Chef Archive

Uh, yeah.. 8 digit precision.. wish I thought about that before I hand crafted a few hundred bottles of brandy

Anyway, don't lose heart yet, Akiko. Not everyone enjoys the math, but there are a few around here that do. Most probably just haven't found much need to refine the formulas any more. Now that you brought up an easier way to test than I was using, I want to go back and try this again myself, but I won't be on much this weekend so it won't be right away.

I've satisfied myself that what we have for determining experimental percentage from the resource stats is quite solid, but getting from the experimental percentages to the final stats needs some work. Some of the cases you mention, like the rounding and the 33% vs 1/3 thing I have done test cases for. I can dig up some numbers if you want, but since resources are server dependent its hard to make something everyone can repeat for themselves.

I actually have a spreadsheet that does the simulator/resource compare thing but just for brandy. I've found it helpful, but it doesn't give exactly the right values (I think because the exp % to final stat relationship is not quite linear). It doesn't really matter when the buff size is a couple points off, but I'm afraid there are cases where it will be wrong about a 50 vs. 51 filling, which could really throw someone off. If anyone wants to see, I put it up here:

http://www.geocities.com/muttonjedi/main.html

Just be warned that it will probably spit out bad values at some critical time, and I can't promise its the most intuitive thing to use either


Mutton

I made a few veghash also, and it looks like the harvest bonus really isn't all that precise. Here is what I saw experimenting one point at a time (great successes only):

Oats: 1000 PE, 858 OQ
Beans: 998 PE, 967 OQ

First test: 15 oats + 10 beans

Exp %, Harvesting Bonus (increase)
28, 13.05
35, 13.5 (.45)
42, 14.1 (.6)
49, 14.55 (.45)
56, 15 (.45)
63, 15.45 (.45)
70, 15.9 (.45)
77, 16.35 (.45)
84, 16.8 (.45)
91, 17.4 (.6)
96, 17.699999


Second test: 20 oats + 5 beans

Exp %, Harvesting Bonus (increase)
27, 13.05
34, 13.5 (.45)
41, 13.95 (.45)
48, 14.55 (.6)
55, 15 (.45)
62, 15.45 (.45)
69, 15.9 (.45)
76, 16.35 (.45)
83, 16.8 (.45)
90, 17.25 (.45)
95, 17.699999

Even though the harvest bonus is a decimal number, it jumps by either 0.45 or 0.6 with each experiment. I'm guessing the expirementation value is actually an integer, just multiplied by 0.15 to determine the final bonus (or maybe 14.9999 or something to get the values like the last one).

Am I right to assume your 19.x value was a typo? If it actually jumps by 2 from 96 to 97%, that would be very interesting. Otherwise, it looks to me like 11.25 - 18 is the best fit, but its hard to be sure since there seems to be some rounding going on somewhere.


Mutton

yup you are right, my 19 was a typo, I meant 17.

Something else I noticed with this is that the ratio of min/max for veghash looks extremely close to the ratio for brandy: 11.25/18 = 0.625 and 150/240 = 0.625. Maybe just a coincidence, but most of the numbers in sciguy's chart have close to the same ratio. I'm beginning to wonder if all the foods are just assigned a max value, and the min is always 0.625*max.