Input needed for recipe analysis technique

[Michael Ohene]

Hello all,

I am trying to develop a way to determine how recipes will turn out. Right now, the calculator is able to determine which recipes are good and which are bad based on whether a recipe falls in its pre-labeled place on a chart. The calculator input takes about 20 seconds at most.

What I did: Using a math formula, I plotted 100's of recipes from Bon Appetit and Gourmet Magazine. The result was a clearly defined pattern. The formula is basically the traditional wet to dry ingredient ratio, but instead of considering all wet ingredients to be proportional to their weight, I give each wet and dry ingredient a pre-defined weighting (weighted sum).

In order of importance...

The highly-recommended chart is at: plot of baked goods. The calculations are performed by a calculator: Baking Calculator. And the calculations are in the excel file at : calculation spreadsheet.

Some people like recipes in grams so I programmed the calculator to convert from weight to volume and vice versa. Hydrations are also covered.

I am not sure if I provided enough information for people to verify things. What more would you all suggest to have this procedure to be more robust? My hope is that this procedure can be used to diagnose problems with recipes and aid in the creation of recipes.

If you don't get this, just look at the chart. Thanks.

great loaves,

ohene

[Lisa Shock]

Weight doesn't just convert to volume, that's why bakers use weight. And, why home cooks who use volume measure for dry ingredients have inconsistent results. For example, a cup of flour can weigh anywhere from 3.5 to 6 ounces.

[Chris Amirault]

Can you help us understand the qualitative judgments -- good, bad, verify, robust?

[Michael Ohene]

First, if you have questions about volume to weight conversions. Email me or message me privately.

A brief justification for using volume, flour can weigh 99g per cup, true. When it does it is called sifted flour. If you look at my calculator I have a separate entry for sifted flour, so the 3.5 oz to 6 oz range quickly narrows. The chart is also not trying to replicate the exact recipe made by the author, it is instead grouping recipes into baked goods with the same qualities as the original recipe.

Understand "good, bad, robust"

bad - is defined as receiving bad reviews, based on approval

good - is receiving good reviews

"robust", is a bad word, "accurate" is better

verify - are there results that are obviously wrong, is there an error in the chart/calculator, I have made tons of improvements made from peoples suggestions. Hydration is something I am sure needs work.

great loaves,

Michael

Edited December 9, 2010 by Michael Ohene (log)

[Michael Ohene]

Let me think about the conversion once again...This is a very icky issue.

I think you are right, I know your point was that flour varies by grams, but it also proportionally varies by compaction. In a standard recipe, highly compacted flour is not recommended, neither is sifted flour.

I used 140g/cup, but 160g/cup might be more accurate.

Michael Ruhlman, had a weight ratio for biscuits, 321 biscuits, which I had as

[0.499, 10.42%, 0E] with 140g/cup and

[0.523, 11.8%, 0E] with 160g/cup.

0.50 is my low figure for biscuits

The second group of numbers (0.523..) is more accurate. Someone also suggested that 160g/cup is the more common figure.

Anyway the change to 160g should only effect conversions to grams for flour and recipes with flour/starter given in mass. It should not affect the analysis of volumetric recipes. I should have everything fixed by the end of the day.

Someone also suggested, letting the user choose the grams/cup for flour. I haven't figured out a clean way to do that yet.

-Michael