gebloom.com

The first principle is that you must not fool yourself – and you are the easiest person to fool.
—Richard Feynman

Sid Meier, a famous computer game designer,¹ once explained that, because of his development experience, when he plays a computer game, he can’t help but see the algorithms—the underlying software processes that create the action. Experts in a domain can envision the below-the-surface complexity of how determinants form the system in action, as when your doctor presses a stethoscope to your chest and translates the heartbeat to the status of your cardiovascular system. Domain experts don’t have Superman’s x-ray vision, but they make do with training and experience.

A computer game, as Sid Meier might explain, is assembled with data and with formulas to define the relationships among the data. An animation layer acts as public relations for the hidden software processes. With either a game controller or a keyboard and mouse, the game player interacts with the animation layer and, if the player isn’t the all-knowing Sid Meier, suspends disbelief to regard the continuous animated updates as a simulated world.

The corporate strategist, costumed in dull business attire, rather than the gamer’s hoodie, is using the spreadsheet on the screen, which is also assembled with data and with formulas to define the relationships among the data, to engage in imaginary play similar to that of the gamer. But unlike game players, corporate strategists interact with their imaginary worlds directly by modifying the data and formulas. These strategists call their spreadsheets “projections” to make them sound businesslike and rational but they’re just stories, with numbers as the characters and the formulas to define the relationships among the characters. A possible plot might be, “What if we buy another company to incorporate their product or their intellectual property?” Or, “What if we reduce the price of our widget to expand our potential market?” The spreadsheet and game are distinct approaches to a similar goal: to play in a fantasy world without the risks found in the concrete world.²

The spreadsheet creates a simulation that you explicitly manipulate (by changing data and formulas in the cells) to look at possible outcomes. The game, in contrast, is a pre-made simulation, in which through the manipulation of the interface, you experiment with possible outcomes that are implicit in the design of the data and formulas.

In short, with a spreadsheet, you explore and experiment by changing the parts from within the skin of the virtual world. In a game, you explore and experiment by changing the whole from outside the skin of the virtual world.

The parts and wholes of spreadsheets and games belong to abstraction, but it works similarly in the concrete world. An engineer designs the parts of a car, then the parts are assembled to test how the whole performs on the road. A football coach selects, conditions, and trains players, plugs them into defined positions, and tests how the whole performs on game days.

Inference in research

What the spreadsheet and the game have in common is there are no counterparts for viewing human behavior. Neither the view from inside (studying the parts of the nervous system) nor the view from the outside (observing and experimenting with behavior) expose the determinants of human behavior. No one has lived inside another’s mind. Research in human behavior depends on cognitive interpretation, which disregards the level of inference that’s required to call the results, “conclusions.”