What enhancement cost actually means
A displayed success rate does not tell you what one player will spend. It describes a long-run probability. A 50% chance can succeed immediately or fail several times in a row. The useful planning number is therefore an expected cost, paired with a larger reserve for unlucky runs.
Expected attempts
1 ÷ success probability
At 50%, the long-run average is two attempts for that stage.
Failure cost
attempt cost + recovery loss
Include repair, restoration, downgrade recovery, or any material needed after failure.
Safer reserve
90th percentile simulation
A planning buffer, not a guarantee. Ten percent of modeled runs may still cost more.
Calculate your own enhancement plan
Choose a range, enter the live-client probability and costs for every stage, then compare unprotected, protected, and mixed strategies. Your inputs stay on this device unless you create a share link.
Use three numbers, not one
| Number | How to use it |
|---|---|
| Expected cost | Compare strategies over many attempts. Do not treat it as the price of your next session. |
| Typical budget (50%) | A middle outcome: roughly half of simulated runs finish below it and half above it. |
| Safer budget (90%) | A more conservative reserve for a planned push. Keep an additional stop-loss for scarce materials. |
If the expected cost consumes your entire wallet, the upgrade is not affordable. A plan also needs currency for repairs, consumables, market opportunities, and the replacement item that may drop tomorrow.
When protection is worth buying
Protection is not automatically good or bad. Compare its cost per attempt with the expected loss it prevents. The calculator shows a break-even value for every stage:
If protection costs less than that value, it is cheaper on expectation for that stage. If it costs more, unprotected attempts are cheaper on expectation—assuming all other consequences are entered correctly.
Expectation is only one decision factor. A player may reasonably pay more for protection when the lost item is difficult to replace, a downgrade blocks group content, or the recovery material is time-gated.
The five-question upgrade gate
| Question | Green light | Reason to wait |
|---|---|---|
| Does this unlock or stabilize the next content? | A real requirement is blocked | The UI only recommends more power |
| How long will the item remain useful? | Known long-lived piece | Story or dungeon rewards may replace it soon |
| Can the investment transfer? | The global client shows the eligible target and final result | Only a regional guide claims it transfers |
| Are all failure consequences entered? | Chance, downgrade, destruction, recovery, and protection are understood | Any cost is still unknown |
| What is the stop-loss? | You keep a reserve after the planned attempts | The push consumes every scarce resource |
Common enhancement mistakes
- Using the expected cost as a guarantee for one session.
- Ignoring downgrade, repair, restoration, or replacement costs after failure.
- Paying for protection at every stage without checking its break-even value.
- Enhancing temporary gear because the upgrade indicator is visible.
- Continuing after the planned stop-loss because the next attempt “must” succeed.
- Sharing a result without sharing the rates and costs used to calculate it.