Mabinogi World Wiki is brought to you by Coty C., 808idiotz, our other patrons, and contributors like you!!
Want to make the wiki better? Contribute towards getting larger projects done on our Patreon!

AP cost.

Fragment of a discussion from Talk:AP Training

I've been investigating how this system works for the past few days as I rank up Puppetry and a few Magic skills. My current model for how I think think the AP cost is calculated is the following:

[AP Cost] = [Exp Gain] x [Skill Modifier] x [Skill Rank Modifier] x [Multiply by 10 if a Production Skill].

Here's what I can say so far:

  • Every skill seems to have a slightly different [Skill Modifier] variable. Two skills (Act 1: Inciting Incident & Act 2: Threshold Cutter) at identical ranks (r7) with essentially identical training methods (70 counts at 0.5 exp, yielding 7 skill exp when AP trained) will have different AP costs; 1 AP and 2 AP, respectively. At first I thought that skills that could be Dan ranked had a different AP cost modifier, since skills like Act 4: Rising Action & Act 6: Crisis had much higher costs using equivalent training methods, but it seems like each skill has a unique constant.
  • The 10-times AP cost for production skills definitely seems true, but this requires more testing. What's more is that it looks like any skill that brings up a menu (things like enchanting) are included in this (requires more investigation). Either that or enchanting has an unusually high [Skill Modifier].
  • The way [AP Cost] is calculated likely ends with a 'round up to the nearest integer' calculation, since it always costs at least 1 AP to train.
  • If a skill has training counts that are not divisible by 5 (for example: Enchant at rC there are 24 counts of "Get a failure.") the number of methods trained will be rounded down (this yields 4 trains, despite 20 % of 24 being 4.8).
  • The AP cost is directly proportional to the amount of skill experience awarded. This means that if you have over 80 % of a given training method complete, the AP cost will keep decreasing linearly as less method counts are available to train. This is also key to determining unknown variables for different skills, since it allows you to see exactly where the breaking points are for when AP training costs change.
  • Higher skill ranks always cost more AP for the same amount of training.
  • Individual training methods within the same skill at the same rank always have the same cost per experience awarded. This means that the efficiency of skill exp to AP spent is identical between different training methods. This may not be true, but with over 200 data points collected at this time, there have been no exceptions.
Hematin (talk)07:39, 16 July 2018