Money management strategies. Martingale. - page 7

 
In short, no one really knows what Martingale is
 
Sorento >> :

That's what this M. So you need a classification to discuss?

Maybe start at the beginning, when the concept of martingale came into existence.

 
Sorento писал(а) >>

That's what M is. So you need a classification to discuss?

No. The definition includes everything. Direction is irrelevant to assignment to M.

 
sanyooooook >> :

Maybe start at the beginning, when the term "martingale" first appeared.

Martingale and martingale are not the same thing.)

 
TheXpert >> :

Martingale and martingale are different ))

explain what is their difference?

 
paukas >> :

No. The definition includes everything. The direction is irrelevant.

A simple increase in the bet? 1-1.1-1.5-1.9 also a martingale then?

Or is it still 1-2-4... where when you fix a loss the result is 1.

And supermartingale class 1-3-7-15 where winnings increase 1-2-4?

I think we should split up.

Because martingale is a swear word to some people.

So is over-sitting. Three points over-sitting?

А 7?

;)

 
Sorento писал(а) >>

A simple increase in the bet? 1-1.1-1.5-1.9 also a martingale then?

Or is it 1-2-4... where the result is 1 when fixing a loss.

And supermartingale class 1-3-7-15 where winnings increase 1-2-4?

I think we should split up.

Because martingale is a swear word to some people.

So is over-sitting. 3 counts of over-sitting?

А 7?

;)

1. Yes, a simple increase in the rate.

2. I already answered about over sitting above, read it.

 
Sorento >> :


So is over-sitting. 3 points - over-sitting?

А 7?

;)


It depends on the dc, on some it's an MK)
 
Sorento писал(а) >>

I agree. Except that the moment of this very discovery is left out, as is the assumption of a possible error threshold.

It's not about error. It's about the probabilistic nature of the game. You can't foresee everything. Take the coin toss as an example:

We play a game of eagle with an opponent whose capital is infinite, ours is finite. We always bet on the eagle.

1. Variation of the problem: The coin is fair (probability eagle/rake=0.5/0.5) then there is no optimal strategy for us, and there is no point in playing at all.

2. Option: dishonest coin (for example eagle/tooth = 0.6/0.4) then we have an advantage and optimal strategy is always one - to enter each time a certain fraction of capital, which is easily calculated. Not by a fixed lot, but by a fixed fraction.

3. Variation: As in the previous example, we have a coin on our side and also a coin (or the process of falling out) has a memory: after falling out with tails two or more times in a row, the fall of heads increases, for example to 0.7. There are essentially two games here. The first is the same as in the previous example and the optimal share of capital on the bet is the same. This strategy is played until two heads in a row fell out. As soon as tails fall out and until a continuous series of tails is interrupted we play another strategy. More precisely, we play another fraction of equity. In this case it is more than in the first strategy because the probability is higher.

4. Option: The probability of getting heads increases along with the length of a continuous series of tails. Then for each continuous series of tails there will be a different optimal share to enter. That will be the martin. And it is only effective when our game consistently becomes more and more profitable for us. If it becomes less and less profitable for example, it is effective to successively decrease the stake, which for market systems can be realized as a partial closing a position.

 
Avals >> :

It's not about a mistake. It's about the likely nature of the game. You can't foresee everything. Let's take the coin as an example:

We play a game of eagle with an opponent whose capital is infinite, ours is finite. We always bet on the eagle.

1. If the coin is fair (probability of eagle/rake=0.5/0.5) then there is no optimal strategy for us, and there is no point in playing at all.

2. If the coin is not fair (eg Heads/Tails = 0.6/0.4) then we have an advantage and the optimal strategy is always the same - enter each time a certain fraction of capital, which is easily calculated. Not by a fixed lot, but by a fixed fraction.

3. As in the previous example, we have a coin on our side and also the coin (or the process itself) has a memory: after tails are hit two or more times in a row, the eagle hit increases to 0.7, for example. There are essentially two games here. The first is the same as in the previous example and the optimal share of capital on the bet is the same. This strategy is played until two heads in a row fell out. As soon as tails fall out and until a continuous series of tails is interrupted we play another strategy. More precisely, we play another fraction of equity. In this case it is more than in the first strategy because the probability is higher.

4. The probability of getting heads will increase with the length of a continuous series of tails. Then for each continuous series of tails there will be a different optimal fraction to enter. This will be a martin. And it is only effective when our game consistently becomes more and more profitable for us. If it becomes less and less profitable, for example, it is effective to successively decrease the stakes, which for market systems can be implemented as a partial closing of a position.

I'll remind you again - we initially assume that we've estimated the probability of market moves in one direction or another...

And if forex is reduced to your example, there is no need for a TA.

And it is impossible to estimate probability of type. It is always 50/50! (And the coin, by the way, has no memory. Multiplication of probabilities will work.)

It's a bit creepy. And even a Wiener wandering or a stretched string is closer to me, and says that my goals will be reached by the market sooner or later.

;)