Guys I've been thinking what to do next with Datamine Network. We've now been running minting for over 4 years, we've got tons of realtime data, charts and all the possible metrics we can think of.
The validators are minting and we've got a decentralized dashboard running. These were my "long term vision" at the start of Datamine Network.
I am very happy with the amount of analytics and maybe it's time to work on something new. There is one area of the project that I feel like we haven't really finished: The forecasting tools.
We have the trends engine done but the frontend is not how I imagine it in my head.
I've been thinking a lot about he we can display trends in a meaningful yet creative way. This is something I've been fighting over in my head for the last 2 years.
How can we show forecasting in a way where 2D is not enough. We have to go 3D but not in the sense you are thinking of.
If a forecasted line is "X and Y" coordinates then we can draw it on a chart.
The problem is that this is not enough data to explain the trends.
Trends have many properties:
Genesis: "When was this trend created?
Age: "How long was the 'retest' bounce?"
Duration: "How long after the 'retest' bounce are we?"
There are all of these foreign concepts that only make sense in my head. If I say "Hey this trend is 15 days away from genesis" you would be very confused.
So how can we fix all of these complex ideas? Well the answer is, let's draw them as 2D characters and convert all the properties into stats.
We represent each trend as a unique character based on the stats. That way you can look at any chart of let's say S&P 500 and see a little battlefield happening.
Speaking of data I am thinking of open sourcing this entire concept. Data and rulesets will be fully customizable and you can import any data set and run the game on your browser.
This means we can now decentralize one more tab of Datamine Network: Forecasting. 