Hello, world.
As currently three of our robots are showing the same direction, I would like to take time and define several ideas that we have as a baseline for our robot's analysis.
First, conceptually we need to define that all robot trading based on a data mining process, which includes a specific education for robot algorithms.
This is where it all starts, the system is concluding results of the robot's trading, and after a lot of iterations, specify which of the educational patterns were right (or got the most profit while keeping a small drawdown).
For Bitcoin, the most certain case scenario for any trader would be a way to understand where do we go on upscaled state, as far as we can define two market conditions: when it's in a trend, or staying in a flat.
Trading on an upscaled level means having more profit in a single order, instead of taking advantage of small movements; and in the same time, most of the critical robot tasks is minimizing a stoploss to be winning by expected value.
Functionally, all of that means a robot should find right factors to see there would be a high peak after the flat, and they will be statistically recognized on a scale that a quantity of profits is going to beat all the loss.
When you understand those points, you can spend time on defining a strategy.
BTCUSD is a perfect model of emotional trading, as it has many retail investors. In contradiction to bot-dominated markets, it has a very small-ranged flat, and very high-ranged emissions, so to have success as an upscaled robot on such exchange, your idea is to predict those emissions and try to catch large sharks.
Currently, three of our robots utilizing different statistical approach, decided to go short with three different sets, which you can see on the graph, or follow the previous analysis.
I would be glad to answer your questions about data mining for Crypto or take a look at your ideas about the algorithms.
Let me know what do you think about the future of the market, will it stay such emotional?
Have a good sailing!
As currently three of our robots are showing the same direction, I would like to take time and define several ideas that we have as a baseline for our robot's analysis.
First, conceptually we need to define that all robot trading based on a data mining process, which includes a specific education for robot algorithms.
This is where it all starts, the system is concluding results of the robot's trading, and after a lot of iterations, specify which of the educational patterns were right (or got the most profit while keeping a small drawdown).
For Bitcoin, the most certain case scenario for any trader would be a way to understand where do we go on upscaled state, as far as we can define two market conditions: when it's in a trend, or staying in a flat.
Trading on an upscaled level means having more profit in a single order, instead of taking advantage of small movements; and in the same time, most of the critical robot tasks is minimizing a stoploss to be winning by expected value.
Functionally, all of that means a robot should find right factors to see there would be a high peak after the flat, and they will be statistically recognized on a scale that a quantity of profits is going to beat all the loss.
When you understand those points, you can spend time on defining a strategy.
BTCUSD is a perfect model of emotional trading, as it has many retail investors. In contradiction to bot-dominated markets, it has a very small-ranged flat, and very high-ranged emissions, so to have success as an upscaled robot on such exchange, your idea is to predict those emissions and try to catch large sharks.
Currently, three of our robots utilizing different statistical approach, decided to go short with three different sets, which you can see on the graph, or follow the previous analysis.
I would be glad to answer your questions about data mining for Crypto or take a look at your ideas about the algorithms.
Let me know what do you think about the future of the market, will it stay such emotional?
Have a good sailing!
www.mql5.com/en/signals/525998/