This post had been in the front page of official website when Trickster R was still open (2008 August). Captured and reposted by this user
To my knowledge, LK has always been a contributing factor ever since the mature compounding was first introduced (2005 May). I don't think it had ever been changed.
1. LK is the ONLY among character stats that affect mature compounding
LK helps increasing mature compounding rates, which determines the probabilty of getting higher values (acting like a Green Elixer).
In conclusion, the higher the LK is, the better chance you will have to get higher value.
2. The items used here are Lv. 110 Night Sheild, and HP stone 110 (min 726~max 1452).
No other Elixer or Compound Master skill was used in the experiement.
The mature compounding was done by 6 characters with different LK stat (0, 10, 50, 100, 200, 900), each 100 times (600 times in total)
Here are the results:
3. The column on the left side represents the different range of the values
, and each column after that represents the total number of the times
that the character ended up with those results.
Note that the character with 0 LK always got a minimum value.
. Don't attempt to do mature compounding with 0 LK characters (obviously....XD)
Even 10 LK character shows better results (albeit minimal) than 0 LK.
Beginning from 50 LK, we can see the results starts to become decent. The average value is almost higher by 100 from 10 LK.
What's interesting is, the highest value that came out of this experiment (through 600 times of mature compounding) was 1422, from 50 LK (which he/she wasn't able to acheive even with 900LK)
At 100 and 200LK, we can see the values are still being increased, but not as much they did from 10 LK to 50 LK. 200 LK's average value was only higher by 2 than 100 LK.
Although, it's important to note that the probability of getting the lowest range decreased almost by 50%.
. LK 100 and 200 characters never got 726 (the min value of the stone) as end result.
LK 900 character has only managed to get 12 higher points than LK 200 character as an average.
As you can see, LK 100, 200, 900 character does not show any noticeable differences in end results. While LK 100 and 200 are reasonably easy to achieve with all characters, the same can't be said about LK 900.
There is definitely a margin of error with this experiment, as it was only done 100 times with each range of LK, but it can
be a good reference.
In my opinion, get at least LK 100 or 200 if you can when you're doing mature compounding--it will definitely help you with the result. Anything above that, as you can see, is a waste of time and effort.
Of course, if you truly want a good result, go buy the elixers