The championship: how the leaders opened - 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
Moderation versus Statistics. Not too long ago I held a parallel Monkey Trading Championship 2007. The conditions are very simple: deal direction, takes and profits are determined randomly, lot size is determined automatically for selected stops and risk level (I used 30%). Among 600 monkeys there were enough "cool traders" who made much more profit than the winner of the last human championship.
Moral: the more participants there are, the greater the chance that the aggressive monkey will win. If the 2007 championship had 1200 participants - 600 humans and 600 monkeys - there is a 100% probability that the monkey would win. It is unknown what will be the proportion of humans and monkeys in 2008, but the future of tournaments is predetermined - with their increasing popularity, the number of participants of all types will grow, and only monkeys will win.
A side moral: no reports on the results of the championship can show who is a human and who is a lucky monkey, i.e. for a strategy purchase it is necessary to know how it works.
+1
IMHO, if the aim is to reduce the element of randomness in determining the winner, there are two ways:
1. increase the duration of the championship to at least half a year - probably not realistic for the companies implementing the project (a very troublesome business);
2. take into account other factors in determining the winner, like Alpari's contests for trading on real accounts;
If we want to have a radical solution, we should move the contest to a real account :) let's say a participant has to deposit 500-1000$ starting deposit so we can observe MM with 0.01 (10000) minimum lot.
I think it will weed out most of the "monkeys" at once.
Not too long ago I held a parallel Monkey Trading Championship 2007. The conditions are very simple: trade direction, take and profit are determined randomly, lot size is determined automatically for selected stops and risk level (I used 30%). Among the 600 monkeys there turned out to be a fair number of "cool traders" who scored much higher than the winner of the last human championship.
Oh, so I was just the luckiest monkey last year? :)
Timbo,
Can you tell me more about the methodology and results of the calculations?
What was the SL, TP used? How many trades? What is the final result of the best monkeys?
Oh, so I was just the luckiest monkey last year? :)
You weren't just the luckiest. You were incredibly lucky. So incredible that it's almost unrealistic that you were a monkey. You must be human after all.
Oh, so I was just the luckiest monkey last year? :)
timbo,
Can you elaborate on the methodology and results of the calculations?
What SL, TP were used? How many trades? What is the final result of the best monkeys?
Trust me, you are not obliged :)
There is a similarity between the winners of the Championship 2006 and 2007: the Expert Advisor consists of three identical EAs with slightly different parameters, and MM - the lot of each system depends on the outcome of positions - the largest lot of the most profitable system for N positions (the profit-loss relation is non-linear).
Such systems are able to work stably with small changes of volatility.
And Soviet monkeys are simple. It's where the poses were opened randomly. Search for it on the forum.
Oh, so I was just the luckiest monkey last year? :)
timbo,
Can you elaborate on the methodology and results of the calculations?
What SL, TP were used? How many trades? What was the final result of the best monkeys?
Not really. Hypothesis testing can't tell you're a monkey, but it can try to prove you're NOT a monkey. You got XXX. If the probability of getting the same amount by trading at random was less than 5%, then you could say that with 95% probability you could claim that you are not a monkey. However, this could not be proven. If you take a crowd of 600 programmers, excuse me, monkeys, then surely (100% probability) someone makes more than your XXX.
The methodology is very simple: EURUSD, constant presence in the market, only one trade with a maximum of 15 lots. Then you randomly choose an entry direction, and then select a stop and profit size (individually - in an interval from 10 to 100 pips, I don't remember exactly). The lot size is set automatically by compuster library on the basis of available money, stop size and risk level (I put 30%). This case has been run 600 times during the Championship period.
The result is almost like a human :-) I found these results in the archive
But this is the first version of the championship, I think there were stop=profit. Then I tweaked it and gave the monkeys more freedom. Two of them beat you here too though.
IMHO, if you want to reduce the element of chance in determining the winner, there are two ways:
1. to extend the duration of the championships to at least six months - probably not realistic for the project implementing companies (very troublesome business);.
Definitely. Increasing the number of deals leads to decreasing the random component.
In general, a radical way - transfer the contest for real :) say the initial deposit participant brings $ 500-1000. that it was possible to comply with the MM minimum lot 0.01 (10000).
I think this will immediately weed out most of the "monkeys".
Don't try to weed out the monkeys. You don't need 500 bucks either. Even a small entrance fee will dramatically change the psychology of participants. It has been described many times in the literature, the risk of even a small loss of one's own money dramatically reduces the level of investor adventurism. No one will open a lot for two depots at once.
There is an idea, which is captivating in its novelty. Obviously, the Metakvots have no reason to bother with participation fees, they have neither the desire nor the capability to do so. But there are organisations for whom this is a daily routine - charitable foundations. If you choose "Wildlife Fund" for example, of course, a bourgeois one, the potential participant transfers there a certain fee, and then the fund sends a list of "sponsors" to Metakvotam. And the risk level is lowered and nature is saved.
Definitely. Increasing the number of transactions leads to a decrease in the random component.
Don't try to cut out the monkeys. You don't need 500 bucks either. Even a small entry fee will dramatically change the psychology of the participants. It has been described many times in the literature, the risk of even a small loss of one's own money drastically reduces the adventurism of investors. No one will open a lot for two depots at once.
There is an idea that lures by its novelty. Obviously, the Metakvots have no reason to bother with participation fees, they have neither the desire nor the capability to do so. But there are organisations for whom this is a daily routine - charitable foundations. For example, they choose "Wildlife Fund" (of course, it is a Bourgeois fund), the potential participant transfers there a certain fee and then the fund sends Metakvotam a list of "sponsors". And the risk is lowered and nature is saved.
The MCs need to promote the product and there's no reason for them to cut monkeys at the championships.
Definitely. Increasing the number of transactions leads to a decrease in the random component.
Don't try to cut out the monkeys. You don't need 500 bucks either. Even a small entry fee will dramatically change the psychology of the participants. It has been described many times in the literature, the risk of even a small loss of one's own money drastically reduces the adventurism of investors. No one will open a lot for two depots at once.
There is an idea that lures by its novelty. Obviously, the Metakvots have no reason to bother with participation fees, they have neither the desire nor the capability to do so. But there are organisations for whom this is a daily routine - charitable foundations. If you choose "Wildlife Fund" for example, of course, a bourgeois one, the potential participant transfers there a certain fee, and then the fund sends a list of "sponsors" to Metakvotam. And the risk is lowered and nature is saved.
Yes, creative. If there really is someone willing to do this, I'd be happy to assist with contacts at the World Wildlife Fund, where I've had the great pleasure of working. The only thing left to do is to figure out how much prize money to raise for the winners.
"Monkey" will be easily distinguished by the maximum relative drawdown. This, by the way, is eloquently demonstrated by the "monkey" championship result given by timbo.
But when the "monkey" beats Better with a relative drawdown, then we can raise the question about Better's class identity hypothesis :)