Thought everyone might like to see some key stats on ROSE AI performance! Obviously we don’t share this type of information too often since it is proprietary, but we are so excited we can’t hide it
The first thing to understand is that part of the magic of the Crowd Wisdom system is it tells us when to trade and when not to. Being a good trader/investor is about knowing when to stay out of the market. The basis of this is understanding risk:reward (R:R)
In the charts I am showing below you will see results from our Mk1 2019 AI algorithms (Rose runs several at any one time) and our more recent Mk2 AI - both on Bitcoin. The MK2 have been live for only a matter of months, and already are showing a strong improvement:
As you can see from the second image, Mk2 is in the market 20.1% of the time with an accuracy of 74.5%, up from 68.2% from MK1.
What’s also very encouraging is Rose is picking the big moves. We give each trade a ‘range score’ based on not only on being on side but how far the market moves. Here you can see MK2 is performing very well, especially when our expert financial team overlay technicals (TA):
While we can already see the system is working on Bitcoin, Gold and Nasdaq Ixic - it will only get stronger with the number and quality of predictions. As those two things happen; the rate of (machine) learning increases.
To date we have many hundred thousand data points from the Crowd. This is a relatively small number of data points for AI; by the end of 2020 we will have millions of predictions in the system.
Prediction quality is increasing also, here you can see the trend in increasing Super Predictors in the Pynk Crowd:
This final chart proves that Pynskters are ‘learning by doing’ in a risk-free environment. It’s very exciting for all involved as proves that we can win as one collective. Later this year we will release a new product that fosters this approach.
Hope that all makes sense, if any questions or insights let me know below.
Also, our lead quant is writing a more technical blog piece on ROSE AI, we will post next month