Request: a chart that gives a relative safe literacy level to read books. IE Ice bolt needs level 10 literacy to have a 80% chance. Corus Maximus 19:29, June 21, 2010 (UTC)
Building this table is unfortunately not as easy as it would seem.
There are some unknown factors that affect the success of spellbook reading. It is definately NOT literacy alone that determines this, something for which I have clear statistical evidence.
When reading spellbooks, you have to read through a certain number of actions, and if you fail at any point then you suffer failure consequences. It seems pretty clear that blessed and cursed on the books influences this chance.
The critical question is-- how many times does it check to see if it fails or succeeds?
You can check "turns" passed, but this number is highly inconsistent and I am unsure how the game determines this. It takes my wizard 3 "turns" to read a contingency book, but a turn only passes every 5 or 6 steps. So one single action is clearly not one turn and a character can clearly take more than one action per turn.
- I am fairly sure that the correct terminology is number of actions, not number of turns. This can most easily be seen by playing races with different base speeds, or riding. A more accurate way is to cast a long buff (ie. holy veil, but not knowledge because that affects the stat we're trying to measure) and record the difference in the spell duration before and after reading. Jimhsu77479 00:10, September 2, 2010 (UTC)
-Spellbook reading success most definitely depends also on MAG. This is most obvious when reading top level spellbooks (aka Wish) with Frisia's Tail equipped. Despite taking over 40 turns, the reading actually succeeds.
Another character I have takes 20 "turns" to read the exact same spellbook. However with a much lower literacy skill and much higher character level and roughly the same learning stat, she has a higher statistical chance of successfully reading the spellbook. So it appears that character level may have a definite impact on whether reading a book of a certain level succeeds.
Mathematically speaking, the more times the game "rolls" to see if you fail, the higher your chance of failure and that chance of failure compounds rapidly. If the game actually checks against your literacy 10 or 15 times, then you have to have a chance of success on each roll approaching 97% or higher to have even a 50% chance of successfully reading. This seems to be unbalanceable from a game design standpoint.
A more logical explanation would be that the game determines when you start reading, how long you will be able to read before failing and if you need more time to read than that you just fail.
This is unfortunately much more complicated than it would seem and I'd need reams and reams of data to drum up a table of even remotely reliable consistency.Elona Sunstrike 05:30, August 12, 2010 (UTC)
I have been building a table, measuring the read time against buff durations. Sadly it requires a tremendous amount of data to have any chance of reverse engineering a formula.
I have lots of higher end data, from literacy score 87 - 130, but limited data for literacy scores below that threshhold.