The coolest advisor, never seen before!!!! - page 12

 
ufkef:
Mathemat:
ufkef, have a look over here: 'What do the numbers in the EA test report mean' (see Expected Payoff).



And I'll tell you this about the clever link, I myself write formulas that are used here!

I don't remember who said, "If you're so smart, why are you so poor?"

And as for buying - not buying. What to buy???? Where's the result???? Buy on the basis of the tester's state (few people are interested, they have not seen enough of this)? five trades from the demo and have nothing to do with the tester? Anyone here will buy a profitable Expert Advisor for a ridiculous 500 coons, I'll be the first, but profitable, SHOW THAT PROFIT!

If you want to sell - lock a demo account for at least a week (the EA should be up 24 hours a day), post the full statement of your allegedly tripled account (and not some pieces of something incomprehensible), show something ... But there's nothing, just verbiage and narcissism.
 
Yes, the statement from the demo account for one day should certainly impress a potential buyer :)
You never ran the demo on historical data and did not show the profit curve. Or is it complicated and there is no time for such nonsense?
 


First of all, you should have read Sergey Kovalev's article 'My First 'Grail ' and found out the difference between trading on a real and on a demo account, especially in a tester, and what are the requirements to an Expert Advisor and its testing in order to get reliable results.
There is nothing to gain in the contest either, if the organizers introduce slippage and requotes in the server, as they promised.
The only place where you can earn some money is micro forex and that is until they switch from automatic to manual quoting or ban EA trading altogether.

 
Valmars:


There is no light on the competition either, if the organisers introduce slippage and requotes into the server provisioning as promised.

At the Automated Trading Championship 2006 there were requotes and serious slippages on the news.
 

About two years ago in the Alpari demo competition, a contestant with a long stretch won the prize. Perhaps Rosh remembers the case. The contestant traded around the clock and sometimes managed to open and close 2 or 3 positions per minute. It is clear that the machine was working, but this is impossible in real life.
I would not want something like that to happen in the EA Championship.

 
elritmo:
Yes, the statement from the demo account for one day should certainly impress a potential buyer :)
You never ran the demo on historical data and did not show the profit curve. Or is it complicated and there is no time for such nonsense?

Where do you get the brokerage company's historical data?
 
Mathemat:
Mathemat:
ufkef, have a look over here: 'What do the numbers in the EA test report mean' (see Expected Payoff).
Have you looked at it? The formula really does show the average payoff per trade, if you know anything about tervers. It can be simplified. It's supposed to be so complicated, but in fact it easily boils down to this:
Expected Payoff = (GrossProfit - GrossLoss)/TotalTrades
Does it make more sense now? When you understand the meaning of this highly theoretical formula, try to work out how much you don't know yet...

And then there is such a thing as Sharpe Ratio, which shows how much this payoff expectation exceeds random market fluctuations. In fact, it is a ratio of Expected Payoff when testing by 0.1 lot to the standard deviation of the price (more exactly, Return that is equal to the difference between two neighboring prices). If the Sharpe Ratio is significantly less than 1, like in your case, the results of your strategy's testing cannot be relied upon. A reasonable value ofSharpe Ratio is several units, or, roughly speaking, several sigmas. And taking into account that the distribution of price "jumps"(Return) is not normal, but fractal in fact, and three sigmas cover much less than 99.7% of cases, you will realize that for the results of testing of your strategy to be believed as much as the "three sigma rule" with a normal distribution, Sharp Ratio should be significantly greater than 3 (in fact, it is about 4.25).

P.S. On the HC quotes, the Return (Close) for minutiae is about 2 pips. That is, you should have the notorious expectation equal to approximately 8-10 pips. It is on one-minute periods. Respectively, on 30-minute Skewers, it may be around 45-50 pips, and on 4-hour Skewers, it may be 130 pips.
.

If I had not had slippage I would have got 20 pips, but the price wouldn't give me the price. That's why I got 30 !!!! Like this!!!!
But the probability of slippage in both directions is 50/50.
 
ufkef:
elritmo:
Yes, the statement from the demo account for one day should certainly impress a potential buyer :)
You never ran the demo on historical data and did not show the profit curve. Or it is complicated and there is no time for such nonsense?

Where can we find historical data of brokerage companies?
You may check it here:
http://www.alpari-idc.ru/ru/quote-archives/
 
ufkef:
elritmo:
Yes, the statement from one day demo account should certainly impress a potential buyer :)
You still haven't run it through the historical data of the DC and haven't shown the profit curve. Or is it complicated and there is no time for such nonsense?

Where do you get the historical data of brokerage companies?

You may open a demo account in any brokerage company, open charts with needed timeframes and press home to download as much data as possible until the chart does not scroll deep into the history. And then in the tester run with a tick "recalculate". By the way, all files with extension .hst should be deleted from history folder in MetaTrader (when terminal is closed). It would be interesting to see how your profitability will change.
 
Figar0:
ufkef:
elritmo:
Yes, the statement from the demo account for one day should certainly impress a potential buyer :)
You never ran the demo on historical data and did not show the profit curve. Or it is complicated and there is no time for such nonsense?

Where do you get the historical data of the brokerage company?
You can e.g. here:
http://www.alpari-idc.ru/ru/quote-archives/

Or have they also found a big archive somewhere and converted it into MT format?
How do I recalculate them in other timeframes? Is there a good tool for that to get candlesticks of higher timeframes? It seems to be a different alignment if I just take the start of a minute in the history and count periods of M5, M15, etc. from the very first M1, they are aligned differently because of the weekend at the end and beginning of the week - I have to deal with it ... :(
Reason: