Price movements can be predicted ! - page 14
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
No, not hurrah, timbo. The predictions of "Kiwi will close the week below its opening price" and "Kiwi will close the week above" are roughly equally likely and almost mutually exclusive (there is a third unlikely event - "will close the week exactly at its opening price", but that can be neglected or put into one of the two classes above). There you have it, the symmetrical coin.
Talking about "targets within one sigma" does not directly define our probability. There are no longer two mutually exclusive events there, but many.
If Sart were to give predictions of closing directions every week for each of the 6 pairs he tracks. The forecast for each could be considered independent. In a couple of months we will accumulate 48 forecasts, which can already be analyzed.
There was a "monkey trading championship" around here somewhere. Nothing special: 600 programmer-traders traded exactly like 600 monkeys opening strictly as they please.
Thanks, I've found and read it, it's very relevant, if you develop this thought, you'll get a simple criterion for screening "smart" Expert Advisors
We can create conditions when it does not matter in what direction and at what distance to place orders so that we obtain a positive EA within a certain period of time.
here I think how to compare the efficiency of an intelligent EA (IS) with a random one (RS)
if at 1000 runs RS gives 84% of profitable runs on a selected interval, what should the test look like in IS to be able to compare IS and RS correctly?
Maybe MS has a number of tests with different parameters, but who has seen an Expert Advisor in which during optimization most of the runs are positive, like in IS?
Or the time interval should be divided into small periods and the number of winning periods should be calculated as %?
I think it is relevant to EA writers because in general their results are worse than those of monkeys (imho of course. I know many would disagree internally)
No, blend, in general - they are not losing monkeys, but about the same (which is even worse). If they were losing, they could be converted and made profitable.
2 timbo: about opening "strictly as you go" you got it right. But this is only one aspect of evaluating the quality of the system in the context of a m.o. deal.
I don't understand why the fast tool went down?
and why you're a bastard who barked at respectable Fourier while you're taking a shit in your own thread?
Timbo is not a good man, but let's face it, he has a brain on his shoulders.
The prognosis did not come true and Sartt is excusing himself like in the army
- Why did you tear up the owl, you pig? Did it bother you? Did it bother you, I'm asking you
? Why did you break Professor Mechnikov?
- He, Filipp Filippovich, needs to be whipped at least once,
Zina said indignantly, - otherwise he'll get completely spoiled. Look,
what he has done with your galoshes.
Listen, Sabluk, there are grown-ups here. Look, they'll kick your ears in for hooliganism. You'll get a boo-boo...
Look, Sabluk, we're adults here. You're gonna get your ears kicked for hooliganism.
and you have nothing to say? You admit that the forecast is bullshit?
I'll forgive the ears joke. I'll take it as an excuse for senility.
>> you're interested in a prediction with a flip of a coin?
I am interested in predictions and a post-mortem will show their veracity. I don't believe in waves myself and I don't follow forecasts, but the topic is interesting even psychologically.
Since Sart is as stubborn as a bull, he will get somewhere in the end. And I don't know why you're getting so hot and bothered, it's not worth it. If you're interested in Sart...
just annoyed by your stubbornness, just say so, it's a mundane matter.
and you have nothing to say on the merits? you admit that the forecast is bullshit?
>> about the ears, I'll forgive the joke.
The thing is, Sabluk, there's nothing substantive to talk about with you. We can only exchange pleasantries like this.
No, blend, in general - they are not losing monkeys, but about the same (which is even worse). If they were losing, they could be converted and made profitable.
not everything in the world can be converted, with a stop loss conversion is usually even worse,
the "as a whole - do not lose" - you need a criterion: if you take 100 codebase experts and go back 2 years or, even better, since 1999 (but this is relative to people), "people" I formulated it this way - "you may create conditions when it doesn't matter in what direction and at what distance to place orders", and what should I measure in an EA? The number of runs on parameters or the number of periods on the time axis. You cannot compare one run with specially selected parameters with 1000 random tests or with one random test
and you have nothing substantive to say? You admit that the prediction is bullshit?
And on the merits, having made the following prediction, Sart has responded. Once we get the statistics, the price of these predictions will be clear. In the end, he does not take money for predictions, what are the claims?
Z.U. And as for the substance, well, if you were a little gentler to people, your way of talking is capable of killing any branch started in the wake of showdown. And in no way should a discussion touch on matters of faith, for example, or turn to insults. That's not the point. I'm talking to you as a long-time scumbag. I hope I understand. >> no, you don't.
Come on, we're not human beings...
Let's make a non-aggression pact.
I'll take my revenge on Fourier, but let that be a lesson to the ignorant