After posting graphs and cold data (quite ilustrative, I believe), and the discussion it has generated (people, why don’t you use the “comment” instead all the other unstructured methods you are using?), please let me write a caveat about graphs and cold data.

In my high-tech gym, you have the option to have a lot of data collected, for your own, private and personal use. It seems like a great idea at first. For example, I can access via a secure web site real time stats of my workouts, such as the “fitness balance” (which shows my emphasis in weight lifting, and then swimming -some data greyed out for privacy purposes-):

Sounds great, doesn’t it? Not that fast.

The following graph (Workout Log) allows me to see how many times I have worked out. Since I tend to go to the gym everyday (but it is not always possible), an average of 4 times per week sounds reasonable, but what is that max. 8 times per week number? Why is there such a dip in mid February?

Data often needs to be contextualized, otherwise we might end up with the wrong conclussion (those 8 workouts per week happened to be visits to the gym to do a personal assesment and training routine design, added to my regular workouts; and the dip… just a long trip!).

But even worse things can happen:

In this case, an obviously strange abherration is showing in the graph. Somehting to be concerned about? Not at all: the scale of the axis make a slight variation (less than 1%, less than a pound) seem like a huge shift. And variables such as measurement thresholds, electronic glitches, etc must be taken into account when considering the validity and presentation of that data.

Let’s just keep in mind: however great quantification and visualization tools are (and I do like them a lot, and believe they can be very beneficial to the way we make decissions and understand the world and ourselves) they must be used with care. After all, this following map might show all the places I have traveled to… but it can not tell you about the experiences lived there…