There are many aspects to writing on the Web which I believe are underutilized, but there is one genre in particular which is especially puzzling, because despite its relative ease of deployment, you only really see it in two places: ultra high-budget digital publications, and esoteric blogs. Bret Victor—a master of the genre—calls it model-driven debate, and I—a much more modestly-skilled practitioner—am inclined to frame it as computational rhetoric.

The gist is that you aren't merely creating an interactive experience, but that you are genuinely using computation as a first-order vehicle to support your argument. The user—the reader—is given handles to manipulate, usually quantities, which are fed into some kind of computation that generates some kind of result, which varies with the input. The argument, for example for a piece of public policy, holds up to the extent that the reader views the inputs—which they control—as plausible. The computation itself can be a piece of spreadsheet math, like most of mine, or it can be something considerably more elaborate.

The ROI of a Solved Problem ()
This one is about the statistical properties of creative work. I consider this essay to be proto-computational rhetoric because while I use computation to make my argument, I nevertheless fail to cede the computation to you.
Reality Check ()
This is my first, very crude attempt at bona fide computational rhetoric which came about quite naturally when I tried to air my skepticism about the carrying capacity of the North American economy for the professions under the UX umbrella.
The Value of Tailored Information Infrastructure ()
Here I get only slightly more sophisticated than a spreadsheet to consider the value of build versus buy.
The HURRDURR Games ()
Here, I use once again what is basically spreadsheet math to argue that so-called hack-a-thons are a scam.
Cost Structure of a Construction Project ()

In a forthcoming article comparing and contrasting software development with other forms of projects, I remind the reader that most of the cost of building construction—minus the land—is the construction itself:

Like Bringing a Gantt Chart to a Casino ()
This is a very simple example, for an essay on productivity, uses two small inputs embedded directly into two different paragraphs to convey nonlinear relationships between inputs and outputs.
Simulation Time! ()

This slightly beefier graphical simulation imagines a fixed allocation of practitioner-hours against a set of projects that vary in size roughly according to a power law. The idea is to show how such a distribution might play out on the calendar.

I have more examples of computational rhetoric in various stages of disarray. I will catalogue them here when they are ready.