Alchemy

[ElderTroll]

Excellent work!

[Myckel]

I've updated the file:
http://myckel.freeshell.org/Morrowind_a ... veling.ods

Now contains for alchemy as minor skill up to level 70.
I'm also trying to get the chances of getting a potion at level 5 and 6 right. But even after > 20 datapoints there is still some variation. On average there should be a successful potion after 7 or 8 attempts for these levels.

[Zini]

Looks like I was cheering too early. The value from the skill record is obviously a plain old multiplier. But the GMST settings give me a headache. Here are my results:

Misc skill (with a GMST value of 1.25): g = (n+1)*0.625
Minor skill (with a GMST value of 1.00) g = (n+1)*0.400
Major skill (with a GMST value of 0.75) g = (n+1)*0.375

where g is the number of required skill gains and n the current level. Right now I can't see any way to integrate the GMST value into the formula. Maybe I missed something obvious and someone else can come up with the solution. If not, then we will require another test run:

Test the skill gain at level 19 with alchemy as a minor skill. Between tests vary the GMST value for minor skills. We will need several values below and above 1. And while you are at it, please check the GMST values for misc and major skills. I got these from scripting for dummies. It would be highly embarrassing, if we spend days to analyse data only to find out that scripting for dummies had the wrong values.

[Myckel]

Will be off for the weekend. I'll take the game with me and see if I can have a look during that time. However, results won't come in untill Monday.

[natirips]

I got these from scripting for dummies. It would be highly embarrassing, if we spend days to analyse data only to find out that scripting for dummies had the wrong values.

Your GMSTs are the same as in Morrowind.esm.

[Lazaroth]

I've started some work on alchemy as well, as per Zini's request.

I've discovered one thing so far in regard to what image/mesh is used when creating a potion:
It's picked at random on load as far as I can tell. I've had the exact same ingredients, put in the exact same order with the same mortar, with both high and low int/alch/luck and each time I reload the save, a new mesh is picked at random. Cell/area change doesn't matter, it stays the same until you reload a save.

If you change either quality of the apparatus, add/remove apparatus, your skill increases or use other ingredients another random mesh is chosen. These also stay the same until you reload a save.

[scrawl]

Thats not really a big issue. Whether a new mesh is chosen depends on the fact if a potion with the exact same values already exists. What would be more interesting, would be if all meshes / icons can occur at all levels of alchemy or if some are specific to high or low levels.

[Lazaroth]

if all meshes / icons can occur at all levels of alchemy or if some are specific to high or low levels.

As I wrote in my previous post; I've gotten exclusive potion meshes with alchemy level 0 and bargain at level 100, and anything in between, so it's as random as you get.

Can't find a pattern as to if some are more common to lower/higher alchemy. To me it seems evenly distributed.

[Lazaroth]

Duration:

I need more data points, but those I have so far all agrees with what UESP says.

(I use brackets below for easier readability)

Duration = ([(Alchemy + [(Int + Luck)/10])/Base_Magicka_Cost] * Qulity_of_mortar) + Other_Apparatus

For positive effects, e.g. fortify:

if(using_retort && using_calcinator)
{
Other_Apparatus = round_to_nearest_Int(Qulity_of_calcinator + (Qulity_of_retort*2));
}
elseif(using_retort || using_calcinator)
{
Other_Apparatus = round_to_nearest_Int(Quality_of_retort_or_calcinator);
}
else
{
Other_Apparatus = 0;
}

I have not tested negative effects yet, but according to the UESP page it's much more complicated and they use a table for that.

Coin value of the created potion:
The UESP page says the same thing as my tests, i.e:
Value = Duration, with Other_Apparatus set to 0.


I haven't got enough data points to say anything about chance yet.

I've attached the values I've gotten so far. It's not complete, but maybe they'll be good for someone.

[Zini]

Not a bad start. btw. final results should go into this wiki page.