2009-08-30, 02:52 AM
Cyanne Wrote:I doubt they individually edit each item's percent chance every single time new things are added in because you always only get one item (the chances of everything must add up to 100%).
Keep in mind they can use lazy stacking to generate a derived percentage.
Under that system it just generates a token between 1 and the sum of all potential and wherever it falls, it falls, like an ever expanding roulette wheel.
example before/after list:
Before:
| Item | Chances | *Cumulative Chance |
|---|---|---|
| Crap A | 100 | 100 |
| Crap B | 120 | 220 |
| Crap C | 150 | 370 |
| Shiny A | 10 | 380 |
After:
| Item | Chances | *Cumulative Chance |
|---|---|---|
| Crap A | 100 | 100 |
| Crap B | 120 | 220 |
| Crap C | 150 | 370 |
| Shiny A | 10 | 380 |
| New Shiny | 50 | 430 |
*Note this field isn't actually in the table, it's just a running summary of where a number would land based on the Chances preceeding it. A token of 150 for example would fall on crap B because it's between 100 and 220.
Under this drop table system new items have a percentage equal to their degree of the overall sum, so adding new items diminishes everything before them unless everything is adjusted as well. It also becomes easier to go back and adjust single items though to give them a larger share of the whole value.
Shiny A for example had a 10 in 380 chance of appearing before New Shiny was added, and now has a 10 in 430 chance, and New Shiny has a 50 in 430 chance.
This is a surprisingly common way of handling random drop tables where percentages are desired because it avoids the need to "equal 100" and in theory provides infinite granularity to the probability while giving easy expansion.
The drawbacks of such a system should be plainly obvious as well though.
It's not having what you want - It's wanting what you've got.