I collect board games, and probably, I collect too many of them. Each game is different and has its own charm and value. Some are fun for large groups, others work best when you play them one-on-one. Sometimes what draws me to a game is a great theme and sometimes it’s a novel mechanic. Regardless, there is something about all of them that makes me want them.
Sadly, I only have so much time in a day and only so much of that can be devoted to playing games. My shelves are getting full with great games that I’ve never played, some even still have the shrink-wrap on them, never opened and likely never will be. To be honest, I have a problem. I have a wealth of entertainment value but no real way to appreciate it. When I buy games, I ignore the fact that I already have 20 games sitting ready to be played that haven’t been touched. And, among those games I have played, rarely have I truly mined their depths, rather I end up playing them once or twice and move on before getting all their worth.
Yes, I have a problem, and I call that problem “Big Board Games.” I have games of great variety and volume, acquired at high velocity and all having (at least to me) great value. All of you reading this probably think what a waste of money my games might be, or, at best what an interesting quirk for a statistician. But if I were to have told the same story in the context of data, we would think of this as a source of universal pride and envy.
It is a rare case that we stumble upon big data by accident or surprise. It is a deliberate process by which we desire and acquire new data sources. Some sources come from changes in technology allowing us to capture more detailed or more frequent data, perhaps a new type of genetic assay. Some sources come from vendors who sell EMR, claims or other rich sources of data. Sometimes we get lucky and find new ways to analyze previously inaccessible data; for instance text mining of EMR notes from a previously purchased source. But all of it we knowingly capture and all of it comes at a cost to us.
Granted, I’m a bit long-winded in getting to the point, but the point is a valid one. The acquisition of big data in and of itself isn’t a point of pride. Rather, the presence of big data indicates a data acquisition strategy that may be out of pace with the corresponding analytic support capability. We will all face times where we step into the realm of big data; this is a necessary product of growth and exploration. But those periods should be short-lived and carefully evaluated. We should be asking ourselves what was the factor that pushed us into a state of big data, and is the corresponding value worth the cost of acquisition – and more importantly, is it worth the cost of utilization?
When I store a new batch of board games on my shelves at home, I find I often have an introspective moment. As I rearrange and push aside old games I’m forced to ask myself whether these new games truly add value to my collection and enjoyment. Through acquisition of games, even great games, my ability to play isn’t increased, yet the potential pool for demand is growing constantly. Some games must be diminished, others forgotten almost completely; and yet, while I’m aware of all of this potential waste and lost value, there’s always one other great game that I just have to get next.