 |
| HP control | Above half HP -16, low HP +0, 1 HP / mercy control +12. | site model v20260703 |
| Status / bonus | Status control adds direct percentage points; player bonus x 0.18. | site model v20260703 |
| Bounds | Result is clamped to 3%-95%; expected spheres = ceil(100 / probability). | site model v20260703 |
Outside the model: factors that still change capture outcomes
These mechanics are not inputs to the estimate above (the model only uses sphere level, HP control, status control and bonus), but they can change real combat outcomes. This section is community-summary level; exact bonus values still require data verification.
| Mechanic | How to use it | Relation to estimate |
|---|---|---|
| Ring of Mercy | Attacks do not kill; the target remains at 1 HP. | Stabilizes HP control at the best tier and removes overkill-to-zero risk. |
| Weak-point hit | A prompt appears when the sphere hits a Pal weak point, increasing success chance. | Not included in the model; treat it as extra tolerance, not a guarantee. |
| Stealth backstrike | Throw from behind before being noticed to increase success chance. | Same treatment: try it as the opener on high-threat targets. |
| Sphere module | Crafted at the sphere workbench; changes trajectory/range and improves capture rate. | Actual probability can exceed the estimate; exact amount is pending verification. |
| Ability Glasses | Check individual strength in the field before spending spheres. | Does not affect probability, but decides whether this target is worth the cost. |
Capture planning is not one estimate click
| Stage | What to do | If it fails, check first |
|---|---|---|
| Before leaving base | Choose target Pal, sphere level and expected bonus; use the material tool to calculate at least one expected-sphere batch. | If materials are short, refill the sphere chain first; do not throw low-tier spheres at high-tier targets. |
| Before engaging | Check element and target strength, then decide on control, backstrike or Ring of Mercy HP control. | If overkill is likely, switch to lower damage or non-advantage element first; do not only increase damage. |
| Repeated failure | Read probability as a long-run average; several failures do not prove the formula is wrong. | Check sphere level, status control, bonus and hit method before questioning the data. |
| After capture | Record it through Paldex tracking or the Pal page instead of relying on memory. | Open Paldex tracker |
Pre-capture checklist
| Check | How to judge | Next step |
|---|---|---|
| Sphere level | If the estimate is below 25%, raise sphere level or add bonuses first. | Calculate sphere materials |
| Target HP | Above half HP usually means damage control is insufficient. | Open type chart |
| Status control | Sleep, paralysis or stable control lowers repeated-failure cost. | Read probability guide |
| Target data | For high-tier and legendary targets, check the data page and version status first. | Search Pals |
Related entries
Before capture, avoid wrong type choices or excessive damage.
Open target version status, breeding relations and capture-data boundary.
Applicable versionP1 route contract / P2 v2.0 content layer
Updated
SourcePalDB extraction layer (data/*.json) + site estimate model
UpstreamConOps current focus: Palworld P2 content-layer rebuild