Chef Archive

Thread: Anyone have some Spreadheat that calculates the Experiment % depending on resources used?

sciguyCO
Wed May 26, 2004 1:44 pm
#14






Elmmx-5 wrote:
Close... but these formulas still need work it seems. I made a very similar spreadsheet myself. In the end, the formulas I used were just as inaccurate at trying to predict initial food values as Sciguy's are. These are very close, but something is definitely off. Here is what I used as test resources for a test glass of brandy.

Resource A: DR 607 FL 823 PE 847 OQ 918
Resource B: DR 640 FL 780 PE 869 OQ 789

When I combine these to make a glass of brandy, I get a product that is 50 filling, 6 uses, 36m30s duration and +170. The values that both my own efforts and Sciguy's spreadsheet predict are, 51 filling, 6 uses, 36m30s and +171. I know they are close but if I can't even predict if a combination will be 50 filling or not these formulas are not terribly useful to me. I do realize my formulas to predict these numbers me be at fault as well. Here is what I used(in excel terms).

Filling =ROUNDDOWN(55-22*(initial percent),0)
Quantity =ROUNDDOWN(4*(initial percent)+6,0)
Duration =FLOOR(30*(initial percent)+30,0.5)/1440
Nutrition =ROUNDDOWN(90*(initial percent)+150,0)




Is the initial percentage coming out as predicted? There is the very real possibility that the stat ranges are incorrect, or some mid-calculation rounding is causing the stat to be off. Based on those stats, the calculated initial percentage would be 17.76%.


No rounding: Filling = (55 - 22 * .1776) = 51.0928
Round percent: Filling = (55 - 22 * .17) = 51.26


Um....huh. Either way the rounded filling would come out at 51, not the 50 you're seeing.


If I adjust the range so that filling is 55-30 (so a range of 25 instead of 22), the filling comes out as 50.56 (not rounding the initial percent) or 50.75 (rounding the percent).


Don't take the ranges as gospel. Between narrow ranges, value rounding, and possible unknown factors they could easily be off.

Message Edited by sciguyCO on 05-26-2004 02:51 PM





Kriles Ch'artoff , Chilastra server
Master Chef (retired)
Currently doing....stuff
sciguyCO
Wed May 26, 2004 3:12 pm
#15


Well, I've been meaning to clean up the stat ranges. I'll sit down and do some serious number crunching tonight.


If you want to play around, here's how I've been calculating the ranges:


1) Get percentage and stat for two data points. For accuracy, the points should be as far apart as possible.


2) I calculated the Min/Max from the two datapoints using these:


Min= Stat1-[ (Stat2-Stat1)/(percent2-percent1)] *percent1

Max = Min + [(Stat2-Stat1)/(percent2-percent1)]*100


Basically, I assumed a linear function with


Stat(Percent) = Min +(Max - Min)* (Percent) / 100

sothe slope of the line is equal to:

(Max - Min) / 100 = (Stat2 - Stat1 ) / (percent2 - percent1)






Kriles Ch'artoff , Chilastra server
Master Chef (retired)
Currently doing....stuff
svin
Wed May 26, 2004 5:52 pm
#16


I have a question in the form of an example:


Vasarian Brandy


Lets say I start out with nutrition of 280 (this is all I care about for now)


Stats are as follows (only for nutrition)


OQ PE

Resource 1 700954

Resource 2 950343


Gives me weighted stats of:


OQ 825

PE 653.50


This is what I don't understand what do the following mean:


Initial Percentage 18.52

Maximum Percentage 71.00


Does this mean if I max out experiments on Nutrition and I get ONLY Great Successes will I get a buff of426.94 or 426?


I did this by taking 71 - 18.52 = 52.48.



Chestie LaRue
Master Chef | 4/4/0/4 Merchant
-1000, -4770 on Tatooine in the Aegis Mall! - 1400m from Anchohead | 1200m from Bestine
-47, -522 on Tatooine in the Darkfire Cantina (shuttleport here!)
Naritus Server
Bah - I'll throw my professions mastered into the mix: 4 - Still no slot
Rolldice
Wed May 26, 2004 5:59 pm
#17






sciguyCO wrote:



2) I calculated the Min/Max from the two datapoints using these:


Min= Stat1-[ (Stat2-Stat1)/(percent2-percent1)] *percent1

Max = Min + [(Stat2-Stat1)/(percent2-percent1)]*100


Basically, I assumed a linear function with


Stat(Percent) = Min +(Max - Min)* (Percent) / 100

sothe slope of the line is equal to:

(Max - Min) / 100 = (Stat2 - Stat1 ) / (percent2 - percent1)







How do you define and add the experimentation results ( Amazing, Great, average, critical )in the formala ?

What about chef with 12 experimentation points ?


As for percent1 and percent2, can you be more specific to which percent theyre refered to ?


I dont think its possible to make a spreadsheet , there are way too many combinations possible,


Now it




sciguyCO
Wed May 26, 2004 7:15 pm
#18






Rolldice wrote:


How do you define and add the experimentation results ( Amazing, Great, average, critical )in the formala ?

What about chef with 12 experimentation points ?


As for percent1 and percent2, can you be more specific to which percent theyre refered to ?


I dont think its possible to make a spreadsheet , there are way too many combinations possible,





As far as I've been able to determine, the success type of an experiment simply determines the percentage gain you get. A great success adds 7% per point spent, an amazing adds 8%. So if you start at 23%, spend five points and get a great success, you end up at 23 + 5 * 7 = 58%. If you start at 23% and get an amazing succes, you end up at 23 + 5 * 8 = 63%. Better successes get you higher results (sometimes with fewer points spent), but I'm assuming that they don't effect the stats in any other way. That assumption may be incorrect, but it's held up for my testing.


percent1 and percent2 are the experimental percentages in the category you're attempting to calculate the range for. So using the "great success" example, you could use 23% for percent1 and 58% for percent2.


svin: The "points to max" is the number of experimentation points you would have to spend to reach the maximum percentage, assuming great success (which give +7% per point). The buff at 71% (not counting additive bonus) would be 218. With an +85 nutrition buff the final stat would be 218 * 1.85 = 403.3. The actual buff values do get rounded, at least to the nearest whole number, although sometimes it looks like it rounds to an even number, so I'm not sure whether the actual value would be +403, +404, or +402. In any case, I don't quite see how you got your number.






Kriles Ch'artoff , Chilastra server
Master Chef (retired)
Currently doing....stuff
svin
Wed May 26, 2004 9:28 pm
#19

sciguy - I simply made up all the numbers as I don't understand the "base" amount for each food. Can you give mea "real world" example for Brandy? I guess I am not grasping it. Sorry.



Chestie LaRue
Master Chef | 4/4/0/4 Merchant
-1000, -4770 on Tatooine in the Aegis Mall! - 1400m from Anchohead | 1200m from Bestine
-47, -522 on Tatooine in the Darkfire Cantina (shuttleport here!)
Naritus Server
Bah - I'll throw my professions mastered into the mix: 4 - Still no slot
sciguyCO
Wed May 26, 2004 11:00 pm
#20






svin wrote:

sciguy - I simply made up all the numbers as I don't understand the "base" amount for each food. Can you give mea "real world" example for Brandy? I guess I am not grasping it. Sorry.





Ah, ok. I admit my ability to explain my mental gymnastics is stunted, I'll do my best.


Let's use some fruit/berry stats I've been using to make my brandy:


OQ PE

Fruits980997

Berries994890


Weighted Stats987 943.5
Initial Percent 28.21

Max Percent95.70


In this case, the stat weighting is a simple average, because Brandy uses the same amount of berries as fruits. For something like Ahrisa (20 greens, 10 flowers), the green's stats make up 2/3rds of the weighting because they take up 2/3rds of the total resources.


So, this means that after initially assembling the brandy, the percentage listed next to Nutrition will be 28%. The decimals get cut off in the display, it's unclear whether they're tracked internally or not.


Now, when you spend experimentation points, the amount the percentage value increase depends on the type of success. A great success adds 7 per point spent, an amazing success adds 8 per point spent. So if you put in 5 points and get a great success, the displayed Nutrition percent will go up to 28 + 5 * 7 = 63%.


I'm not sure what the increases are for the other success types. A critical failure subtracts 7 for each point spent.


Now, the second part is how the food's stats get determined by the experimental percentages. This part is a bit fuzzy, but here's what I've been assuming.Each food stat has a minimum and a maximum value, obtained at 0% and 100%. Between 0 and 100%, the stat increases linearly, so if going from 10 to 20% adds 50 to the buff, going from 80 to 90% will also add 50 to the buff.


To calculate the food's stats from the experimental percentage you are currently at, I believe the game does this:

CurrentStat = StatMin + (StatMax - StatMin) * CurrentPercent / 100


I've tried to determine the minimum and maximum stat values listed in my chartby recording the stats I get with different percentages, matching those to a line, and seeing where that line crosses 0% and 100%. There is no guarantee that these are 100% accurate, since the crafting system is definitely doing some rounding.


For now, let's assume that the buff value range I have for brandy is correct: +150 to +240.


So, the brandy starts off at 28%. The starting value would be:

150 + (240 - 150) * 28 / 100 = 175.2


Now, actually making the brandy puts the buff at 174. I think there's some "round down to nearest even number" going on, but the buff does occasionally come out with an odd value. /shrug.


So, now I spend 5 experimentation points, and get a great success. This puts the percentage value at 63%. The calculated stat buff would be:


150 + (240 - 150) *63 / 100 = 206.7

The drink displays +206.


I continue spending experimentation points, eventually maxing the nutrition out at 95%. The stat at this point is:


150 + (240 - 150) *95 / 100 = 235.5

And just to confuse me, the displayed buff is 235.


The differences in the displayed vs. the calculated stats could be a couple of things: the experimental percentage is tracked to more decimal places than is displayed, the min and max stats could be off, or the devs may have just decided to do some odd rounding on the stat values.


Now, that whole scenario was done without a BE additive. The additive bonus is a percentage increase of the calculated "base" value using the food's natural min and max. So a BSN with "Nutrition: 85" would take the calculated value and multiply that by 1.85.


Initial with 85 BSN: 174 * 1.85 = +321

5 points with BSN: 206 * 1.85 = +381

Maxed Nut. with BSN: 235 * 1.85 = +434


Those might also be a bit off from what you actually see in the game, since I'm not sure if the base buff gets rounded before the BSN multiplier is applied.


And now that I've probably thrown around enough numbers to scare off most normal people, I think I'll give it rest.





Kriles Ch'artoff , Chilastra server
Master Chef (retired)
Currently doing....stuff
svin
Thu May 27, 2004 8:11 am
#21






sciguyCO wrote:


For now, let's assume that the buff value range I have for brandy is correct: +150 to +240.






I understood everything but this. Where do you get the assumptions of min/max of foods?


Thanks for the long dissertation, I hope i helps others!





Chestie LaRue
Master Chef | 4/4/0/4 Merchant
-1000, -4770 on Tatooine in the Aegis Mall! - 1400m from Anchohead | 1200m from Bestine
-47, -522 on Tatooine in the Darkfire Cantina (shuttleport here!)
Naritus Server
Bah - I'll throw my professions mastered into the mix: 4 - Still no slot
Elmmx-5
Thu May 27, 2004 9:29 am
#22

Sciguy used the following formulas to estimate the stat ranges.

Min = Stat1-[ (Stat2-Stat1)/(percent2-percent1)] *percent1
Max = Min + [(Stat2-Stat1)/(percent2-percent1)]*100

Using these formulas, I plugged in a couple values to try to get some solid stat ranges for brandy. Here are the following data points I used for brandy and the results. 6% filling = 52 fill, 87% filling = 38 fill, and... 6% nutrition = 154, 91% nutrition = 232. If you use Sciguy's above formula you get a range of 148.4941 to 240.2588 for brandy's nutrition and 53.03704 to 35.75309 for filling. These seem to do a pretty good job finding the max value, but leave the min value open for interpretation. I think maybe if we can get some critical fails we'll have a more accurate min value. Any 4-4-4-0 chefs want to test this?



_________________________________
Akiko' Saito
+25 Master Chef (14pt artisan, force experimentation)
Shop at -6009 -447 Corell Peaks, Corellia


svin
Thu May 27, 2004 1:11 pm
#23

Gotcha, thanks! I am going to work on this some and try to fit it into an excel spreadsheet.



Chestie LaRue
Master Chef | 4/4/0/4 Merchant
-1000, -4770 on Tatooine in the Aegis Mall! - 1400m from Anchohead | 1200m from Bestine
-47, -522 on Tatooine in the Darkfire Cantina (shuttleport here!)
Naritus Server
Bah - I'll throw my professions mastered into the mix: 4 - Still no slot
Page 2 of 2