Bio Engineer Archive
Thread: BE Nutrient Crafting Guide
Page 2 of 2
Ketu_Sringa
Sat Jun 05, 2004 4:05 pm
#14
After plugging both formulas provided so far into a spreadsheet I found that some were almost dead on to the fully MBE experimented results I had recorded, but some were as much as 4 points low. I looked at the resources involved in these and in those which were accurate and noticed that the OQ was typically very high on the ones which the formulas provided too low a projected max value....
Given this, and a previous suspicion that it's not a pure average across the resources but a weighted average based on how much went in I worked out the following formula which appears to be more accurate so far:
R1 = AVG (FL, PE, OQ, OQ) * resource count
Disclaimer: I have done no other crafting classes seriously so I don't know if things such as resource weighted averages is typical for those classes. Then again, so much of this is seen as voodoo that only a handful of people have reasonable information about How Crafting Works for their profession of choice.
Ketu_Sringa
Sat Jun 05, 2004 4:28 pm
#15
Sorry, st00pid forums making life difficult for me, full post in the works (again...) with real life examples and full numbers.
Meplorium
Sat Jun 05, 2004 4:28 pm
#16
The value is weighted depending on how much resource is used. If you had a schem calling for 900 of one material and 100 of an other and used 1000 OQ for the 900 and 100 OQ for the 100, the average would be 910 OQnot 500.
Ketu_Sringa
Sat Jun 05, 2004 5:28 pm
#17
After plugging both formulas provided so far into a spreadsheet I found that some were almost dead on to the fully MBE experimented results I had recorded, but some were as much as 4 points low. I looked at the resources involved in these and in those which were accurate and noticed that the OQ was typically very high on the ones which the formulas provided too low a projected max value....
Given this, and a previous suspicion that it's not a pure average across the resources but a weighted average based on how much went in I worked out the following formula which appears to be more accurate so far
(changed slightly from original mis-posts for readability):
R1 = AVG (FL, PE, OQ, OQ)
R2 = AVG (FL, PE, OQ, OQ)
etc
approximate_result =
item_min + ((R1 * units_used + R2 * units_used + ....) / total_units_used / 1000) * (item_max - item_min)
Thus, for IntNanonutes made with 90 parts Nebe [dom rice] and 20 parts Kactoa [carn meat], you'd get:
R1 = AVG (192, 925, 998, 998) = 778 (705 "normal" avg)
R2 = AVG (462, 783, 219, 219) = 421 (488 "normal" avg)
OQ-heavy_weighted_result = 90 + ( (778*90 + 421*20) / 110 / 1000 ) * (120-90) = 111.393
normal_avg_weghted_result = 90 + ( (705*90 + 488*20) / 110 / 1000 ) * (120-90) = 109.966
Using 0smspiff0's formula:
slot_weighted_result = 90 + (421 + 778 + 778) / 3 / 1000 * (120-90) = 108.98
unweighted_result = 90 + (421 + 778) / 2 / 1000 * (120 - 90) = 107.895
The actual result I got, maxing out MBE experimentation and with great successes, is 112. All projections fell short, but the OQ-heavy_weight_result was definately closest.
This is simply a good example of where another formula is far off, and not the onlycombine I checked. Doing the same comparison for BSNs and MNS outcomes I have shows it is within about .8 of the outcome each time. In any case where the OQ-heavy average differs significantly from the straight average, the formula above is far more accurate. When it doesn't differ by much then all three methods - 0smspiff0's, mine without OQ-heavy avgs, and mine with OQ-heavy avgs - give pretty accurate estimates of your outcome given the potential to full experiment.
I compared this against 20 combines I had recorded the outcomes and resources for.
This does _not_ take into account the possibility that the DR is actually playing a part as I haven't been recording it. Let's not forget that our docs are awful, and BE hates us as a profession 
so ruling out the potential for the formulas to be borked in some way - either counted OQ twice in the average or averaging all 4 stats rather than the 3 we are told in the docs are important - would be more trusting than we probably should be.
so ruling out the potential for the formulas to be borked in some way - either counted OQ twice in the average or averaging all 4 stats rather than the 3 we are told in the docs are important - would be more trusting than we probably should be.Disclaimer: I have done no other crafting classes seriously so I don't know if things such as resource weighted averages is typical for those classes. Then again, so much of this is seen as voodoo that only a handful of people have reasonable information about How Crafting Works for their profession of choice.
MilkToast
Sat Jun 05, 2004 6:35 pm
#18
0smspiff0 wrote:
"I've gotten up to 89 using high OQ wood, high stat flora and meat.
To get 89 using 1000OQ wood, your meat would have to be OQ1000/PE1000/FL950. Are you sure you got 89?"
and
"I got 89 on a BSN schematic last night using the following resources:
Corellian Deciduous Wood Ahazi/Corellia OQ 995
Rori Vegetable Fungus Ahazi/Rori FL 961, PE 994, OQ 959
Yavinian Carnivore Meat Ahazi/Yavin FL 925, PE 990, OQ 933
Not sure how that matches your formula but I know what the results were, lol."
Plug that into myh formula:
0 0995
961 994 959
925 990 933
----------------
AVG 943 992 962
AVG2 =AVGof the AVG = 965.7778
60 + ((AVG2 / 1000) * (90 - 60))
(60 is the min, 90 is the max)
Result is 88.97333.
Add to that that amazing successes can push it over the top and you get an 89 out of it.
If the flora had been used instead of the wood you wind up with
AVG2 = 964
60 + ((AVG2 / 1000) * (90 - 60))
Result is 88.92 which should also be able to be pushed up to an 89 I would suspect.
If there were no amazing success to push them over then you are still better off using wood, since, as stateb before, it is easier to get good OQ wood than it is to get good FL/PE/OQ flora.
The thing that stands out to me out of this thread is that people think that it is (FL+PE+OQ) / 3 that determines the "value" of the ingredients - that is not the case. An 800/900/900 and a 900/900/800 are NOT the same when it comes to making BE tissues.
Yes I'm sure. I did get an amazing success on the assembly and 1 during experimentation. I actually tried making a schematic just using the meat and flora but couldn't do better than +88.
0smspiff0
Sat Jun 05, 2004 7:16 pm
#19
How is this for the formula?
(this assumes wood - I have not had a chance to test it without wood - and I don't have time tonight to run the numbers).
Using the numbers from above (the one that I used to show mu formula):
BSN - 35 organic, 25 flora, 20 creature
FL total = 25 + 20 = 45
PE total = 25 + 20 = 45
OQ total = 35 + 25 + 20 = 80
FL = 961 * (25 / 45) + 925 * (20 / 45) = 945
PE = 994 * (25 / 45) + 990 * (20 / 45) = 992.2222
OQ = 995 * (35 / 80) + 959 * (25 / 80) + 933 * (20 / 80) = 968.25
Total = (945 + 992.2222 + 968.25) / 3 = 968.4907
Bonus = 60 + ((968.4907 / 1000) * 30) = 89.05472
That gets the 89 bonus without the need for amazing successes.
I have not had time to try out the forula further - and I am not sure if it is different than the one posted above (I think it is - but I didn't read it enough - and I have to run :-)
Ok had just enough time to try it with the stuff above but using the flora in place of the wood:
FL = 952, PE = 993, OQ = 952.5
Total = 965.8333
Bonus = 60 + ((965.8333 / 1000)* 30) = 88.975
I have always assumed that amazing successes let you exceed the calculated bonus cap - I don't think that is the case now. So the 88.975 should be the max bonus regardless of how many amazing successes there are. That holds up to what the poster said actually happened with those resources.
I'd love some people to try this formula out (and again sorry if is the same as the one above - I really don't have time to check it tonight).
The weightedness also explains why it shows up as several points with heavy additive - the wood organic component is weighted much heavier in that example.
Message Edited by 0smspiff0 on 06-05-2004 08:43 PM
Page 2 of 2