Casino gambler - page 8

 
Mathemat >> :

Forex is not roulette. The other probability distributions rule there. I won't go into it - search here for the phrases "fat tails", "Peters", "(non)Gaussian distribution", "martingale" (not "martingale"!).

You are one of the sensible people on the forum. All your posts say so, your opinion can be respected.

So notice, you don't send me to "[eq]", or shut me up, you send me to the articles - that's another level of discussion.

I know I'm provoking some people in my posts, asking questions that I actually know the answers to - but it's a method. Because asking a question - you're pushing interlocutor to the right answer, and I want to put questions so as not to push the interlocutor to say what I already know without him.


Further to the point. Of course I've read about fat tails etc. And of course I understand the difference between roulette and the market, albeit partly speculative, but still a market where real goods - currency - are traded.

But I still want to hear people's opinions. It is not for nothing that we entrust market movements to a mechanical machine (a computer), instead of deeply studying macroeconomics, collecting information on M1, M2 aggregates from official websites of central banks... etc., do not track trade balances from official sources and in short everything associated with it - you have to admit, we do not. If we do not do that, but write MTS, then we have come to the conclusion that it makes no sense to study economic macro-and even more so micro-indices, the movement of the currency cannot be predicted that way. So, logically we are still dealing with an unpredictable process - like in the case of roulette (again, I do not put roulette and Forex in the same line, there are simply common features).

So based on two points: first we are trying to predict the unpredictable process, and secondly, a lot of far from stupid people are doing it and can not come to a decision - based on this, has anyone ever suggested that the process is not necessarily white noise or something like that, so it can not be predicted? For example the trajectory of a fly is not white noise - but of course the fly knows what it is doing - but it won't tell us !...

 

I don't know whether you're disappointed or happy...

I've been doing some research - watching a fucking fly swish around under the chandelier

concluded that the fly is a predictable mechanism.

how do you know who's doing the analysis, i'm not limited to currency quotes

 
antoxa_zelyonyj писал(а) >>

So on the basis of two considerations: first, we are trying to predict the unpredictable process, and second, a lot of far from stupid people are doing it and can not come to a decision - based on all this, has anyone ever suggested that the process is not necessarily white noise or something like that, so it can not be predicted? For example the trajectory of fly's flight is not white noise - but of course the fly knows what it is doing - but it will not tell us !...

1. So far it is a statement that the process is not predictable. Where did you get that from? You have already been given a link to an indicator which shows that the process is correlated, hence there is a possibility of predicting it.

2. There are many processes which cannot be predicted, not necessarily white noise.

3. the fly is easy no problem. I can suggest an even more difficult problem. Two flies - you are on one of them and need to aim and hit the other. Already done and working.

 
antoxa_zelyonyj >> :

But I would still like to hear people's opinions. It is not for nothing that we all entrust market movements to a mechanical machine (a computer), and do not deeply study macroeconomics, do not collect information on M1, M2 aggregates from official central bank websites... etc., do not track trade balances from official sources and in short everything associated with it - you have to admit, we do not. Since we do not do that, but write MTS, it means we came to the conclusion that it makes no sense to study economic macro-and even more so micro-indicators, the movement of currency cannot be predicted that way. So, logically we are still dealing with an unpredictable process, just like in the case of roulette (again, I do not put roulette and Forex in the same line, there are simply common features).

Wrong conclusion. People don't study macroeconomics not because they have come to some conclusions, but because of laziness. It's much easier to race optimizations to the glow of the processor. To conclude that something does not make sense, it must be studied to the full, while here the thesis is that exchange rates do not depend on the fundamentals, but are set by the organizers in order to extort money from traders. There are very few people willing to learn.

Even a random process can be predicted. And yes, forex can be compared to roulette - for example, an order with an equal stop and loss, say 20 pips, has a 50/50 chance of profit or loss minus spread, so it is red/black on roulette minus zero.

 

откуда вам знать кто как делает анализ

I agree. There are people who flip coins, others who appeal to the stars, others who draw real works of art. Antoha, I have no right to force anyone, but I don't think provocation is the way to go.

 
Prival писал(а) >> Two flies - you are on one of them and need to aim and hit the other. Already done and working.

Not only does it work, but both flies really like it :)

About roulette: this is the analogy that came to mind. Let's imagine that Foreh is a lot of roulettes in one room (one per couple). Each of them is almost perfect (in the sense of "randomness"), but at least some of them are also rigidly, deterministically correlated. Well, for example, because EURUSD * USDJPY = EURJPY. Arbitrators make money on deviations from this equality.

 
antoxa_zelyonyj писал(а) >>

Firstly, we are trying to predict an unpredictable process, and secondly, a lot of far from stupid people are doing it and cannot come to a decision.

That's not quite true, forex is predictable, but it's not so easy to make real profit out of it in our sandbox environment. I have systems with ~60-70% entry/exit percentage (roughly the same number of times I can predict next candle's colour). That's great! The result must be positive, but this is conditioned by even distribution of these entries. In practice, the market becomes the least predictable exactly in periods of highest volatility, and hence losses of 30-40% of transactions, together with the costs can more than "compensate" for profits of 60-70%. Basically, it is possible to talk about unpredictability. Another thing is whether it is necessary to make money.

-----

And to come to a decision, you must first adequately formulate the problem. For many people it is secondary, for me too, for a long time overrated demands to the system did not allow me to see acceptable solutions just by looking around carefully. You can compare forex and casino (by the way MT and autotrading are not just forex), except I'm not sure that as you gain experience the odds of a player in a casino increase, as far as trading is concerned, it does.

 
Prival >> :

1. While it is a statement that the process is not predictable. Where did you get that from? You have already been given a link to an indicator which shows that the process is correlated, hence it is possible to predict it.

2. There are many processes which cannot be predicted, not necessarily white noise.

3. the fly is easy no problem. I can suggest an even more difficult problem. Two flies - you are on one of them and need to aim and hit the other. Already done and working.

>> your argumentation is stronger than that of your colleagues.

but the conclusion is that you vote for predictability. accepted :)

 
Mathemat >> :

Not only does it work, but both flies really like it :)

About roulette: this is the analogy that came to mind. Let's imagine that Foreh is a lot of roulettes in one room (one per couple). Each of them is almost perfect (in the sense of "randomness"), but at least some of them are also rigidly, deterministically correlated. Well, for example, because EURUSD * USDJPY = EURJPY. Arbitrators make money on deviations from this equality.

you vote for predictability too. accepted :)

 

In general, the korefes choose predictability, ok.

But I've decided for myself, I'm testing a couple of other systems and if I don't see steady profits - screw them all.

Although I think Forex quitting will be much harder than quitting smoking or even drinking, the strongest drug is money chasing :)

Reason: