The unhealthiest sandwich ever? Ask a numbers wonk

In the munch-inducing lull between election cycles, fivethirtyeight's Nate Silver turns his attention to the KFC Double Down:

I've created an index based on the amount of fat, sodium and cholesterol that the Double Down and a variety of comparable sandwiches contain as a portion of the USDA daily allowance. (In the fat category, saturated fats are counted double and trans-fats are counted triple.) The index is scaled such that the Original Recipe version of the sandwich receives a score of 1.00, a measure of gluttony that will hereafter be known as The Double Down (DD).** (source)

Based on the graphic above, the Double Down doesn't look all that bad, right? It's surprising how items you'd think must contain the most fat, based on sheer delicious deep-fried goodness, aren't as bad as some of the Chicken Club Diet Compromises. At least the Double Down isn't deceptive about being ludicrously fat-laden. But just wait until Nate divides by calories - the beauty of dataviz makes the Double Down flip from seemingly attractive to deviously destructive. It all depends on whether you think (fat+cholesterol+sodium)/calories, or (fat+sodium+cholesterol)+calories, is the more important fast food metric - and there is a valid physiological case to be made either way, depending on your diet and eating habits.

On some level, all of this analysis is silly, since it doesn't matter how you parse it: the fast food products in this graphic are not healthy - even by comparison with one another. But that doesn't stop earnest commenters on Silver's post from advocating everything from the Subway diet to vegetarianism. People: if someone is seriously using data visualization to justify eating a Double Down, I don't think they're receptive to your arguments about the ethics of meat consumption. I'm just saying.

Update: I had originally included an image from the blog post to illustrate Silver's great dataviz, but I was asked today by a representative of fivethirtyeight to take it down never to share any of their content even with attribution. I'll comment on this shortly, but it's very interesting: I almost never get such requests, and I certainly wasn't expecting one from fivethirtyeight - for several reasons. More later. . .

Update #2: The rep from 538 objects to how I described our interaction above, specifying that he/she was referring to graphic content specifically. I didn't get answers to the questions I asked about the policy, so I can't fill you in beyond that. I'll simply reiterate that the goal of this blog is to promote interdisciplinary discussion and highlight brilliant examples of sciart and dataviz, and if any artist or designer ever has an issue with their work appearing here, they are welcome to contact me and I'll remove it.

Update #3: Here's a different, angrier take on the fivethirtyeight Double Down post. I actually don't agree with it, although you may differ, so go check it out.

First, I doubt very much that Silver would argue the simplified definition of "bad" used in his post - based on fat, cholesterol, and sodium - is the best physiological approximation of "bad." However, those three values are values you can actually get for (all)? fast food chains' food, and they are strongly related to an unhealthy diet and obesity, and a data wonk's got to have something to work with. Second, I think that the Double Down post does make the point that the Double Down has fewer calories overall than a lot of the foods with comparative levels of fat, cholesterol, etc. Once that point is made, we can all argue until the sun comes up: we do NOT as biologists have consensus on exactly what causes different people, especially obese people, to feel "full", we can't generalize about how their metabolisms adapt to dietary input, and we really have no clue whether for such people, it's "better" to eat a gigantic carb-stuffed Chipotle burrito or two grease-drenched Double Downs. For me, the carb-stuffed burrito would make me (uncomfortably) fuller longer; for you, maybe the fat shot to your dopamine circuitry would put you into satiated nap-mode. Who knows? Reasonable people could come out either way, and I think that's the point Silver's post makes effectively, with the two graphics. Of course the most reasonable people of all aren't eating any of the stuff in the post regularly, especially if they're obese. As I said above, the methodology of comparing a bacon cheeseburger to a deep fried sandwich is kinda silly, so I think critiquing that methodology is . . . meta-silly. But if it gets people thinking about their diets, bring on the meta-silliness! w00t.

More like this

Wendy's for dinner it is.

I saw that Double-Down thing at a KFC just the other day and joked with the employee about how it sounded more like a comedian's take on the weird meals the fast food introduces to the American public. But real. Of course, I think the double down on its own is probably no worse than getting a huge bucket of fried chicken slathered in ranch dressing, which is one alternative.
A bad sandwich? Yes, but the other stuff on the menu eaten together isn't exactly healthy either. Tasty though.

Thanks for the link!

I'll say up front that I probably took Nate Silver's post a lot more seriously than (a) he did and (b) it was intended. (I've spent the weekend marking "numerical argument" papers, so I'm primed to be sensitive about this stuff right now.) What really set me off about his post was how carefully he constructed his metric to put the Double Down at the top of his chart, and how little he justified it ("fat, sodium, and cholesterol (bad)").

"Meta-silly" needs to go up on my brag board.