Matt Meshulam's Brain Dump

I’m sure that at this point everybody has heard about Jdimytai Damour, the temp employee who was killed by a herd of overly eager Black Friday shoppers at a New York Wal-Mart yesterday morning. The metaphors that can be drawn from the incident are pretty obvious, but I still had to point out the final paragraph in a follow-up article from the Associated Press:

Items on sale at the Valley Stream Wal-Mart included a Samsung 50-inch Plasma HDTV for $798, a Bissel Compact Upright Vacuum for $28, a Samsung 10.2 megapixel digital camera for $69 and DVDs such as “The Incredible Hulk” for $9.

If I recall correctly from my middle-school newspaper days, the last paragraph of a news article often provides additional background information relevant to the story that doesn’t fit into the overall narrative of the piece. Did the author choose to report these Wal-Mart rollbacks to help readers decide if the death was justified? You make me sick, Associated Press correspondent Colleen Long.

I just came across a fun little site called StateStats, which uses Google’s state-by-state search history data to correlate search terms with demographic information. This article from Silicon Alley Insider found a few good ones, such as that searches for “myspace” correlate moderately with state illiteracy rates.

A few good finds of my own:

• Wine has a strong negative correlation (-0.71) with the proportion of the population who voted for Bush in 2004, implying that wine is more popular in less Republican states. Interestingly, arugula has the exact same correlation with the 2004 election, but it should be noted that there is a restaurant of the same name in Connecticut.
• Obviously, I also tried latte but was surprised to find that Minnesota blew past every other state for this search term. On a per capita basis, Minnesotans searched for “latte” 50% more than did the coffee mecca of Washington, its closest competitor.
• Guns have a 0.76 correlation with voting for Bush. I’ll leave this one to the gun control groups, but that search also correlates 0.7 with suicide rates.
• Meth strongly correlates with the area of a state, and negatively correlates with the density of a state and its longitude. This leads to the intuitive finding that methamphetamine (or at least Google searches for it) is most prevalent in the mountain west region.
• I recalled Google’s Flu Trends project, which hopes to track influenza outbreaks faster than any currently available system through finding regional variations certain searches. So I looked at the map for searches containing “influenza:”I suggest you get in contact with any friends or family that you may have in South Dakota.

Parts one and two can be found here and here, respectively.

Given my newfound interest in eating cheap taken to the extreme, you can imagine my excitement when I came across this snippet from December’s issue of Harper’s. Abridged and slightly paraphrased:

From a list of recipes from residents of Leningrad during the Nazi siege (September 1941-January 1944). The letter, dated 1942, is from a Russian compilation, “I Saw It: New Letters about War,” published by Vremya in 2005.

[...]

Leather-belt soup: It’s better to use undyed belts. Cut the belts into small pieces, then rinse them in water and let soak. After boiling, season with nettle, saltbush, chickweed, or other herbs. It’s good to add a little bit of vinegar.

[...]

Soup from pets and domesticated animals: Meat is ranked by taste in the following order: dog, guinea pig, cat, rat. Gut the carcass, wash well and place in cold water. Add salt. Cook for one to three hours. For aroma: bay leaf, pepper, any sort of herbs, and, if available, grain.

Of course, the food choices available to you when your town is under siege isn’t quite the same constraint as simply being poor. Also, I’m skeptical about the nutritional value of a leather belt, but since there isn’t really anything even close to it in the USDA nutrition database, that will have to remain a suspicion.

If you use the internet, you’re probably familiar with the concept of “tags.” Instead of categorizing articles or links or anything into predefined folders, you can apply any number of arbitrary tags to the object. The idea is that this increased flexibility makes it easier to find entries pertaining to a specific topic when you can tag them with anything.

But there are some big usability problems with tags that make them pretty ineffective in many applications where they’re used, and I would like to bring up some of these issues.

The major benefit of tags that I hear is that it is general/elegant/simple/etc. This is a commendable goal in design, but it is possible to construct a framework that is so flexible, it ceases to be a framework at all. This is what happens with tags in many cases. It’s up to the author to enforce any organization scheme they choose to follow, since it’s usually just as easy to make a new tag as it is to use an existing one. This leads to tags which are virtually duplicates, making it a real pain to actually find what you’re looking for. Take, for example, the tag cloud for my del.icio.us bookmarks:

A few comments: first, I know what you’re thinking, and I’ll be discussing tag clouds specifically in a paragraph or so. Second, I’m not a hardcore Delicious user (tags just aren’t for me) so I probably have fewer tags and bookmarks than the average user. Nevertheless, it’s pretty clear that there’s some serious tagging redundancy going on here: stl, stlouis, 63105; bicycle, bike, cycling; et cetera. I remember bookmarking a great recipe, but did I put it under cooking or recipes? Or just food? Or vegetarian? Or all four? I think you get my point. Sidenote: I’m not sure why I decided that carrots deserved their own tag. Next thing I know, every ingredient in every recipe I bookmark will be clamoring to be tagged. Talk about a slippery slope.

And that brings me to visualizing tags. Tag clouds, as seen in the figures above, are the method of choice for displaying tags on the web. Often the size of each word correlates to how many items are described by each tag. Pro: provides an intuitive visual cue to guide the user to the most important tags. Con: ugly as crap. But my biggest problem with tag clouds is that they completely ignore the relationships between the tags themselves. Redundant tags wouldn’t be an issue if a visualization instantly made it clear that most things that are tagged with election are also tagged with politics.  This is a clustering problem, which is a bit above my pay grade, so I’ll leave it to Zach.

I’ve come up with a few criteria in my search for an alternative to tagging: It must require minimal additional input on the part of the author/categorizor. It needs to be easy to implement, so no crazy semantics algorithms. And there should be no learning required to do it right; it should just work. The best that I’ve come up with is something I call “thesaurus-augmented search.” Google and the other major web search engines do this a little bit, but not nearly to the extent that I’m imagining. Basically, once I finish writing this article, I shouldn’t have to scroll down to the next box and type in anything else. I just post the article. And when I want to find it, I search for it. It doesn’t matter if I don’t use all the right words, because as long as I get the right idea I’ll get my search result.

Of course, there are problems with this approach, mainly that you’ll always get way more results than you would with regular search. But with a little refinement, I think it would  be pretty fun to play with while I’m waiting for the semantic web.

In my quest for better data on food prices, I got sidetracked exploring the U.S. Department of Agriculture’s Economic Research Service. They have an impressive quantity of food-related data, and I’ll probably be exploring it for quite a while. Some interesting things I’ve found:

• The per-capita availability of chicken in the U.S. has steadily doubled over the past 30 years, to 87 pounds per person per year in 2006. Pork has been pretty much constant and beef has decreased by about 12%. Source (Excel spreadsheet)
• North Dakota is responsible for 20% of U.S. honey production, more than double the share of any other state. Source (PDF)
• In 1996 the USDA commissioned a survey to determine how Americans prefer their hamburgers to be cooked, as a part of studying foodborne illness. I assume that internally it was known as “The Hamburger Study.” Source
• “U.S. Cabbage Statistics” would be a great name for a prog rock band.

In December, I will be going to Israel for ten days. I recently thought to take a look at the exchange rate to see if the dollar has gotten any stronger since the summer, when some of my friends were busy getting poor in Europe:

Not too bad. It’s not quite where it was two years ago, but it looks like the summer really was the worst of it. Also, it seems like it’s been pretty volatile lately, what with the global financial crisis and all. So naturally, that got me to thinking: Would it be a worthwhile gamble to invest in foreign currency, speculating on the dollar drastically losing value?

I know it’s possible to invest in currency from the comfort of your own country, but just for fun I wondered what the odds would be of profiting during my trip abroad. Simply put, if I buy Shekels the day I get to Israel, and sell them the day I leave, what would be my expected return?

I got the past two years of USD-to-ILS exchange rate data from Oanda.com, which has a very nice interface for historical currency data. I assumed a 2% premium when exchanging currency, approximately what your credit card company charges. I brought the data into Matlab and made a pretty picture:

As expected, it is highly likely that I would lose money because of the 4% in exchange fees (two transactions: USD-to-ILS and back). Sure enough, the red dotted line indicating the average return is pretty close to 0.96. But there have been a few peaks over 1, indicating a net profit. All of these have occurred in the last year when the economy has been going crazy. In the past two years, there were twenty instances where it was profitable to hold Shekels for ten consecutive days. This translates to about a 3% win rate, which is not very encouraging. Furthermore, the highest-ever return was 1.023, netting just above two cents per dollar invested.

The bottom line is that it may make sense to make long term investments in foreign currency (or stock) as a hedge against the U.S. economy, but the exchange premiums make it pointless to try beating the market on a shorter scale.

For part I, click here.

In part one of this ongoing series, I attempted to verify whether it is actually common for people to eat dog food out of necessity. My results were inconclusive, so I decided to perform some additional analysis. In this part of the series, I will be assessing the economic viability of eating dog food. The third part will cover the nutritional implications of this unconventional diet.

For my economic analysis, I relied heavily on two federal databases: the Bureau of Labor Statistics’ average price database, which is used to calculate the Consumer Price Index; and the USDA’s national nutrient database. Both of these resources were incredibly helpful, and you’ll be seeing me using the USDA database quite a bit in the future (I’ve got some big plans brewing).

I compiled nutrition information and the most current average national price for a small basket of foods that I feel reflect a good variety of low-cost sustenance options. I also estimated calorie content of both dry and wet dog food, based on the “guaranteed analysis” that is found on pet food instead of nutrition facts. This data is presented as a percentage (by mass) of fat, protien, fiber, and water. I estimated that nearly all the remaining mass is carbohydrate, a fairly accurate assumption. Since average national price data was not available for dog food, I used the prices at my local supermarket: $11.98 for a 20-pound bag of Purina™ Dog Chow, and $0.80 for a 13.2-ounce can of Pedigree™-brand wet food. I used this data to plot how many calories’ worth of each item can be bought for a dollar. Let’s look at the results:

Sure enough, dry dog food provides a significantly better value than any meant-for-human-consumtion alternative. Canned dog food doesn’t fare nearly as well due to the fact that 82% of what you are buying is water. I found it surprising that oatmeal fared so well, but note that the average price according to the BLS is much less than a carton of Quaker. Their price is 79 cents per pound, which is actually a bit more than I pay for bulk oats from the store.

Also important to note is that many of the foods listed require additional gas or electric costs for preparation. I may do a more in-depth study that takes this into account, but I suspect that the difference isn’t severe. Intuitively, I would expect a PB&J to be a slightly better deal than black beans, but it probably wouldn’t catch up to spaghetti.

Of course, calories per dollar is the crudest calculation possible to determine the cost effectiveness of food. A proper ratio of macronutrients (proteins, carbohydrates, and fats) as well as dozens of other nutrients are required to stay healthy over even a short period of time. But for that analysis, you’ll just have to wait for part three.

I just came across a story about McDonald’s opening a no-brand eatery in Japan, simply called “Quarter Pounder:”

Apparently there have been so-called “no-brand” establishments in Japan before, but just from the top of my head I could only think of one example of nonbranding in American culture:

Of course, the whole idea of a product or company being brandless is a lie. There are hundreds of design choices required when opening a retail establishment or developing a product. From obvious decisions like a name to more subtle considerations like colors and the look and feel, every choice influences the branding. If you start a store with a generic name written in black-on-white Helvetica, and decorate it as sparsely as possible, that isn’t brandless, it’s American Apparel. My problem is more with the term “no-brand” than with the actual idea. It’s quite a common development which already has a name: Minimalism.

I’m going to stop now before the word “brand” completely dissociates from its meaning in my mind. Brand brand brand.

With all the news about the economy, I’ve been hearing a lot of news stories about how people are changing their behavior to save money. Take this story in the New York Times from a while ago about how people are eating differently nowadays:

Burt Flickinger, a longtime retail consultant, said the last time he saw such significant changes in consumer buying patterns was the late 1970s, when runaway inflation prompted Americans to “switch from red meat to pork to poultry to pasta — then to peanut butter and jelly.”

“It hasn’t gotten to human food mixed with pet food yet,” he said, “but it is certainly headed in that direction.”

<b> added. I’ve heard references to people being so poor they had to eat dog food, but I suppose I just assumed it was a figure of speech. But an expert said it! It must be true! Right? So I turned to my trusted advisor in all important matters: Yahoo Answers. In case you aren’t familiar with the service, Yahoo Answers is a place where the stupidest people on the internet go to answer the stupidest questions on the internet. It is responsible for the “How is babby formed?” meme. So let’s see what wisdom the Hive Mind has for us today.

I found two questions on human consumption of dog food out of necessity. For each posting I will quote a summary of the question as well as some highlights from the answers.

Post #1: Would poor people really eat dog food? There is a stereotype image of poor people having to eat dog food because they don’t have enough money for anything else. Isn’t a tin of tuna or Chef Boyardee just as cheap or cheaper?

• I have never been poor, but I doubt if they have had to eat dog food. Chef Boyardee is pretty cheap. Some of that dog food, is pretty expense [sic]. I am sure they would go to the $1.00 menu at McDonald’s if necessary
• Can of tuna does not have the nutrition [sic]. If you look at the dog food ingredient [sic], you see a wide variety of food. Corn, meat, fat, protein, everything’s in there.
• If a poor person was hungry enough, they wouldn’t have a dog just sitting around, they would have it run deer [sic], or they would eat it. Poor people also grow there [sic] own food, hunt there [sic] own game. They are way more independent then rich people, because if the world was to go to heck…they wouldn’t suffer that much!

Post #2: How did the urban legend that poor elderly people eat dog food get started?

• unfortunately, it is not an urban legend. it is quite true.
• WHEN MY GRANDPA GOT COUGHT [sic] POURING MILK ON HIS KIBBLES AND BITS FOR BREAKFAST THEN CHUCK NORRIS KICKED HIM IN THE FACE
• Elderly people on fixed incomes started buying dog and cat food because it was cheaper, and has the same if not better nutritional value. However, that was quite some time ago – pet food is not always cheaper than cheap canned foods. This was a big concern and story in many places in the late 70s, early 80s.

So there’s some disagreement on whether there has ever been a widespread trend of eating dog food out of desperation. It’s worth noting, however, that the only response that didn’t appear to be written by an idiot supports the idea. But does eating dog food actually make economic sense? Does it provide all the basic nutrients needed by the human body? Find out next time, in a very special episode of…Matt Meshulam pathologically overthinks everything.

Introduction

Last night I happened upon a carton of colored Goldfish-brand cracker snacks at a party. So naturally, I set up a single-blind ABX test to determine if it is possible to distinguish the colored fish from their “naturally” colored friends simply by taste. Color-taste correlation studies have been attempted before-I’m sure everybody remembers the controversial Von Braun M&M experiments in the 1960s-but the results have always been tainted by the fact that most of these studies were bankrolled by Big Dye.

Method

For my study I employed the assistance of a single volunteer, who received no compensation other than approximately fifteen Goldfish crackers. Standard ABX test procedures were followed: the subject was asked to close her eyes, was given a Goldfish known to be uncolored (the “A” sample), and then given a colored Goldfish (“B” sample). The test administrator informed the subject of the specific color of the colored Goldfish, but the main focus of the study was to simply distinguish colored from uncolored samples. Determining a specific color by taste was outside the scope of this study.

Once the A/B calibration was completed, a pseudorandom sequence of Goldfish crackers was given to the test subject one-by-one. The subject was instructed to make a best effort at determining whether the Goldfish was colored. After each trial, the subject was informed of the true coloration of the cracker. This breaks from typical ABX protocol, but was deemed acceptable since an A/B calibration was not performed before each individual trial.

Approximately 15 trials were performed. Detailed data was not recorded since a pen and paper was not readily available. Even without detailed data analysis, it was clear that the test subject was unable to correctly guess the color of the Goldfish more than 50% of the time. This percentage, of course, indicates zero correlation between the color of the Fish and the taste.

Conclusions

Unrelated: Poutine sure has been coming up in conversation a lot lately, hasn’t it?