Neat! I feel like I've seen something very similar to this somewhere else, but I can't remember where...
Also,
They will also get the same price history going back to July 26th, 2020
I wonder why this specific date in history?
It probably performs very poorly after that. I don't think this is a serious thing that can be relied on strictly with on chain data. If it were something that could be relied on then time period wouldn't be a factor. If it can't even go back one epoch then it's not something that can go forward one epoch either. Interest project, but anyone can take some model within a specific time period and find the algo to fit their assumptions.
reply
It's kind of like fee estimation. Even if you develop an estimation model that works really well, it'll stop working when everyone uses that model.
reply
LOL. This ^^
reply
Actually, I misunderstood. This doesn't try to predict the future at all. It just allows you to pretty accurately determine a given day's USD price (assuming the day is within their specified range).
I don't fully grok it, but most of the magic is in a historical "stencil" in the shape of a bell curve that matches output patterns to price points. It then lines up this stencil with a bell curve of the output patterns on a given day and voila.
What's cool about this is that if you need historical price data you can use this to deduce the price rather than rely on some external source. I guess it could break if onchain behavior for a day is anomalous and not accounted for by the model?
reply
Oh, thanks for explaining this. That is useful and cool.
reply
"but anyone can take some model within a specific time period and find the algo to fit their assumptions"
A bit dismissive lmao
reply
I heard more about it today at bitdevs and it sounds like it has reasonable assumptions based on common amount analysis but needs a solid baseline of historical price matching for it to work. And still can't be relied on in any meaningful way.
reply