Guilds/NERF

Intro

This guild is dedicated to cooking research (but that's not a promise we wont ever diverge!). We also have a secondary goal of teaching people how to cook correctly, without any myths from previous tales creeping in. The GH is located to the east of Meroe CS, donations of bricks/boards/foods appreciated. We can use up 1000s of a food in one experiment.

Guild warehouse for donations for research is at 813, -3656.

Links

---

The Dictionary

The dictionary lists all the definitions of the more complex words and ideas we use on here. Reference it often when writing!

---

Current StatMap

Various data exports, for everyone's perusal.

• http://www.atitd.org/wiki/tale4/Image:Foods.xls.zip
• Ikuu's data, mostly just + or - for each stat. Most of the data is simply culled from Foraging.
• Web based Spreadsheet

The More Advanced Stuff

Recovery Rate

If an item is never used in a recipe, its potency will increase slowly with time (as our current understanding is). To test the recovery rate, we took the average potency of mushrooms at the start of the tale, per Arame's data. Mushrooms are used since they're generally forgotten ingredients. That potency was about 2384 (need Arame's data to find that again!) and approx 226 days later (this is really accurate!) the 'entry level' of never-cooked foods was 2814. This means that over the course of 226 days, the potency recovery has raised potency by 430. We then divide that number to find the potency gain/day, which equates to 1.903 points/day. For general purposes, lets round that to 2. This means that, under the multiplier settings (see the multiplier map for the settings at the time of writing), it would take about a real-life year to raise camel meat from a *1 multiplier to a *2 multiplier.

Ratio Calculation

This is one of the Holy Grails of cooking. This would let us predict the dead zone and harmony zone, as well as the base-heavy zone. This calculation has a basis in potency, shown from the harmony tracking done by Orrin with camel and discorea. (Image to be uploaded.)

Exact Stat Calculation

The stats shown in this spreadsheet clipping seem impossible to achieve. However, an explanation has been found and is supported by data in Media:Cooking_Data_3_weighted.zip. This spreadsheet exists to determine, as best as possible, the actual ratios per food that occurs within a recipe. Once enough data is found, the ratios can be compared to other known data related to the recipe to hopefully find correlations. Currently at a state of being able to predict large <0.8 modifiers, although all relative to that one food. Possibly select a 'base value' and do extensive testing. Problematic if Camel Meat shifts potency. The predictions of the sheet have been basely verified by experimentation.

A Vexing Mixing Problem

Sobe has found examples of when two good pairs don't add up in a total recipe in a way that makes sense. For example:

• 6 Coconut Meat, 1 Thunder Plant - 42 minute duration
• 6 Camel Milk, 1 Sky Gladia - 29 minute duration
• 6 Coconut Meat, 1 THunder Plant, 6 Camel Milk, 1 Sky Gladia - 48 minute, 22 sec duration

The Center spot

A recent discovery, it allows us to find the value of the more potent ingredient exactly, with minimal interference. If there is interference, then it can be negated by moving the ratio close to 1:1.

Discovery: Base's Multiplier

Using the center spot, an approximate formula has been found for the multiplier used for the base ingredient. Through many tests of simple base/additive pairs, the multiplier was seen to increase with the potency of the base and decrease with distance from base to additive. Plotting the potency*distance as x and the apparent value of the multiplier as y revealed a curve that went to 0 at a distance of 1000 units, so the plot was redone with x=potency*(1000-distance). At this point, a sqrt() curve was apparent. A general formula for the curve, basically an inverse to a quadratic equation, is y = sqrt(a*x + b) + c.

Since the center spot is an artifact of the additive's multiplier going to zero as the ratio goes to 1, and the multipliers' values can only be determined at isolated, discrete points, getting exact results for where the multiplier changes value is not possible. Nevertheless, an approximation is possible using curve fitting. An approximation, with accuracy of about 2-3% in the tested multiplier range of 2-7 is:

multiplier = sqrt(18.3*potency*(1000-distance)/1000000 + 2) -.79

Data outside the 2-7 range for the multiplier is scant, due to a lack of ingredients with high potency and the difficulty of finding a good center spot for low valued potencies. The values in the above formula are not set in stone. In particular, it is likely inaccurate as potency*(1000-distance) goes to zero.