Okay—What’s The Scoop on Two Scoops

By The Metric Maven

Bulldog Edition

I have no idea when I first saw the commercial. It’s part of our collective commercial culture. We all know there are “two scoops”  of raisins in a box of Kellogg’s Raisin Bran. Internet academics ask that if there are “two scoops” of raisins in a box, then is there a larger ratio of raisins to cereal in the small boxes than in the large ones? Gregory J. Crowther, Ph.D. and Elizabeth A. Stahl, J.D have done the research and published it in the Science Creative Quarterly. They formalized the hypotheses into: always two scoops, or the scoops are proportional to the box size. The boxes come in 15, 20 and 25.5 ounce sizes. Or when related to people with refined culinary sensibilities:  425, 567 and 723 gram sizes. These intrepid explorers of knowledge at SCQ counted the raisins in these different size boxes, and have reported their results as a range. The credibility of these scientists suffers as they report their results in Ye Olde English units, but I have converted them to the metric system so they may be seriously discussed:

425 gram box  201 (47.29 raisins/100 g) — 241 (56.71 raisins/100 g)

567 gram box  381 (67.12 raisins/100 g) — 294 (51.85 raisins/100 g)

723 gram box  308 (42.60 raisins/100 g) — 331 (45.78 raisins/100 g)

This data forced them to abandon their original hypotheses which they labeled A and B. Like most research it creates more questions than it resolves. They now offer these alternative hypotheses to contemplate:

(C) Kellogg employees are poorly trained in the operation of the scoops.

(D) Kellogg factories are equipped with a very large number of scoops of different sizes such that no two scoops are alike.

(E) Kellogg allocates raisins via some stochastic process rather than with scoops.

I have translated their conclusion to SI so that my readers might understand their weighty observations:


If you like raisins, you should buy Kellogg’s Raisin Bran in [567 gram] boxes, which appear to contain the most raisins per [100 grams]. If you dislike raisins, we recommend the [723 gram] boxes or, better yet, a raisin-free cereal.

To achieve truth in advertising and avoid lawsuits, The Kellogg Company should replace its misleading “Two scoops!” slogan with a statement listing both the mean number of scoops per box (presumably 2) and the standard deviation (roughly 0.4).

Number 50 Disher — click to enlarge

Their research did not provide an answer to “what size is the scoop used for allocating raisins to the boxes?” They did not even offer a hypothesis of what its size might be. Thankfully I have my friend Pierre to diligently work his way through the US culinary forest of literature where there are “ounces, and pottles and quarts—oh my!” The question of scoop size first entered my mind when Alton Brown of Good Eats was discussing the dispensing of—probably cookie dough? He pointed out there is a number printed on the inside of the disher, on the sweeper. My sweeper has a 20 on it. So how big is this scoop? Why 1/20 of a quart of course. You all can visualize that—right? Pierre obtained this information from a top cooking reference which explains the volumes found in US scoops (and confuses mass and weight):

Well, this graphic uses the Scoop  Number like a gauge and 20 is 1/20th of a quart or 0.05 quarts–but only tell you that in the text. The quarts are suppressed and you are offered alternating fluid ounces and cup values to explain the fractional gauge values. I’m even more confused when I use my conversion program to check the table. Well, number 20 should be 0.05 quarts which is 1.6 ounces? The answers are 1.5 fluid ounces and 1.75 ounces. Wow, my converter doesn’t offer either of those:

Ok, let’s get this straightened out. Certainly it must get the metric volume right—right? Well the output is 47.31 mL instead of 45 mL. Ok, that’s enough of this. I truly appreciate Pierre’s hard work finding the cooking reference, but I’m going over their head to Wikipedia. Their entry for scoop has this table:

Wow, there it is, Wikipedia explains the number is scoops per quart, has 1.6 US fluid ounces, and 47 mL, which would be the correct rounding from 47.31 mL. I also have a number 50 disher, which is conveniently left off of the list.

This mess, and other culinary metrology disasters, inspires me to write a one sentence book with the title: Why Johnny and Jane Can’t Cook. The sentence: Because the US does not have the metric system.

But all of this has been for not, as Wikipedia explains, there are more than one kind of scoop:

In the technical terms used by the food service industry and in the retail and wholesale food utensil industries, there is a clear distinction between two types of scoop: the disher, which is used to serve ice cream, measure a portion e.g. cookie dough, or to make melon balls; and the scoop which is used to measure or to transfer an unspecified amount of a bulk dry foodstuff such as rice, flour, or sugar.

Alfred Cralle

The disher or ice cream scoop was created by a Pittsburgh inventor one Alfred L. Cralle in 1897. Mr Cralle at least had the good sense to create a scoop which is calibrated. Even if it is in Ye Old English volumes.  This would certainly allow a merchant to keep track of the amount of ice cream or other commodity they sold to the public which would in turn help them stay in business.

Wikipedia has an illustration of a transfer scoop:

Transfer Scoop — Wikimedia Commons

Scoop of Raisin (85 Scoop)

Transfer Scoop of Raisin (85 Scoop) — Two Scoops would still be two scoops of raisins.

“Two Scoops? I love the idea Darrin”

Uh—oh. This image looks like one of the two scoops shown on the Raisin Bran cereal box, which are expertly utilized by Sol who is apparently a two fisted scooper. I’ve seen this kind of scoop many, many times. I’ve seen it vending screws and nails at hardware stores. When this is done, one always uses a scale to measure the quantity for pricing purposes. These scoops are ubiquitous in grocery stores and supermarkets. They all have one thing in common, I don’t recall ever seeing one with any sort of graduation on it. They are just used to transfer bulk quantities to a scale of some sort, which does measure them. So, at the end of our measurement quest, we have been yet again taken in by a marketing scheme. A transfer scoop does not imply any manner of quantity. It only will transfer the raisins to a device, such as a mass or volume scale, which will then be used to quantify the substance. So kids, there is no such thing as two scoops of raisins, no matter how much that amiable animated sun cheerfully claims otherwise. There is only an unaccountable advertising campaign, which almost certainly designed it that way. Sorry you had to hear it from me first kids.

If you liked this essay and wish to support the work of The Metric Maven, please visit his Patreon Page


The Metric Maven has published a new book titled The Dimensions of The Cosmos. It examines the basic quantities of the world from yocto to Yotta with a mixture of scientific anecdotes and may be purchased here.


Opening Up a Can of Morass

By The Metric Maven

Bulldog Edition

My father once worked in a corn canning factory, and has many tales from his tenure there. The one story which always comes to mind is that of a co-worker, called Roscoe, who marched to a different, and rather stochastic drummer. Roscoe never seemed to be able to get his time card to remotely match his apparent work schedule. One day his boss blew his top after seeing Roscoe’s time card and demanded he immediately come into his office. The boss looked at Rosco and inquired about the entries on his time card.

Boss:  I’ve looked at your card for this week and on Wednesday it indicates you worked 25 hours on that day.

Roscoe:  Yeah.

Boss: Roscoe, could you please explain to me how you could possibly have worked 25 hours on Wednesday.

Rosco: I didn’t take lunch that day.

It was tales like that which made me believe that the canning plant was operated by a mad hatter, with help from Bizzaroworld.

Recently my father sent me this table from a new cookbook:

How could I help but wonder if Roscoe had been tasked with determining the can nomenclature and quantity. An eight ounce can is 1 cup, which makes as much sense as any of the Ye Olde English Arbitrary Grouping of Weights and Measures do. Who would have thought that the next size up from an eight ounce can would be a picnic sized can? Every Midwestern picnic I’ve ever attended would always have cans of Van Camp’s Pork and Beans, “freshly opened” for the occasion, right in front of my eyes. Their website indicates the available cans are in 8 oz, 15 oz, 31 oz, 53 oz, and 114 oz sizes. Well, the cookbook has the wrong table so we’ll re-write it to conform with the VCPB default units:

Contents of Standard Cans:

8 oz Can = 8 ounces
Picnic = 10 ounces
No. 300 = 14 ounces
No. 1 tall = 16 ounces
No. 303 = 16 ounces
No. 2 = 20 ounces
No. 2 1/2 = 28 ounces
No. 3 = 32 ounces
No. 5 = 58 ounces
No. 10 = 80 ounces

The cookbook only matches one value of  VC Pork and Beans. They do not have a “picnic” size can. How on earth was I able to go on all those picnics? Was it a massive cover-up by the well-meaning women of my childhood? My father indicated that the corn canning plant at which he worked, exclusively used No. 303 cans. But how would he know if someone slipped some contraband No. 1 tall cans into the to loading dock?

Among my eclectic collection of books about engineering and science, I have zero references on standard sized “tin cans.” Wikipedia, as usual does not disappoint, and has a section on standard sizes. I’m sure the diligent volunteers there can clear up the confusion:

Can sizes in the United States have an assortment of designations and sizes. For example, size 7/8 contains one serving of half a cup with an estimated weight of 4 ounces; size 1 “picnic” has two or three servings totalling one and a quarter cups with an estimated weight of 10½ ounces; size 303 has four servings totalling 2 cups weighing 15½ ounces; and size 10 cans, most widely used by food services selling to cafeterias and restaurants, have twenty-five servings totalling 13 cups with an estimated weight of 103½ ounces (size of a roughly 3 pound coffee can). These are all “U.S. customary” cups, and not equivalent to the former Imperial standard of the British Empire or the later Commonwealth.

Wait a minute? The picnic size has 1 1/4 cups, which the last time I checked my Ye Olde English references was volume, which would translate into 10 fluid ounces, but it is designated as about 10 1/2 weight ounces? My understanding is that one fluid ounce of water = 1.0425 avoirdupois ounces more or less, so 10 fluid ounces of water is approximately 10.425 weight ounces. This is the picnic size which has an estimated weight of 10.5 ounces. This assumes the density of beans is the same as the density of water.

Apparently this means I should have written the table:

8 oz Can = 8 fluid ounces = 8.34 ounces
Picnic = 10 fluid  ounces = 10.43 ounces
No. 300 = 14 fluid ounces = 14.59 ounces
No. 1 tall = 16 fluid ounces = 16.68 ounces
No. 303 = 16 fluid ounces = 16.68 ounces
No. 2 = 20 fluid ounces = 20.85 ounces
No. 2 1/2 = 28 fluid ounces = 29.19 ounces
No. 3 = 32 fluid ounces = 33.36 ounces
No. 5 = 58 fluid ounces = 60.46 ounces
No. 10 = 80 fluid ounces = 84.4 ounces

When Van Camp’s gives their can sizes in ounces, is it fluid ounces or weight ounces?—with avoirdupois assumed? Perhaps Wikipedia can shed more light into this morass:

In the United States, cook books will sometimes reference cans by size. These sizes are currently published by the Can Manufacturers Institute and may be expressed in three-digit numbers, as measured in whole and sixteenths of an inch for the container’s nominal outside dimensions: a 307 x 512 would thus measure 3 and 7/16″ in diameter by 5 and 3/4″ (12/16″) in height. Notice that this is not in millimetres. Older can numbers are often expressed as single digits, their contents being calculated for room-temperature water as approximately eleven ounces (#1 “picnic” can), twenty ounces (#2), thirty-two ounces (#3) fifty-eight ounces (#5) and one-hundred-ten ounces (#10 “coffee” can).[9]

Ok, so “new” cans are all given as a volume of the outside of the can in 1/16 inch increments, but the “old” can numbers relate to the weight of room-temperature water? As the rest of the world might actually think we in the US did something rational with a measurement, they have to warn them that the can numbers do not relate to millimeters.

The Can Manufacturers Institute has this to say:

The CMI Voluntary Can and End Dimension Reference Manual is a compilation of technical information developed by committees of the Can Manufacturers Institute (CMI). Intended for use by CMI members and other interested industry representatives, this publication is available to the public as a service of the Can Manufacturers Institute. CMI does not provide either an expressed or implied warranty as to their viability or accuracy.

Ah, yes, following long established US tradition, the values provided by industry are all voluntary and they are not to be held responsible if the values are not used or met or whatever. How dare you think they might be held to measurement standards. Here are the very, very, voluntary values:

So what values are used in metric countries? According to Wikipedia:

In countries and regions that use the metric system of measures, most tins are made in 250, 500, 750 ml (millilitre) and 1 L (litre) sizes (250 ml is approximately 1 cup or 8 ounces). In situations where products from the USA have been repackaged for sale in such countries, it is common to have odd sizes such as 3.8 L (1 USA gallon), 1.9 L (1/2 USA gallon), and 946 ml (USA 2 pints / 1 quart).

In metric countries one would expect volume in milliliters and/or mass in grams. Both would be very good.  My pantry shelf indicates that cans of beans and sauces in the US are all given in weight, so the only two values on the label are in weight in ounces and mass in grams. Therefore one much ask a simple question: “why did the cookbook offer the contents in cups?–which are clearly volume—when they are sold by mass?” The simple answer is I don’t know, and I doubt they would have a rational answer either. The designation of cans in the US is archaic, irrational and yet again shows that leaving the magic of technical Darwinism to determine the labeling of quantities in our economy only produces a chaotic situation which only confuses and does not offer clarity. Because this is the case, we are only left peering into an an open can of worms because we have never had a mandatory metric system switch-over in the US.

If you liked this essay and wish to support the work of The Metric Maven, please visit his Patreon Page