5 ms·

Visualizing good behavior

43 points by stevehaz 10y ago

JacksonGariety 10y ago
>Be the hero, save the world >We’re drowning in low-quality information Right. Self-promotion on Hacker News does not help.
- SpaceInvader 10y ago
  Good point :D
- stevehaz 10y ago
  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-information-cf9566d407a9 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.
TeMPOraL 10y ago
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?
- stevehaz 10y ago
  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.
  - dhimes 10y ago
    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.
    - stevehaz 10y ago
      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/ 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.
      - dhimes 10y ago
        So it scrolls by like an ECG chart? Which way do you read those?
        stevehaz 10y ago
        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/ 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.
  - pantalaimon 10y ago
    > 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!
metasean 10y ago
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/ http://taskwarrior.org/
- stevehaz 10y ago
  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.
  - metasean 10y ago
    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/ http://the-happy-button.myshopify.com/
    - 10y ago

43 points by stevehaz 10y ago

>Be the hero, save the world >We’re drowning in low-quality information Right. Self-promotion on Hacker News does not help.

SpaceInvader 10y ago
Good point :D
stevehaz 10y ago
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-information-cf9566d407a9 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.

TeMPOraL 10y ago

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?

stevehaz 10y ago
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.
- dhimes 10y ago
  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.
  - stevehaz 10y ago
    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/ 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.
    - dhimes 10y ago
      So it scrolls by like an ECG chart? Which way do you read those?
      - stevehaz 10y ago
        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/ 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.
- pantalaimon 10y ago
  > 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!

metasean 10y ago

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/ http://taskwarrior.org/

stevehaz 10y ago
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.
- metasean 10y ago
  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/ http://the-happy-button.myshopify.com/
  - 10y ago