Visualizing good behavior | Hacker News Reader

Visualizing good behavior

medium.com

43 pointsstevehaz16 comments

It may be useful to a wider range of potential users to have an api for task-specific information. Then the open source community could use Benome to display the results of other tools (e.g. `done` tasks in TaskWarrior[0]).

This would increase actual productivity (i.e. users don't have to mark an existing todo item as done in one tool _and_ click a button in Benome). It would also mean that the Benome is available where ever the user is, without Benome having to store or transmit task-specific data.

[0] http://taskwarrior.org/

You make a fine point, an API would definitely have a place. Right now because this is only a one-man side project, I'm focusing on behavior that isn't already digitized and wouldn't be without something like Benome.

My view is that the vast majority of goal-relevant behavior is not available to individuals in digital form and so it isn't visualized and there's no insight to be gained or value extracted. While data-aggregators or ad networks may have a lot of digitized behavior data, people themselves tend not to. I'm hoping that Benome (or the ideas behind it) can help improve that situation.

I agree. Visualization of goal-data is an important piece that isn't well covered at this point.

For users that want a simple, "get me up and running" solution, the Benome button approach to data collection could be very satisfying (in a Happy Button[0] way).

I just see the possibility for wider adoption and contributions if it were more modularized to separate the data input and the visualization.

There is definitely value in Benome's visualization. One of my personal goals is to use Taskwarrior. Because I'm trying to optimize my workflow –not add more overhead– I'm not going to add the Benome data input to my workchain just for the visualization. However, if I could simply connect my Taskwarrior data to Benome, I would absolutely use it (and it would probably reinforce my self-reflection and subsequent data input).

[0] http://the-happy-button.myshopify.com/

It sounds like you're saying the information might as well be able to come in the back door (via API) as well as the front door (via direct input like buttons), and let users choose their balance. That way early adopters or innovators can benefit without having to change their workflow, such as in your example.

I'll add it to my big list, though there's several prerequisites to take care of first.

RE: API and 'let users choose their balance' Exactly :)

RE: 'big list' Totally understand! It seems ever expanding todo lists are becoming a modern perpetual epidemic. Thinks to self, "If only we had good tools to make them easier to wrangle and learn from." ;)

Cool, though I have to ask - why the direction of all those graphs is backwards? I.e. the future is on the left, and past on the right?

Author here, thanks for your comment.

I think left-to-right works in this case because my target audience is the everyman, not only technical or highly-educated folks. In my experience, people care mostly about the meaning of the information and don't care much about graphs or data in themselves. In that sense, I think of this kind of graph as more of a word or a sentence than a traditional graph, and mostly we read words left-to-right.

I'm with TeMPOraL on this. Every kid learns graphing with the horizontal axis increasing to the right. Only the tech folks play with other styles very much.

Because this is backwards, I'm trying to figure out what I'm missing. Am I supposed to be thinking backwards in time? I get confused on what I'm trying to get out of it, decide I don't understand what you're talking about, and move on. Just my $0.02.

The less people have to relearn the easier it is to get them on board, all else being equal.

I'll have to do some A-B testing.

The key here is that I've integrated this particular type of graph as part of a larger interactive visualization or UI. Within that larger system, the graph is generated in real-time in direct response to your input about your own behavior.

I put up an interactive demo here: https://benome.ca/Intro/

So in the intended context, graph's meaning is more clear than it appears at first glance. But in any case, it's easy enough to switch a graph's direction.

So it scrolls by like an ECG chart? Which way do you read those?

I can't speak to ECG's, but rather like an animated time series except in the reverse direction.

As you say, the norm is right to left, highest priority (now) on the far right, like this: https://square.github.io/cubism/

It must be an acquired taste, or contextually dependent, because even though I do plenty of metrics visualization which is right-to-left, left-to-right for a high-level view of behavior feels fine to me.

> I think of this kind of graph as more of a word or a sentence than a traditional graph, and mostly we read words left-to-right.

But your graphs go right-to-left!

In these behavior-quality graphs, "now" is on the far left. In my experience, how I'm doing right now is the top priority because then I can make a good choice -- I can do something aligned with my goals -- and the graph will immediately thicken.

Being "now", it's always changing and thus everything on the graph - one's past behavior - gradually moves to the right.

Check it out here: https://benome.ca/Intro

>Be the hero, save the world >We’re drowning in low-quality information

Right. Self-promotion on Hacker News does not help.

Good point :D

I appreciate that you've taken a deeper look into my work. I'm doing it all out in the open, including open source, which I think many people on HN appreciate.

For others wondering what this comment is referring to, the article in question is https://medium.com/@SteveHazel/we-re-drowning-in-low-quality.... In it I lay out a lot of the concepts behind the graph-like visualization of human behavior that is the current topic.

Reply on news.ycombinator.com