The percentages given are NOT accurate, my basis for this is empirical testing (Mr. Puzz3d would be proud). I ran (by saving and reloaded), for multiple assassins, several attempts of varying difficulty (11%, 16%, 23%, and 35% iirc) 25 times each to see if they succeeded. Not *once* did they accomplish their goals, and the ratio of agents dying in the attempt vs just failing was roughly 50%. Before anyone berates me or disagrees, this is hard math and statistics. If a probability is given at say (just for giggles) 16%, that means 16 times out of 100 my attempt should succeed, or reduced this would be 4/25 attempts. 0/25 is *not* 16%, that is 0 percent. Before anyone says "you should have done it more", that argument is invalid, based on the probabilities given, I should have succeeded at least once given my attempts, and the probability of me trying 100 times and it succeeded the last 25 in a row is about as close to nil as one can get. One successful attempt out of 25 would be 4%, not 16%, and so on. As such, I list this is a bug and believe there are several possibilities:
1. The percentage is accurate but the code is broken.
2. The percentage is NOT an accurate portrayal of their chances, this also qualifies as "broken" as the correct indicator should be displayed.
3. The percentage is accurate, at the beginning of each turn the agent "pre-rolls" against all possible characters and it's already decided how it will happen when you make the attempt, as opposed to "making the roll" when you try to execute the action. I *highly* doubt this is how it's coded as it would seem to be an utter waste of RAM, but if it is, it should be fixed to function as "calculated upon attempt" imo.
Whatever the case may be, it is HIGHLY annoying.
*** Addendum: I just tried this for a priest on a heretic with a 33% chance. Tried it 33 times, there were 2 successes. Given the probability listed in the game I should only have had to do it ~3 times max for it to succeed. For those who didn't do the math, 2/33 is 6%.
****** Addendum 2: FINALLY after countless reloading, got a few of them above 1st/2nd grade, up to 5-6 each after a few turns. Even still, they aren't as effective as what I feel they should be, and the rule I stated above about inaccurate percentages and failed attempts still holds true. bleh.
Bookmarks