Proposal to the organisers of the Championship - page 3
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
And if the Surgeon hadn't revealed that he had a regular MACD? How would his victory have been treated then?
There will always be a share of pure luck winners. The other question is, "How can we tell if we have a monkey in front of us or a brainy robot?" And it's not about nerves, by and large, it's about the technology used, allowing you to really "sniff" the market.
And the rules can be changed very simply to weed out the monkeys. For example, the condition of a minimum number of trades per trading day, say, 5. Then the "rich long-livers on high timeframes who will die rich" will be eliminated. They will be eliminated.
Naturally, robots that trade very often, more often than most, are least likely to be thought of as monkeys. And so on, with such and similar criteria for sifting out leaders will give those very, clever robots.
If anyone who looks at the championships of previous years, and puts their hand on their heart, they will say - yes, most of the winners are just monkeys....
What do you mean? What pairs are you missing in the tester?
Maybe I have something wrongly configured, but my tester only has
Major:EURUSD,GBPUSD,AUDUSD,USDCHF,USDCAD,USDJPY
Crosses3:AUDNZD
Crosses:AUDCAD,AUDCHF,AUDJPY
But EURAUD,EURCHF,EURJPY,GBPJPY,GBPCHF,EURGBP do not
Yes. It's not the 'terminal developers and championship organisers' fault. It's "crooked configuration" on the user side. :) Look at the testing section in the Handbook.
...Think about it: if at least one of the symbols allowed in the Championship did not work in the tester, how could participants test their programs by such a symbol? :)
The rules will have to change if you want to popularise robot trading. I'm sick and tired of looking at insignificant MACD variations, which are only useful for the time of the Championship. In the meantime, it is better not to put them on the real market, because all too many of them know very well what their real value is.
I am in favour of having fewer monkeys. We need smart human EAs, not simia Bernoulli ("Bernoulli monkey" is my term: it's a robot that makes trades at random and on average leaks a spread per trade - for a monkey!).
I want ATS to have intelligence and I want ATS to become really interesting. I remember one (2007) when the neuro advisor Better won. Never mind what happened to it afterwards. The important thing was that it was a work of thought, not dumb MACD fitting without any confidence in the result. And people did seriously wonder - could it be that neurocellsets are really worth it?
None of the other ATCs (perhaps excluding the "ATS pipsetters", i.e. ATC 2008) caused such a surge of optimism.
We will have to come up with mechanisms which make people think, rather than stupidly clone known ideas without risking anything.(And what's that got to do with drawing styles (see green text tooltip)?)
You cannot just say: "People, think about it, and stupid advisors will be excluded. We'll have to come up with more sophisticated rules that exclude such advisers in advance and make the championship more interesting as a result.
In short, ATC should become a purveyor of money managers to potential investors, rather than a bunch of monkeys, by hook or by crook, eager for a slice of the prize hundred thousand. This will raise the profile of both ATS itself and the MT5 platform.
I don't believe Xupypr is capable of anything other than MACD!
P.S. I apologize for being harsh, if I went too far.And I see the problem as something else entirely. In the conditions of the competition of course.
The problem is in uploading the code to the organizers' server.
Suppose a manager got interested and decided to upload a formatted strategy portfolio as a robot. But he would not do so, because it is possible to build a profitable business on this robot. And here they can decompile it or use it for their own purposes.
In my opinion, we should come down to earth and allow participation from remote terminals.
Then the 100500 problems and issues associated with pre-testing would die off. Why, if I can afford Z IBM server, it doesn't matter what optimization the robot performs on each tick, as long as it shows results.
For me giving my robot to someone is like providing funds to control without seeing the article and without knowing the robot's handler.
As a result, it is much easier to use 20 strategies based on 2 MA with different parameters for trading one of them, and 19 only virtually considered signals and after each trade to choose the best of 20. But this is the way to nowhere. Although out of 400 monkeys at least 10 will randomly show a decent result.
To me, we should come down to earth and allow participation from remote terminals.
Then it would be a regular competition that all brokers run.
It would have been the same. The drawdown from 73 to 20, after which it went upwards - now permanently - spoils the impression of the robot greatly.
Then it would be a regular competition run by all the brokers.
And why are there thousands of participants at the LCI, all the brokers' efforts end up with a hundred and all are extremely uninspiring?
Maybe the developers + sponsors have a lot more motivation and opportunity to make a flashy show than the individual?
It is all the same. The drawdown from 73 to 20, after which he suddenly went upwards - this time for good - strongly spoils the impression of the robot.
That's what I'm talking about. Winning here is pure chance. First of all, the small number of trades proves it.
One of the conditions, along with the limit on the minimum number of trades, may be a limit on the maximum drawdown. Otherwise, you may win with only 1-2 trades during the whole Championship.
In general now with the current rules it is not profitable to trade often, as it dramatically reduces the probability of winning purely statistically.