Bio Engineer Archive

aswex, if you're getting much higher results than you expect occasionally with the same resources, its not because you put the wrong ingredients in its most likely you got an amazing sucess on assembly. I've had the max leap from 58% to 68% with amazing sucesses.

---

Biytor wrote:

This is they way I've had it explained and it works well enough.

Resource OQ / PE / FL

Flora933 857 864

Meat 966 773 617

Wood 983 xxx xxx

-----------------------------------------

Total 2882 1630 1481

Divideby 3 2 2

-----------------------------------------

AVG 961 815 741

Heh, I never notice it either - until I get to the experimentation phase and get much better tissues than I expected

I have done some more testing. But first a correction in (2), (5) and (6) from the previous samples. (2) had 479 instead of 419. (5) had 777 instead of 774 and (6) had 968 instead of 909.

---

2 OQ PE FL
O 771 261 379
F 778 572 575
M 771 261 379
A 795 419 391 -> 53.0% (75)
Great Success 13% (64) -> 60% (78)

5 OQ PE FL
O 819 518 723
F 724 866 286
M 759 810 437
A 774 700 515 -> 66.3% (79)
Great Success 17% (65) -> 69% (80)

6 OQ PE FL
O 1000 839 919
F 968 972 668
M 804 951 948
A 941 909 848 -> 89.9% (86)
Great Success 25% (67) -> 91% (87)

---

I made four new runs using only three materials. Berry/Bean/Insect, Berry/Insect, Bean/Insect and Insect/Berry/Insect.

10 OQ PE FL (Note this is the same as #6)
O 1000 839 919
F 968 927 668
M 804 951 948
A 941 909 848 -> 89.9% (86)
Great Success 25% (67) -> 91% (87)

11 OQ PE FL
O 1000 839 919
F 1000 839 919
M 804 951 948
A 951 867 926 -> 91.5% (87)
Great Success 26% (67) -> 92% (87)

12 OQ PE FL
O 968 927 668
F 968 927 668
M 804 951 948
A 927 967 737 -> 87.7% (86)
Great Success 25% (67) -> 90% (87)

13 OQ PE FL
O 804 951 948
F 1000 839 919
M 804 951 948
A 865 916 937 -> 90.6% (87)
Great Success 25% (67) -> 91% (87)

---

First I tried including the DR into the equation, which is why I used four set. This gave me two negative weights. I then made four set of linear equations using this tests.

([10],[11],[12]) = (91, 92, 90) ==> OQ = 40%, PE = -18%, FL = -14%
([10],[11],[13]) = (91, 92, 89) ==> OQ = 53%, PE = 27%, FL = 20%
([10],[12],[13]) = (91, 90, 89) ==> OQ = 51%, PE = 30%, FL = 18%
([11],[12],[13]) = (91, 90, 89) ==> OQ = 53%, PE = 28%, FL = 19%

The first set gives negative numbers, however the other three gave something close to 50/30/20.

---

Using 50/30/20 as a basis, I revisited all my tests and plugged the weighted average stats into the forumla:

(0.5 * OQ) + (0.3 * PE) + (0.2 * FL) = Final Quality

Below are all of my tests plugged into this formula with the actual percentage given by the run in brackets []:

1) (0.5 * 806) + (0.3 * 581) + (0.2 * 457) = 669 [66]
2) (0.5 * 795) + (0.3 * 419) + (0.2 * 391) = 609 [60]
3) (0.5 * 795) + (0.3 * 681) + (0.2 * 462) = 694 [69]
4) (0.5 * 805) + (0.3 * 718) + (0.2 * 462) = 696 [69]
5) (0.5 * 774) + (0.3 * 700) + (0.2 * 515) = 700 [69] ***
6) (0.5 * 941) + (0.3 * 909) + (0.2 * 848) = 913 [91]
7) (0.5 * 717) + (0.3 * 982) + (0.2 * 846) = 822 [82]
8) (0.5 * 773) + (0.3 * 632) + (0.2 * 541) = 684 [68]
9) (0.5 * 960) + (0.3 * 971) + (0.2 * 186) = 809 [80]
10) [see (6)]
11) (0.5 * 951) + (0.3 * 867) + (0.2 * 926) = 921 [92]
12) (0.5 * 927) + (0.3 * 967) + (0.2 * 737) = 901 [90]
13) (0.5 * 865) + (0.3 * 916) + (0.2 * 937) = 895 [89]

As you can see, using this formula matches all of my results perfectly with the exception of (5), which is likely due to a rounding/truncation anomoly.***

The three corrections at the top were the only ones that did not match, which is why I went back to check my calculations and numbers. That is when I found (2), (5) and (6) were in error.

*** If I calculate this from the base components and never round, the number comes out to 700.069. I would bet they truncate this somewhere in the equation, thus resulting in 699.x.

---

In conclusion, the formula is 50% OQ, 30% PE and 20% FL.

Please test your own results with this and see if it matches so we can confirm it or go back to the drawing board. Once confirmed, we can have the schematic text updated to 50/30/20 or we can ask for the formula to be set to 33/33/33.

[Edited for typos and grammar.]

Message Edited by Seiryuu on 07-13-2004 06:01 PM

Supplementary material:

I also checked my two runs using wood with the 50/30/20 formula.

a OQ PE FL
O 982 --- ---
F 743 1000 798
M 804 951 948
A 862 978 865 -> 89.7% (87) [50/30/20 not 33/33/33]
Great Success 25% (67) -> 89% (86)

b OQ PE FL
O 982 --- ---
F 778 572 575
M 759 810 437
A 862 678 514 -> 73.7% (82) [50/30/20])
Great Success 18% (65) -> 73% (82)

The weighted OQ includes all three materials. The weighted PE and FL are calculated only using the flora and creature food resources.

They match the 50/30/20 formula perfectly.

In addition to providing further support of 50/30/20, it also shows that wood only contributes to OQ. Use this to your advantage when cutting tissues with it.

Been through my tests using a 50/30/20 split and they fit the pattern perfectly.

Only gobby forum spammers are allowed to edit their posts

An amazing sucess on assembly *should* give you a higher base percentage and a higher cap. That has always been the case in my experience