Loopy: a tool for thinking in systems(ncase.me)
ncase.me
Loopy: a tool for thinking in systems
https://ncase.me/loopy/
26 comments
Google Sankey diagrams Budgets. There are lots of tools - eg - https://github.com/ricklupton/floweaver, sankeymatic.com etc
Not my area but complications in visualizing flows dramatically increase when the flows criss cross. Even small budgets start looking like chaotic weaving.
Which I think is why we still have buildings full of accountants and analysts to make sense of what is going on even though lots of people have access to the data.
Not my area but complications in visualizing flows dramatically increase when the flows criss cross. Even small budgets start looking like chaotic weaving.
Which I think is why we still have buildings full of accountants and analysts to make sense of what is going on even though lots of people have access to the data.
There is a game about that Democracy 3 (https://www.positech.co.uk/democracy3/).
I think it should be a mandatory exercise for any politician - and voter.
I think it should be a mandatory exercise for any politician - and voter.
This is a good idea. Why do you suppose they don't do it?
Based on my own experience at local government meetings, I don’t think the knowledge is all in one place.
The meeting I’m thinking of was the city council advocating extending a tax that was due to expire. They had a presentation deck full of the things they planned to use the money for. That’s all well and good, but when I raised my hand and asked what proportion of incoming revenue for the public works department and the overall city budget the tax currently made up... crickets.
No one there knew exactly how much public works was allocated, because the director of public works. They had the overall city budget numbers handy, but no one there had a breakdown by source, and no one knew where to find that information.
The meeting I’m thinking of was the city council advocating extending a tax that was due to expire. They had a presentation deck full of the things they planned to use the money for. That’s all well and good, but when I raised my hand and asked what proportion of incoming revenue for the public works department and the overall city budget the tax currently made up... crickets.
No one there knew exactly how much public works was allocated, because the director of public works. They had the overall city budget numbers handy, but no one there had a breakdown by source, and no one knew where to find that information.
I see. So they didn't even have the tools to gather the data in the first place it sounds like, let alone figure out how to put together a simulation for it.
There's a much much more advanced version of this tool used for systems design in game design called Machinations. You can model really complex economies with it.
Now that I've left the games industry I still think about this tool often and how it might be useful for modeling load on a backend. Kind of waiting for an opportunity to try it out.
https://machinations.io/
Now that I've left the games industry I still think about this tool often and how it might be useful for modeling load on a backend. Kind of waiting for an opportunity to try it out.
https://machinations.io/
That's really cool!
I see the value for game design, but also for lots of other system simulation (modelling load, as you said, among other things). I only tinker with game dev, but I can see myself using this tool on other projects too.
Thanks for sharing.
I see the value for game design, but also for lots of other system simulation (modelling load, as you said, among other things). I only tinker with game dev, but I can see myself using this tool on other projects too.
Thanks for sharing.
It says it's in open beta. Checking the FAQ indicates that a desktop version used to exist. Which version did you use?
This is indeed an interesting tool, thanks for sharing.
This is indeed an interesting tool, thanks for sharing.
I had used it back when it was an Adobe Air program. Maybe the death of Flash contributed to the decision to build a web version instead; but I also think the Google Docs metaphor is a much better fit for the tool commercially. I hope they do well.
It's a nice idea, but it's too simple for most systems simulations. It is, however, a splendid interface for building simulations. I'd hope the author hooks it up to a simulation framework at some point, exposing a much wider range of tools.
(Or builds their own framework, IDK. I just want to build more complex systems by drawing on a canvas :)
(Or builds their own framework, IDK. I just want to build more complex systems by drawing on a canvas :)
Anyone got a chance to try the recently launched Hash.ai?
From their announcements it seemed like a tool aiming to more complex scenarios.
I've messed around with it a bit, and it's actually pretty great for agent based modelling & Monte Carlo simulations. It does have nice visualizations, but not so much geared toward representing complex relationships, but rather showing realtime/step-by-step representations of simple relationships. But, I have only played around with it for a bit - I may very well be grossly underselling it.
It's a very simplistic tool, but it is great to help visualise and illustrate some systems. It helped me demonstrate the dynamic between technical complexity, delivery pressure and velocity (then add refactoring) https://medium.com/pageup-tech/the-system-of-technical-compl...
I've been also thinking about systems like this, but wouldn't you need constraints and counting?
I mean let's say you modeled human behavior based on basic economic principles, then you'd have constrains/rules like:
if money < x, stay home (and you wouldn't meet new people and have an impact)
if money < y, walk instead of taking a bus (and you would get only so far, your reach is small)
if money < z, have no energy to do new things (your behavior would stay the same)
if energy < c, die
edit: adjustments
if money < x, stay home (and you wouldn't meet new people and have an impact)
if money < y, walk instead of taking a bus (and you would get only so far, your reach is small)
if money < z, have no energy to do new things (your behavior would stay the same)
if energy < c, die
edit: adjustments
There’s also Minsky. Looks to be a bit more comprehensive with an eye to economics.
https://sourceforge.net/projects/minsky/
https://sourceforge.net/projects/minsky/
I love loopy for the crazy simplicity, it's good just to illustrate the moving pieces. But when you want to model something, the next tool I'd turn to is Insight Maker
Someone ported https://en.wikipedia.org/wiki/World3 to it: https://insightmaker.com/insight/2531/My-World3
So you can do a lot with it. You can even do sensitivity analysis with it! The devs are super responsive, I made a feature request and they implemented it in a day.
Someone ported https://en.wikipedia.org/wiki/World3 to it: https://insightmaker.com/insight/2531/My-World3
So you can do a lot with it. You can even do sensitivity analysis with it! The devs are super responsive, I made a feature request and they implemented it in a day.
Loopy is pretty neat. I played around with modeling glucose, but I couldn't get the arrows to model glucose uptake quite the way I wanted. https://bit.ly/2CjjaO0
Not sure how we can debug it. It seems the simulation is sensitive to
1. When you start each circle
2. Length of the wires
It would be neat if actual numbers were shown and rates could be adjusted via a hard reset.
I was messing with a procrastination model in https://bit.ly/2ZPNXdo .. it seems if you are not feeling good and you procrastinate less, things improve over time. To test this, decrease the feel good and then decrease procrastination.
1. When you start each circle
2. Length of the wires
It would be neat if actual numbers were shown and rates could be adjusted via a hard reset.
I was messing with a procrastination model in https://bit.ly/2ZPNXdo .. it seems if you are not feeling good and you procrastinate less, things improve over time. To test this, decrease the feel good and then decrease procrastination.
My procrastination model looks a bit different
https://bit.ly/38Jto6M
Never give up vs give up is such a classic dilemma. I do think the positive aspect of procrastination is exploration, satisfying curiosity but on the flip side exploration can also increase anxiety. If we did not explore then we would be using the same fishing net to catch the same fish. All progress depends on procrastination.
I was playing with it to try to model a simple regenerative system from permaculture design -- the "Three Sisters" plant guild: https://preview.tinyurl.com/ydxqmmwd
The model is really off since I can't specify units or connection weights. I was hoping the tool would be adequate enough to use as a kind of visualization or a kind of sandbox/lab to illustrate different regenerative processes. Might still work.
The model is really off since I can't specify units or connection weights. I was hoping the tool would be adequate enough to use as a kind of visualization or a kind of sandbox/lab to illustrate different regenerative processes. Might still work.
Love it, always liked iThink to model systems but the price is going through the ceiling in the last decade.
The purpose would be to show that everything the gov't does is connected through incoming taxes and outgoing expenditures - and there are very tough choices when it comes to the direction of taxes.