wattahell is happening? i have an huge percentage of guilds from the very beginning of the age, also lots of channeling sciences, but my off spells fail A LOT... like 15% success rate... why was spells success so nerfed?
Printable View
wattahell is happening? i have an huge percentage of guilds from the very beginning of the age, also lots of channeling sciences, but my off spells fail A LOT... like 15% success rate... why was spells success so nerfed?
i have 15.6% guilds. i'm gnome, sage. 97% BE... why did i just fizzle 4 paradises in a row? 1 or 2, sure... but FOUR??
bad luck. also 15.6% is not that much. So.. don't think it's any bug there.
i should not be failing 4 self spells in a row with 15.6% guilds...
Lúthien... did you fail 4 spells in a row over the course of 10 seconds?
I haven't read the whole thread, most everything up to page 5 is just people ranting "it doesn't work!!!!" with little constructive conversation. But I do know that the system has a slight random-number-generator flaw.
If you've ever played morrowind, you'd know that it's fairly easy to get chains of successes with low skill levels by timing it in a very precise manner. What occurred was rather simple, like true random series you had "clusters" of numbers (though, I would say far more than is truly random)... and because the game recognized certain clusters as being "successes" you'd have chains of successes and chains of failures.
The system, in fact, also allowed you to chain fail after getting very high skill levels... which is why I point it out.
In short... the problem two fold:
A) What we think of as random is not mathematically random...
B) Computers are incapable of true random sequences.
The solution is actually fairly complex. They need to replace the current algorithm with a new one... that isn't actually random.
What you all want is an algorithm that generates bell curve results. Meaning that most numbers will reside around 0.5 and very few numbers will be 1 or 0.
But a true random sequence does not generate bell curve. Every number has an equal likelihood of occurring... and there-in lies the problem.