I think the reason why computers and companies always fail to predict big market crashes even though they happen about every 10 years is because of these failure to realize If you can make your script more self aware of this who knows. Good luck
That's a really interesting point here ! The kind we can talk days about and trust me, I love that
If I had to articulate an answer, I would start by telling you that for an alternative truth to be accepted, the majority must always start by rejecting it. This is the most basic cycle there is ! You'll hear me talk about cycles a lot. Why ? Because this is the way energy moves in the universe. It's everywhere : from day/light to left/right political structure, from the tides to human emotions... Everywhere !
Why models, and particularly from genius people, trying to mimic the financial markets all failed (remember Long-Term Capital Management) ? Because they were trying to treat events linearly and not dynamically. Everything is caused by multiple influences so why keeping on using a +b = c from Aristotle to forecast complex things ? What we found is that there are cycles to the strength of an influence, the efficiency of an agent in others words. Imagine a giant dish of spaghetti where each spaghetto is a cycle treating one matter. Look how the whole thing is a product of infinite external incluences... Yet there is an order : the grouping (aggregate), or not, specify the end result.
Now take financial markets. What if I tell you we can forecast wars ? You would think and tell that we are crazy stupid. But picture this : a general knows a war is coming because he received top secret orders. He wants to benefit from this info. He starts speaking about it to his personal financial advisor. And voilà : the info is out and impacting gold market for example. Our dynamic modeling can pick up the shift in gold and correlate with others sectors and historical data and then... comes the war forecast !
All influences can be analyzed, sorted and classified but that takes a computer to do so
I completely agree in cycles. Something that even the oldest religions have a tenet of and don't get me wrong I don't want to get religious here but examine Hinduism and Buddhism for example, which are thousands of years old and practiced by billions as in actually practiced (I'm not Hindu or Buddhist so anyway). Everything is a cycle for them too, for the Hindus even in precise dates (we are currently not even a tiny fraction of the universe cycle, which has a known end), and comes the phrase also "all of this has happened before".
Sounds familiar in markets and even in Bitcoin right?
So as I was mentioning, huge crashes seem to happen regularly for over 100 years, almost every decade, and yet and yet supercomputers never predict it but as you say when majority rejects it, it happens. So couldn't a model predict precisely this trigger like "ookay way to many people think BTC will hit $100k so the crash is in 5 weeks".
Interesting and of course I wish I could see something like this model publicly