Sometimes while in the depths of what would normally be a friendly conversation, you arrive at a point where in order to proceed, you must define the terms you’re working with, so that all parties involved can be on the same page. Oftentimes this is as simple as a quick statement of definitions and then everyone can agree. Other times it involves spirited debate about the criteria by which you make definitions. It is in this latter case that my roommate Adam and I were pressed to graph the relations and divisions among different varieties of snack mix.
To begin with we were primarily concerned with how you find the dividing line between what makes a trail mix, and what makes a party mix. Could there be overlap where a given snack mix is both party mix and trail mix? In order to determine just that we have to take an overview of what is available snack mixes, and plot them on an appropriate spectrum.
Both of us being long time consumers of various kinds of snack mixes, Adam and I proceeded to name our axes based on our initial understanding of the world of snack mixes.
Initially we had named the lower end of the Y axis “chip content” but found that to be too narrow in scope. So we went ahead with our established axes with X being a scale from savory to sweet, and Y being a scale between starch content and nut content. With our axes defined we were ready to begin plotting as wide an array as we could manage of various kinds of snack mix.
We decided to try and plot as many different kinds of snack mixes as possible, including mixes that we ourselves had not experienced, in order to get a better sense of the overall world of snack mixes. With a good number of snack mixes now plotted on our graph, we had now a great sense of how different mixes group, what their various regions are, and if and how those regions intersect and overlap.
One interaction that we were particularly interested in was between Planter’s Spicy Nut Mix (please ignore that the picture includes “cajun sticks”), and Imperial Nuts Sweet and Spicy Bar Mix. The makeup of a standard bar nut mix, almost resembles a true trail mix, whereas something about a spicy nut mix on its own seems to take it out of the realm of trail mix.
We also felt that certain relationships required new rules and exclusions to be made. The first is that the inclusion of almonds in a nut mix that is not made up exclusively of nuts, will often be a trail mix. And like the common misconception that peanuts are nuts, peanuts do not always contribute to “nut content” on our graph. In many cases, unless peanuts are included as a primary individual component, they can in a sense be drowned out by more overpowering components in a mix. So while the presence of peanuts may often be thought of as a foundational component of trail mix, their presence alone does not necessarily influence a snack mix’s position on the nut/starch scale. With some more rules defined and our plotted snacks in place, we were ready to start drawing conclusions and defining regions on our graph.
As you can see more than the expected “trail mix” and “party mix” regions became apparent. We have defined four distinct regions in this graph, those being: dessert mix, trail mix, crunch mix, and party mix. The most surprising component for us was the distance between true party mixes and trail mixes, which is taken up on one side by what we have called “crunch mixes” which do share some overlap with party mix, but is largely comprised of snack mixes somewhat more experimental than those that exist in true party mixes, but too savory to be trail mix. On the other side is somewhat of a no man’s land of ill-defined sweet-and-savory chip related mixes and granola, this region suffers from severe homogeneity and is far less exciting than the named regions and thus continues without a name. Areas of future exploration include the crossover area between dessert mix and trail mix, how the graph changes when a third axis relating to fruit content is introduced, and how snack mixes on the whole fit into the larger snack food continuum. Our findings are somewhat subjective, and if you have anything to add or that you would like to further discuss, please reach out to me via email or twitter.