Cauldrons 102: Production V1.0

[Katwick]

I downloaded the link that Silly Bubbles posted here, and it works fine for me. We even verified that we get consistent results for some mixed objective scenarios. I didn't try to directly download LibreOffice 7.5, but @Silly Bubbles can probably figure it out.

When I convert to the various platforms, I have to dump and reload the Solver Extension, but the Named Ranges and stuff all seems to be stable. Have you tried using the HTML web version of Excel? Web to Web conversions seem to be a lot cleaner than Web to Desktop.

 

[Katwick]

An Astonishing Breakthrough
For some time I've been unhappy with the "By Guess and By God" approach for limiting the number of each type of Ingredient. If you just add your BEST Ingredient, the costs balloon very quickly, as multiple instances of the same Ingredient are compounded by 154%.

Square of the Sum / Sum of the Squares is a well behaved, much utilized approach for spline fits on data sets. Very much to the point, it can be differentiated so you can use the entire Calculus zoo, which is what the GRG Non-Linear Solver is really good at. If you have the courage, take a look at The Analysis Factor, Statistics How To, Nonlinear Regression, and Inequalities.

I was also gnawing on minimizing the need to fiddle with the constraints in the actual Solver popup.

If you look at the bottom of the How Many column, you'll see an 89 that equals 5^2 + 5^2 + 3^2 + 3^2 + 3^2 + 3^2 + 1 + 1 +1 which is the Sum of the Squares. Compare that to the Square of the Sums, which is 25^2=625. In a sort of reverse logic, you MAXIMIZE the ratio by keeping the DENOMINATOR as small as possible. Now simply multiply the % Success Rate by that ratio, and we have a rigorous methodology for when and how to AUTOMATICALLY bump up the How Many of Each Type constraints, so that the algorithm can't just run up the amount of it's best Ingredient.

I'm using an Effect VLOOKUP() to get the right % Success Rate, and then an INDIRECT(AA5) to create a formula that reflects the rate that's associated with the checkbox effect. The parser refuses to deal with an indirect function, but because the value doesn't change DURING a calculation, the Solver doesn't need to know HOW the formula was created, merely that it won't change when the Ingredients are manipulated. That's the 5.4236 number that's next to the 89. I'll be posting this version of the Beta Cauldron 102 model next week, as soon as I've nailed down a few more of the patterns in the data set.

Did I mention that it worked pretty well? I would have been happy to get within 10%, so it was a total surprise when the Modeled Witch Points were only off by 5110 - 5107 = 3 parts in 5000. That's WAY to close to be a coincidence.

 

[Katwick]

Presets	Effect	Rate	Dpl	#	WPs	$$	Result	11	12	13	14	21	22	23	24	31	32	33	34	Date
11Same 01	01_C_	8.57%	10	1	50	0	8.57%	1	0	0	0	0	0	0	0	0	0	0	0	2023D05
11Same 02	01_C_	11.33%	10	2	105	0	11.33%	1	1	0	0	0	0	0	0	0	0	0	0	2023D05
11Same 03	01_C_	14.05%	10	3	166	0	14.05%	1	1	0	0	0	0	0	1	0	0	0	0	2023D05
11Same 04	01_C_	16.91%	10	4	232	0	16.91%	1	1	0	1	0	0	0	1	0	0	0	0	2023D05
11Same 05	01_C_	21.06%	10	5	305	0	21.06%	1	1	0	1	0	0	0	1	0	1	0	0	2023D05
11Same 06	01_C_	20.39%	10	6	386	0	20.39%	1	1	1	1	0	0	0	1	0	1	0	0	2023D05
11Same 07	01_C_	26.64%	10	7	496	0	26.64%	2	1	1	1	0	0	0	1	0	1	0	0	2023D05
11Same 08	01_C_	26.48%	10	8	594	0	26.48%	2	1	1	1	0	0	0	1	1	1	0	0	2023D05
11Same 09	01_C_	29.84%	10	9	728	0	29.84%	2	2	1	1	0	0	0	1	1	1	0	0	2023D05
11Same 10	01_C_	28.22%	10	10	845	0	28.22%	2	2	1	1	1	0	0	1	1	1	0	0	2023D05
11Same 11	01_C_	26.23%	10	11	975	0	26.23%	2	2	1	1	1	0	0	1	1	1	1	0	2023D05
11Same 12	01_C_	35.55%	10	12	1214	0	35.55%	3	2	1	1	1	0	0	1	1	2	0	0	2023D05
11Same 13	01_C_	37.34%	10	13	1414	0	37.34%	3	2	1	2	1	0	0	1	1	2	0	0	2023D05
11Same 14	01_C_	37.89%	10	14	1638	0	37.89%	3	2	1	2	0	0	0	2	1	2	1	0	2023D05
11Same 15	01_C_	41.95%	10	15	1918	0	41.95%	3	3	1	2	0	0	0	2	1	2	1	0	2023D05
11Same 16	01_C_	40.96%	10	16	2127	0	40.96%	3	3	1	2	1	0	0	2	1	2	1	0	2023D05
11Same 17	01_C_	50.28%	10	17	2623	0	50.28%	4	3	1	2	1	0	0	2	1	3	0	0	2023D05
11Same 18	01_C_	50.40%	10	18	2959	0	50.40%	4	3	2	2	1	0	0	2	1	3	0	0	2023D05
11Same 19	01_C_	52.15%	10	19	3403	0	52.15%	4	3	2	3	1	0	0	2	1	3	0	0	2023D05
11Same 20	01_C_	55.53%	10	20	3932	0	55.53%	4	4	2	3	1	0	0	2	1	3	0	0	2023D05
11Same 21	01_C_	58.52%	10	21	4452	0	58.52%	5	4	1	2	1	0	0	3	1	3	1	0	2023D05
11Same 22	01_C_	58.75%	10	22	4933	0	58.75%	5	4	2	2	1	0	0	3	1	3	1	0	2023D05
11Same 23	01_C_	60.45%	10	23	5594	0	60.45%	5	4	2	3	1	0	0	3	1	3	1	0	2023D05
11Same 24	01_C_	64.80%	10	24	6467	0	64.80%	5	5	2	3	1	0	0	3	1	3	1	0	2023D05
11Same 25	01_C_	65.66%	10	25	7314	0	65.66%	5	5	3	3	1	0	0	3	1	3	1	0	2023D05

Take a close look at the irregular progression of the Witch Points. Try fitting a curve to that beast, which is what a couple of the other models are trying to accomplish. The % Success Rate is even worse; it actually goes BACKWARDS a couple of times.

 

[Katwick]

Avoid Weekly Diamonds 23_31_32_33

To Zoom, use Ctrl +/-, Ctrl Mouse Wheel, or a two fingered stretch.

This chart is also available, full sized, in the Weekly Diamonds directory for Caldrons 102. There's also a stripped down Google Sheet, that only includes the Recipe Library and Recipe Charts, but not the Recipe Creation sheet. Using the spreadsheet will allow you to Filter and Sort the data, as well as allowing you to record your own Recipes to see what they do to all the fancy charts. The data, which is a LOT more convenient when you have the spreadsheet tools that are readily available, is in the next two posts.

 

[Katwick]

Effect	Rate	Dpl	#	WPs	$$	Actual	11	12	13	14	21	22	23	24	31	32	33	34	____Date____	Title_with_an_optional_screenshot_Link,_or_perhaps_a_HoneyDew_List.
01_C_	75.34%	10	25	9180	0	01_C_	6	5	5	3	0	0		4				2	2023-04-09	Avoid 23_31_32_33 #25 01 used preset of 1 on 11
02_WS	8.57%	10	1	50	0	02_WS	0	0	0	0	1	0		0				0	2023-04-09	Avoid 23_31_32_33 #01 02
02_WS	11.33%	10	2	105	0	02_WS	0	0	0	0	1	1		0				0	2023-04-09	Avoid 23_31_32_33 #02 02
02_WS	12.54%	10	3	166	0	02_WS	0	1	0	0	1	0		0				1	2023-04-09	Avoid 23_31_32_33 #03 02
02_WS	21.13%	10	4	232	0	02_WS	0	1	0	0	1	0		1				1	2023-04-09	Avoid 23_31_32_33 #04 02
02_WS	24.72%	10	5	305	0	02_WS	0	1	0	0	1	1		1				1	2023-04-09	Avoid 23_31_32_33 #05 02
02_WS	22.77%	10	6	386	0	02_WS	1	1	0	0	1	1		1				1	2023-04-09	Avoid 23_31_32_33 #06 02
02_WS	28.96%	10	7	496	0	02_WS	1	1	0	0	1	1		1				2	2023-04-09	Avoid 23_31_32_33 #07 02
02_WS	33.55%	10	8	620	0	02_WS	1	1	0	0	1	1		2				2	2023-04-09	Avoid 23_31_32_33 #08 02
02_WS	35.31%	10	9	760	0	02_WS	1	2	0	0	1	1		2				2	2023-04-09	Avoid 23_31_32_33 #09 02
02_WS	39.05%	10	10	917	0	02_WS	1	2	0	0	2	1		2				2	2023-04-09	Avoid 23_31_32_33 #10 02
02_WS	44.18%	10	11	1114	0	02_WS	1	2	0	0	2	1		2				3	2023-04-09	Avoid 23_31_32_33 #11 02
02_WS	45.45%	10	12	1309	0	02_WS	1	2	0	0	2	2		2				3	2023-04-09	Avoid 23_31_32_33 #12 02
02_WS	43.45%	10	13	1466	0	02_WS	1	2	1	0	2	2		2				3	2023-04-09	Avoid 23_31_32_33 #13 02
02_WS	46.71%	10	14	1729	0	02_WS	1	2	1	0	3	2		2				3	2023-04-09	Avoid 23_31_32_33 #14 02
02_WS	52.77%	10	15	2060	0	02_WS	1	3	1	0	2	2		2				4	2023-04-09	Avoid 23_31_32_33 #15 02
02_WS	56.49%	10	16	2393	0	02_WS	1	3	1	0	2	2		3				4	2023-04-09	Avoid 23_31_32_33 #16 02
02_WS	59.32%	10	17	2780	0	02_WS	1	3	1	0	3	2		3				4	2023-04-09	Avoid 23_31_32_33 #17 02

 

[Katwick]

Effect	Rate	Dpl	#	WPs	$$	Actual	11	12	13	14	21	22	23	24	31	32	33	34	____Date____	Title_with_an_optional_screenshot_Link,_or_perhaps_a_HoneyDew_List.
02_WS	59.11%	10	18	3125	0	02_WS	2	3	1	0	3	2		3				4	2023-04-09	Avoid 23_31_32_33 #18 02
02_WS	62.43%	10	19	3626	0	02_WS	2	4	1	0	3	2		3				4	2023-04-09	Avoid 23_31_32_33 #19 02
02_WS	68.37%	10	20	4227	0	02_WS	2	3	1	0	3	2		4				5	2023-04-09	Avoid 23_31_32_33 #20 02
02_WS	71.57%	10	21	4883	0	02_WS	2	4	1	0	3	2		4				5	2023-04-09	Avoid 23_31_32_33 #21 02
02_WS	70.86%	10	22	5401	0	02_WS	2	4	2	0	3	2		4				5	2023-04-09	Avoid 23_31_32_33 #22 02
02_WS	73.55%	10	23	6245	0	02_WS	2	4	2	0	4	2		4				5	2023-04-09	Avoid 23_31_32_33 #23 02
02_WS	75.84%	10	24	7025	0	02_WS	2	5	2	0	4	3		3				5	2023-04-09	Avoid 23_31_32_33 #24 02
02_WS	80.03%	10	25	8063	0	02_WS	3	5	2	0	3	3		3				6	2023-04-09	Avoid 23_31_32_33 #25 02 - Katwick 2023D09
03_SB	61.41%	10	25	8692	0	03_SB	1	0	7	4	3	3		2				5	2023-04-09	Avoid 23_31_32_33 #25 03 used preset of 1 on 13
04_T1	81.84%	10	25	33418	0	04_T1	9	4	0	0	0	0		0				12	2023-04-09	Avoid 23_31_32_33 #25 04
05_T2	64.34%	10	25	9437	0	05_T2	6	4	5	1	0	6		2				1	2023-04-09	Avoid 23_31_32_33 #25 05
06_T3	47.87%	10	25	11566	0	06_T3	0	7	6	5	5	2		0				0	2023-04-09	Avoid 23_31_32_33 #25 06
07_ST	56.19%	10	25	11291	0	07_ST	0	8	4	4	0	0		6				3	2023-04-09	Avoid 23_31_32_33 #25 07
08_GS	46.00%	10	25	30720	0	08_GS	0	8	0	0	0	0		0				17	2023-04-09	Avoid 23_31_32_33 #25 08 use preset of 1 on 34
09_OR	60.50%	10	25	9252	0	09_OR	5	1	0	7	3	3		1				5	2023-04-09	Avoid 23_31_32_33 #25 09
10_GM	62.49%	10	25	10392	0	10_GM	0	0	5	4	6	2		7				1	2023-04-09	Avoid 23_31_32_33 #25 10
11_SM	60.78%	11	25	14487	0	11_SM	5	8	3	1	0	0		0				8	2023-04-09	Avoid 23_31_32_33 #25 11
12_GD	44.16%	12	25	9166	0	12_GD	4	4	5	2	0	0		4				6	2023-04-09	Avoid 23_31_32_33 #25 12

 

[Katwick]

There are a couple of things worth noticing.

• The Diamond Ingredients for the week of April 10th, 23_31_32_33 were AVOIDED. While there's a different group of four each week, things just move around a bit. It's like a balloon animal; if you squeeze it in one place, it pops out somewhere else.
• There's a 25 Ingredient Recipe for each of the 12 Effects, plus a full 25 ingredient run for 02_WS Supplies.
• For every Effect there are three Ingredients with a negative Value Contribution. While it's more apparent in the model, the same pattern also shows up in the Brewery if your first Ingredient dumps you into a hole. The short version is that every Effect has 9 good starting points for a Recipe, plus 3 useless starting points. For modeling purposes I'm starting with blanks every time and allow the model to choose the Least Cost next Ingredient, UNLESS the model stalls in which case I force the starting point by defining the first Ingredient as the highest available positive value, which is usually the 3 Valued cell for that effect. So that's where the "preset" stuff comes from.
• There aren't any Diamond to be considered in the AVOID state, so the Witch Points always go up rather than sawtoothing when a Diamond Ingredient is added, but even so the Witch Point progression is not smooth.
• The % Success Rate actually goes DOWN once in a while, when you add a Least Cost Ingredient. Models that JUST try to raise the % Success Rate won't be able to find the Least Cost Recipe.

Once taxes are out of my hair, I'll be adding a 25 Recipe run for EACH of the Effects, and with Diamonds allowed as needed. The Brewery is designed with a LIMITED number of Witch Points each week, and a few Diamonds can replace a LOT of Witch Points.

If the 25 Recipe runs go smoothly, I'll be releasing the Production Version of my Cauldron model by the end of April.

 

[ajqtrz]

@Silly Bubbles

I tried the Recipe Maker and found I don't get the same results this week. Turned off diamonds. (7, 9, 10, 11 when numbering game squared from upper left to lower right, 1 to 12). Have 5 goblets, 12 diplomas, 6133 WP. Selected Mercenaries Mayhem as my target.

Incremented the following according to the recommendations: 34 12 34 22 24 12 34 22 24 34 12 22 22 22 24 22 24 22 14 24 (20 ingredients), Took the numbers and converted them to the chart in the game with 1 being in the upper left corner and 12 in the lower right (12 total ingredients) Entered 12 2 12 6 8 2 12 6 8 12 2 6 6 6 8 6 8 6 4 8 In your SS I got: 38.2% 5972 WP and 20 ingredients. In the game it comes out 28.6% 8336 and 20 ingredients. I've done this 3 times and get the same results. Either the Recipe Maker has a hitch or I do...and I'm betting on me. Can you run this and see if you get the same results? (Or maybe somebody can find where I'm off in my procedure?)

In exploring the problem I then grouped the original order into groups of 3 and entered them so the least WP were used by the group of 3 (except the last which was 2) The order did make a difference in some groups. In the end the probability remained at 28.6 but the WP spend went down from 8336 to 8139, almost 200 pts. It would appear to me that this should not be if your SS is optimized by WP spent for probability gained in each round...which, if I understand it, should be the recommended path. For instance, in the first group of 3 I found the recommended order of 34 12 34 (12, 2,12) cost 190 WP. But if you put them in as 34 34 12 (12 12 2) it costs 188. In every group there was a lowest cost order but it wasn't always the SS suggested. I do know that grouping by 3 may mean that the boundary between the groups made some of them do better, but the overall results in the game were pretty consistent. My thought is that if you choose the least cost in WP (as a % probability gain) you should maximize your probability but you will also increase the cost. Obviously if you just consider WP cost alone you can lower it but at the price of lowering the probability.

As I ponder this I think my first objective is to get any mistakes I've made out of the way to see if I can get the results expected from your model to match what I get in the game. After that I'll worry about the actual optimized order of entry per WP spent and probability gained.

AJ

 

[Katwick]

I got: 38.2% 5972 WP and 20 ingredients. In the game it comes out 28.6% 8336 and 20 ingredients. I've done this 3 times and get the same results. Either the Recipe Maker has a hitch or I do...and I'm betting on me. Can you run this and see if you get the same results? (Or maybe somebody can find where I'm off in my procedure?)

11_SM is an unfortunate objective because the lack of 33 Cottoncloud Bud knocks out your +3 value. So it's a scramble to find enough points for a reasonable Success Rate. I only have data for 25 Ingredients, but it's scary.

Effect	Rate	Dpl	#	WPs	$$	Actual	11	12	13	14	21	22	23	24	31	32	33	34	Date________	Title with an optional screenshot Link, or perhaps a HoneyDew List.
11_SM																				__ Strength Mercenaries
11_SM	58.77%	11	25	9463	0	11_SM	2	4	1			5		5	1	0	7		2023-04-11	Avoid 14_21_23_34 #25 11 - Katwick 2023D07
11_SM	60.78%	11	25	14487	0	11_SM	5	8	3	1	0	0		0				8	2023-04-09	Avoid 23_31_32_33 #25 11

If you're not scared off by 150 Diamonds, 1925 Witch Points isn't all that bad for a 65.9% Success Rate.

 

[Katwick]

Katwick Cauldrons > Production V1.0
Release Notes: There will always be two Production Copies of the Cauldrons 102 Recipe Sheet posted at Katwick Cauldrons > Production V1.0, except during the Weekly downtime on Saturday morning, while Ingredients are being gathered.

• 1) During the Brewery downtime on Saturday morning, the current Comprehensive Least Cost Recipe Library sheet will be archived.
• 2) Witch Points Only aa_bb_cc_dd will be immediately renamed as Comprehensive Least Cost Recipe Library, and the the Solver will be reset to ALL ingredients.
• 3) As soon as we know the new Weekly Diamond Ingredients, a copy will be made and renamed as Witch Points Only ee_ff_gg_hh and a 25 Ingredient Recipe for each Effect, plus anything interesting, will be added. You are VERY MUCH ENCOURAGED to download the Witch Points Only sheet, and use the Solver to see if you can afford the additional Witch Points that will be required for your favorite Least Cost Recipe, if you don't use any Diamonds.
• 4) The mere addition of a few Recipes will NOT be regarded as a Version change.

 

[Katwick]

Ouch! That Shoe Doesn't Fit!!

If you scan up for the selected 10_GM, you'll notice that 24 Mocking Tongues requires Diamonds this week, which clobbers our highest value cell, the 3, if we're trying to avoid the use of Diamonds. As a consequence we have to overload the remaining 8 Ingredients.

The second best Ingredient for 10_GM, the 2, is ALSO the best Ingredient, the 3, for 08_GS. So no matter WHAT we do to Maximize the Success Rate for 10_GM, we'll never be able to catch up with 08_GS. This isn't at all uncommon. The Recipe Library uses a Green Conditional format to catch your attention when you bet on the 2nd place horse, and we flag the checkboxes as well.

The way we've set things up, the only thing you EVER need to do with the Solver is to click on the Solve button. "Old" Witch Point Only configurations aren't useful as Brewing Recipes, but they're a dandy way to look at the importance of each Ingredient.

The GOOD New is that the Witch Point / Diamond ratios don't change much from Week to Week, because a Least Cost Recipe necessarily has a lot of small amount ingredients, and STILL does a decent job of delivering a respectable Success Rate.

So, with 25*12=300 or so Least Cost Recipes to work with, plus the various Witch Points Only recipes to consider, you should be able to find SOMETHING that suits your fancy.

Also, every column in the Recipe Library can be Filtered and Sorted, and the Graphs are dynamic, so it's pretty easy to compare stuff.

Note that we are NOT optimizing the Total Number of Witch Points, they're calculated after the optimization is done, but the Total Number of Ingredients will get you close, unless you went wild on a single Ingredient or two.

If you do get interested in the Solver, ALWAY start with blank Ingredients so that you'll get repeatable results. The Solver (and the Brewery for that matter although it's less obvious) is sensitive to preloaded configurations, so if you're sequencing, and you mistype, you'll get some pretty baffling results that are nearly impossible to replicate, because you mistyped "something." Save yourself the aggravation; select the _Initialize range and then use the Delete Key so that you'll always have a clean start.

Now that I have some data to work with, I'll be looking at the symmetry in the Recipe Space, and will hopefully be able to show that we actually have ONE basic series of Recipes, and ELEVEN shadow Recipes. If that works out then I'll have a Recipe State Network that I can use to chase down the Shortest Global path, rather than the current approach which calculates a likely path, and then goes for a spline fit.

If you're the least bit intrigued by the heavy duty math behind the Sum of the Squares stuff, you can bend your mind at
Least Squares - The Gory Details and The Art of Problem Solving.

 

[Katwick]

The short version is that there's a LOT of recipe symmetry, but the symmetry is not simple enough to be intuitive nor predictive.

You can use all of the Google Sheet tools by downloading or viewing Symmetry Analysis.

I'm currently working with a bug-free version of the Recipe Space. The main tool that I've added is the Stacked column, which eliminates the offset that's inherent in the Recipe Space, so that I can compare "the same" Recipes for various Effects, by jamming the various Ingredients into a single number, with a shift that compensates for the diagonal offsets in the Recipe Space. For the most part, if the % Success Rates are identical, the Recipes will be identical once you compensate for the offsets.

There are a couple of very interesting patterns:

• 01_C_ is usually the same as the previous 12_GD, except for one additional Ingredient
• Most of the Recipes are very similar, which I've illustrated by listing the twelve Ingredient Recipes for all of the Effects
  • But sometimes there are three different patterns that are interlaced, which pretty much eliminates any simple Rules of Thumb
• You can dodge most of the interlacing patterns by never allowing Ingredients to be removed from the mix, but the costs explode
• Trying to fix the 06_T3 by 34 Ear Truffle and the 09_OR by 31 Eye Sprouts Recipe Space anomalies probably isn't worth the bother.

If you care to look at the following data, you'll see that there are some very regular patterns. The fundamental problem is that there's no inherent limit on the number of each Ingredient that you should be using. Without some sort of restraint on quantities, the Solver will just grab 25 of your BEST ingredient, for an enormous price. Mathematically, using a Square of the Sum / Sum of the Squares is a very elegant, well understood methodology for providing a Least Cost (best fit) incentive, but it's not the sort of thing that you can easily wrap your head around.

Effect	Rate	Dpl	#	WPs	$$	Actual	Title with an optional screenshot Link, or perhaps a HoneyDew List.	Brew	Stacked	same
01_C_			0				______________ Coins	11	0.000000000000	same
01_C_	8.57%	12	1	50	0	01_C_	Symmetry No-Bugs D12 #01 Add 11 v3	11	0.000000000001
01_C_	11.33%	12	2	105	0	01_C_	Symmetry No-Bugs D12 #02 Add 12 v2	11	0.100000000001
01_C_	14.05%	12	3	166	0	01_C_	Symmetry No-Bugs D12 #03 Add 24 f2	11	0.100000100001
01_C_	16.91%	12	4	232	0	01_C_	Symmetry No-Bugs D12 #04 Add 14 v1	11	0.101000100001
01_C_	21.06%	12	5	305	0	01_C_	Symmetry No-Bugs D12 #05 Add 32 f1	11	0.101000101001
01_C_	20.39%	12	6	386	0	01_C_	Symmetry No-Bugs D12 #06 Add 13 v1	11	0.111000101001
01_C_	26.64%	12	7	496	0	01_C_	Symmetry No-Bugs D12 #07 Add 11 v3	11	0.111000101002
01_C_	26.48%	12	8	594	0	01_C_	Symmetry No-Bugs D12 #08 Add 31 f1	11	0.111000111002
01_C_	29.84%	12	9	728	0	01_C_	Symmetry No-Bugs D12 #09 Add 12 v2	11	0.211000111002
01_C_	28.76%	12	10	845	0	01_C_	Symmetry No-Bugs D12 #10 Add 21 v-1, % Drops	11	0.211100111002
01_C_	36.01%	12	11	1071	0	01_C_	Symmetry No-Bugs D12 #11 Add 11 v3, Add 32 f1, Sub 21 v-1	11	0.211000112003
01_C_	39.45%	12	12	1257	0	01_C_	Symmetry No-Bugs D12 #12 Add 14 v1	11	0.212000112003
06_T3	39.45%	12	12	1257	0	06_T3	Symmetry No-Bugs D12 #12	12	0.212000112003	same
03_SB	35.55%	12	12	1214	0	03_SB	Symmetry No-Bugs D12 #12 Mismatch	13	0.211100112003
09_OR	39.45%	12	12	1257	0	09_OR	Symmetry No-Bugs D12 #12	14	0.212000112003
02_WS	39.45%	12	12	1257	0	02_WS	Symmetry No-Bugs D12 #12	21	0.212000112003	same
05_T2	39.45%	12	12	1257	0	05_T2	Symmetry No-Bugs D12 #12	22	0.212000112003	same
07_ST	39.45%	12	12	1257	0	07_ST	Symmetry No-Bugs D12 #12	23	0.212000112003	same
10_GM	39.45%	12	12	1257	0	10_GM	Symmetry No-Bugs D12 #12	24	0.212000112003	same
08_GS	39.45%	12	12	1257	0	08_GS	Symmetry No-Bugs D12 #12	31	0.212000112003	same
04_T1	39.45%	12	12	1257	0	04_T1	Symmetry No-Bugs D12 #12	32	0.212000112003	same
11_SM	39.45%	12	12	1257	0	11_SM	Symmetry No-Bugs D12 #12	33	0.212000112003	same
12_GD	39.45%	12	12	1257	0	12_GD	Symmetry No-Bugs D12 #12	34	0.212000112003	same
01_C_	39.22%	12	13	1466	0	01_C_	Symmetry No-Bugs D12 #13 Add 13 v1	11	0.222000112003
01_C_	37.67%	12	14	1638	0	01_C_	Symmetry No-Bugs D12 #14 Add 21 v-1, % Drops	11	0.222100112003
01_C_	41.95%	12	15	1918	0	01_C_	Symmetry No-Bugs D12 #15 Add 12 v2, Add 33 f-1, Sub 13 v2	11	0.312000212103
01_C_	41.53%	12	16	2127	0	01_C_	Symmetry No-Bugs D12 #16 Add 21 v-1, % Drops	11	0.312100212103
01_C_	46.10%	12	17	2498	0	01_C_	Symmetry No-Bugs D12 #17 Add 11 v3	11	0.312100212104
01_C_	46.42%	12	18	2827	0	01_C_	Symmetry No-Bugs D12 #18 Add 13 v1	11	0.221002121043
01_C_	52.52%	12	19	3403	0	01_C_	Symmetry No-Bugs D12 #19 Add 14 v1, Add 32 f1, Sub 33 f-1	11	0.323100213004
01_C_	55.87%	12	20	3932	0	01_C_	Symmetry No-Bugs D12 #20 Add 13 v2	11	0.423100213004
01_C_	58.52%	12	21	4452	0	01_C_	Symmetry No-Bugs D12 #21 Add 11 v3, Add 24 f2, Add 33f-1, Sub 13 v1, Sub 14 v1	11	0.412100313105
01_C_	60.58%	12	22	5068	0	01_C_	Symmetry No-Bugs D12 #22 Add 13 v1, Add 14 v1, Add 31 f1, Sub 24 f2, Sub 33 f-1	11	0.423100223005
01_C_	57.84%	12	23	5475	0	01_C_	Symmetry No-Bugs D12 $23 Add 24 f2, Add 33 f-1, Sub 14 v1	11	0.422100323105
01_C_	60.19%	12	24	6190	0	01_C_	Symmetry No-Bugs D12 #24 Add 14 f1	11	0.423100323105
01_C_	63.70%	12	25	7122	0	01_C_	Symmetry No-Bugs D12 #25 Add 12 f2	11	0.523100323105

We already have some Traveling Merchants, and I'm speculating that it won't be all that long before we'll have a
Medicine Show that deals with Lotions & Potions.

 

[ajqtrz]

In looking at post #30 I have to admit I love the color patterns. I have no real idea how to use them, but they sure are pretty!

Thanks!

AJ

 

[Katwick]

In looking at post #30 I have to admit I love the color patterns. I have no real idea how to use them, but they sure are pretty!

Thanks!

AJ

The Rigorous Methodology chart is the most interesting, as it allows you to see patterns that are hidden in the data thicket.

The Ingredients are simply stacked, and visually you can easily track which Ingredients are added (or occasionally subtracted) as you step through 1-25 Ingredients. Picking the same information out of the raw data is impractical if not impossible.

 

[Gkyr]

Huh?

 

[Darielle]

Hey, ummm, can ya explain how this works for those of us who are not math PHDs? Honestly, my eyes rolled back in my head half way through the first post.

I am with you, Elramike. I showed this to my son (who has two degrees, in Math and Computer Science from Arizona State University) and he said ... um ... on second thought, I think I'd be banned for relaying what he said (or at least implied, because he doesn't cuss in front of his mommy). Suffice to say that he did understand it, but he thinks I'm crazy to play this game. He also suggested that if I want to use my brain that much on my BREAKS TO RELAX AND PLAY a game, I should just ditch it and work 24/7. I agree.

 

[Moho]

He also suggested that if I want to use my brain that much on my BREAKS TO RELAX AND PLAY a game, I should just ditch it and work 24/7. I agree.

This is the reason why I recommend an empirical method. Since my cities are parked in chapter 6 and there is one Squad Size research I've skipped, the number of effects I can obtain is limited. But I've discovered how to increase the strength of the troops trained in the Barracks, and how to boost my T1 and T2 goods productions. My recipes are cheap, simple, and effective.

 

[Katwick]

I am with you, Elramike. I showed this to my son (who has two degrees, in 1Math a2nd Computer Science from Arizona State University) and he said ... um ... on second thought, I think I'd be banned for relaying what he said (or at least implied, because he doesn't cuss in front of his mommy). Suffice to say that he did understand it, but he thinks I'm crazy to play this game. He also suggested that if I want to use my brain that much on my BREAKS TO RELAX AND PLAY a game, I should just ditch it and work 24/7. I agree.

Katwick Cauldrons > Production is refreshed weekly.

If you prefer to stay in the shallow end of the pool, it's pretty simple.

• Enter your Diploma level
• Click which Effect you want
• Enter how many Ingredients you can probably afford
• Click on Solve (Which is only available on Desktops, not Mobiles)

If you just want to look up a recipe, the Recipe Library tab has a Least Cost recipe for 1 thru 25 Ingredients, for each of the first 12 Effects/Diplomas, which you can sort and filter to a fare thee well, on both Desktops and Mobiles

After that it gets hairy, especially if you're trying to dodge Diamonds.

 

[Darielle]

Katwick Cauldrons > Production is refreshed weekly.

If you prefer to stay in the shallow end of the pool, it's pretty simple.

• Enter your Diploma level
• Click which Effect you want
• Enter how many Ingredients you can probably afford
• Click on Solve

If you just want to look up a recipe, the Recipe Library tab has a Least Cost recipe for 1 thru 25 Ingredients, for each of the first 12 Effects/Diplomas, which you can sort and filter to a fare thee well.

After that it gets hairy, especially if you're trying to dodge Diamonds.

Yeah, the basics are fine, but they should have simplified it before rolling it out. When you offer too many fancy twists and turns, people get turned off even to the basics. I just hate looking at it. And I hate that I was forced to complete the research.

Reminds me of the IRS forms. The basic ones are easy peasy. When you get into the forms for cashing in IRAs or for windfall profits from outside the US, those take hours. And even the IRS volunteers don't understand those, so you either have to slog through them yourself or have an accountant do it at megabucks ... and this is not like the IRS. Elvenar is supposed to be stress relief, not a grind. But for those who like this sort of stuff, to each his own, I guess.

 

[Katwick]

Elvenar is supposed to be stress relief

Ahhh, but there's the rub. I like murky issues where there's enough complexity to make it worth figuring out, and I needed to learn more about Google Sheets and Frontline Solver in any case. Very much to the point, Programmers RARELY understand all of the implications of what they're creating. They're trying to tick off bullets in the Program Specification, and they're just grateful when it actually works, somehow.

To me "the grind" is the Tournament and the Spire, because there's nothing to learn nor investigate once you've been through it a couple of times. The only real challenge is figuring out how to set up your city to produce enough stuff to throw away as soon as you have it. And for WHAT? Some buildings and stuff that just makes the entire model MORE erratic? No Thank You!

On the other hand, it's a great joy to wade into something complex enough to be worth understanding, and in a gaming environment I can keep at it as long as it's still interesting, without somebody bugging me about budgets and schedules. And if something useful is accomplished, so much the better.

For me THAT'S stress relief. Grinding out Troops and Goods that I can throw at the Tournament, Spire, and FA Pit, is just mind-numbing.