Board Thread:News and Announcements/@comment-26581304-20150906194347/@comment-26871289-20150913220344

Caltain wrote:

My resulting opinion is that the code generates a single pseudorandom number. We should all remember that all possible outcomes of any event must total 100%. It isn't necessary to split the event into two stages. For the case I used, Rare X + X, there is no breed option. It must be an X. If the chance of a Rare X versus a normal X is 30%, then 1-30 produces a Rare X, while 31-100 nets a normal X. Caltain: I certainly appreciate your point - the two stages of my scenario aren't necessary; it could all be managed with one pseudorandom "roll". If I understand your hypothesis correctly, a result of Rare Z4 is always a subset of the range normally allocated to Z4. This is akin to saying that the likelihood of breeding Rare Z4 is correlated to the probability of breeding Z4. I submit that this is substantially similar to my two roll scenario where the Type of monster is determined first and Rarity (where available) is determined second. Under this algorithm it is best to choose the Rare Z4 + Z4 breeding combo since the desired Rare Z4 outcome will be a subset of all possible outcomes. Choosing another breeding combo will just produce some nonproductive outcomes like X1 and Y3 that cannot produce the desired result and therefore just represent wasted breeding time.