Lorrie and I have lately been taking part in a protracted conversation with some colleagues about the virtues of experimentation. I am firmly of the opinion that it is crucial to investigate new things in order to remain productive – despite whatever other responsibilities are placed upon us.
An example is Stack View. This is one of two experimental user interfaces I’ve come across lately that vary the traditional process of browsing digital content (the other is Library Observatory, which I hope to write about soon).
A product of Harvard’s Library Innovation Lab, Stack View visualizes digital collections similar to the way visitors browse books on a shelf. You can see more about the project this introductory video:
Lorrie pointed me toward Stack View a few weeks ago, but at the time none of my projects offered a chance to evaluate it in context.
As has become my custom in these situations, I turned to soccer. Why not? Soccer has been a consistent playground for me for over a decade, in part because it offers me a platform for self-driven investigation. There are no expectations other than those I put on myself, I have amassed a large amount of data about players over the years, and there is a community of fellow fans who share my interests and are usually willing to provide feedback.
So, Lorrie and I brainstormed about “if soccer players were books, what would they look like?” This may be an odd question, but from a data perspective there are similarities. “The story of a player” is certainly familiar to sports fans, or fans of biographies in general. After tinkering for half an hour or so, I had succeeded in putting something simple together:
In terms of this experiment, the challenge was to find parameters in the soccer world that could correlate into a library-style experience. StackView uses three numeric parameters to determine a record’s visualization: page count, binder size, and shelf rank. Page count and Binder size determine a book’s size, while shelf rank (essentially how often it is checked out) influences the binding color.
These three values occur naturally with books, and their application in StackView is straightforward. With soccer, however, we had to discuss a variety of parameters before being able to translate the experience. Two of the three parameters came easily, but the third is one we still struggle with.
The length of a player’s career – measured in terms of minutes played – determines the thickness of a player’s “book”. This makes sense – a player who appears in only a handful of games would have a much shorter “story” than a player who is a mainstay over several years.
Shelf rank, in contrast to page count, is a qualitative parameter – it can be seen as a rough proxy for quality, or the value of a book to its library community. In this experiment, binding color is derived from the amount of impact a player has: goals and assists, yellow and red cards earned. The notion here being that a player who anonymously plays a number of games should be treated differently than a player who finds a way to score goals, or otherwise influence the game. This is an imperfect proxy, but it serves well enough for an investigation (I’ve written in soccer circles about some potentially better proxies, but that discussion isn’t really relevant here).
The final parameter, binding height, currently lacks an obvious analogue. We’ve used player age for now, with older players having a longer binding. This is imperfect – particularly when players appear over several years – and we continue to evaluate better alternatives.
Beyond any specific details, however, I’m glad to have had the chance to pursue short-term experiments like this (from start to finish, this experiment took less than an hour to implement – starting while I waited for the bus one morning and wrapping up over a lunch) discussion). The projects I undertake at work vary widely, and frequently involve technologies or problems I’ve never encountered before (this is especially true as I adapt to a new job, in a new industry, with new systems). Unless I am willing to say that I have all the skills I will _ever_ need, I must be always looking for new tools, new techniques, and new experiences. Otherwise I run the risk of having an inadequate base of experience for the next project to land on my desk.
This is not to say, by the way, that open-ended experimentation is the most important activity for me to pursue. My job is not to experiment, but to contribute to my organization by identifying and resolving the problems that emerge naturally (and continually). Many times those problems may require responses with which I am well versed.
Such is not always the case however – and I do myself and my organization a disservice if I become complacent, assuming that I have all the tools and experiences I will ever need.