Bio Engineer Archive

ph me pr ps ag final final+ final++ diff
hard 57 50 63 48 51 11 54 54 -1
fort 8 0 20 3 9 2 10 10
dext 54 46 59 49 53 11 11 51 -147
endu 511 500 512 505 514 332 332 373
inte 54 54 57 48 54 11 54 54 4
clev 129 127 134 125 137 37 131 131
depe 1007 1004 1011 1004 1015 1167 1167 1167 -8
cour 510 508 515 507 508 336 336 336
fire 242 232 238 232 237 92 197 237 0
powe 125 113 125 123 121 33 121 121
final 0 2 17 75 6
final+ 3 9 17 75 15
final++ 3 9 21 75 17
%maxed 100 100 75 100 100

OK, what does it mean?

1. The slots line up like they used to. The first 2 numbers on the dna allign with physical. The second 2 with mental, going down the line.

2. Final represents the stage in the combine just before any experimentation is done. Final+ represents after a little has been done. Final++ is after experimenting is finished. The finals on the bottom of the matrix are the %s reported at each stage of experimenting. The finals on the right are the values for each stat.

3. Difference represents the difference between the sample numbers and the final++ numbers. This shows that the % maxed is acurate to a few degrees of error.

4. %maxed solves a big question I had a while back. The %s in the experiment window represent the addition of the 2 primary stat values in comparison to the number 2000. I.e. 1167+336 = 1503 which is 75% of 2000. Now, as we can see, the prowess did not max out it's experimenting. It was at 21%. I counted the little pips on there and it was at 18 out of 24, which is 75%. The sample had 59+512 = 571 (571/2000 = .28)for it's prowess score and the final++ had 51+373 = 424 (424/2000 = .21) 424/571 = .742 which is within certain degrees of error. In contrast, the physical did max out its experiment and the numbers for that are: sample had 57+8=65 (65/2000 = .03), final++ had 54+10=64 (64/2000 = .03). 64/65 = 98%. Of course, thereare always rounding errors.

OK, how can we use this?

Hopefully, someone will figure out if the CL on a creation is the result of averaging the CL of the samples or from the numbers on the resulting template. maybe they can use this finding in their research.

Grelgen, this is close, but not quite it I don't think. Try working with some samples that have larger numbers because the %'s don't marry up.

Take your Physique. At 57 and 8, thats 65/2000 = 3%, however your starting template was (11+2)/2000 = 0%.. less than 1%. The numbers are so small that your first experiment into physique bumped you up to the 3% level... and it capped there, but this is again a trick of small numbers and rounding.

One example would be a clone I did that was 707 and 615 in the physique slots. Thats (707+615)/2000 = 66%. The starting template however was (558+463)/2000 = 51%. After maxing out the physique slot I had (675+610)/2000 = 64%. Rounding error isn't enough to account for the 2% discrepency... 2% of 2000 is 40 points, a rather large chunk.

I do believe there is something tied up in these numbers but I"ve been racking my brain to determine how the DNA Samples combine to give you the initial template. i.e. 675 and 610 turning into 558 and 463. I've got about 50 pages of paper crunching thru combinations looking for a formula to tell me WHY 558 and 463. It is the same problem we had wiht the previous system.... the combinations can be done in so many ways that it'll be a stroke of insight to find the formula. And I do believe a formula exists because the experimentation on categories follows a very strict formula for how the numbers change. I just can't find the samples to template conversion

Sumo,

Have you considered downloading a neural network or genetic algorithm package from someplace (like http://www.aridolan.com/ga/gaa/gaa.html)plugging your data into it, and asking it to chew for a while and try to get a formula for you?

You'd have to make sure to feed it ALL the data though (including the quality of the DNA samples, even though gut feeling tells us that the numbers themselves should tell the whole story).

I don't have enough data to try this myself yet, and probably won't until I get back from my upcoming vacation, but it's something to think about.

--Insomnia

im going on right now, ill do 3 combines with heavily varried numbers just to solidify my therory. But, im pretty sure that it matches up that way. BTW, that 3% final was the max for experimenting. Where im going with that statement is that in the experiment phase, the max you should be able to go is within 1% of the parents.

Im also thinking that the resulting stat pools, hp for example with corolate with the combined numbers, like 400/2000 = 10khp while 40/2000 = 100hp. Im sure it wont be a direct relation, maybe a squared or a cubed.