What is it? - page 24
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
Хорошо что не лоботомия...
:)))
Sberbank of Russia or something?
Ага, а Милиса - мировая закулиса. Это ответный подарок :)
Thanks ;) Then by analogy, Milibia is backstabbing. Worldly, self-seeking.
Vapche terminology seems to be quite in the spirit of the topic. More precisely in the letter.
Here's a linguistic contrivance invented: Follos is "successful forex trader".
;)
Практическое применение уже как бы состоялось. )) См. выше.
Появилось предположение, что этот реальный выигрыш не случаен. (Во всяком случае ему нельзя найти простое объяснение)
Это необходимо доказать (и желательно это сделать чисто математически - цель моего обращения здесь).
Если получится доказать, то система достаточно легко проэцируется на Форекс.
Очень коротко, но надеюсь ответил Вам.
Is this meant by practical application or martin?
Is that meant by practical application or martin?
Yeah, right. Tried to re-read the thread myself from a re-reader's point of view... Complete confusion and nothing is clear. And the linguistic acrobats "help" ....
And you want to find the origins, where the "legs grow". Have I understood you correctly?
By practical application I mean the following:
........
One day a group of comrades (GT) wanted to use roulette in the casino as a stable source of income.
They approached the matter "responsibly". It took about a year and a half to find, test, etc. Eventually came up with a game system (IS). It was time to test it under combat conditions.
The GT was well aware of the negative MO, the impossibility to win at roulette, in short: no illusions. Just as much effort was put into finding mathematical justifications (proofs) of the methods used. But the battle was won without the support of mathematicians.
For the game a sum of approximately one hundred stakes was determined. In the course of the game there were "bumps" (drawdowns), and different bets were used (it can be called a kind of Martin's 1 to 3, very seldom reached odds of 5).
We were in deficit a little (in the beginning), most of the time we were in the plus.
Maximum balance reached about +3.5 depo. Then there were ideological contradictions within the GT. As a result, I moved away from the IS and downgraded the account to +0.5 depo. Then a return to ~ +1.5 depo. And termination of the experiment. Played ~ 30000 games. Every day logs were kept. ))
.......
Many will say, yes you had a martin, so it helped. Do the math for yourself. Our result would have been between (minus) -6 and -7 starting deposits. Martin would have only accelerated the defeat.
Conclusion?
Did you get practical data that does not fit into the ideal SB model? Or you have theoretical reasoning leading to a contradiction of the TV axioms or their implications?
I think I have answered you at the same time. And you have theoretical reflections too. ))
And if so (as applied to the market) . We open a position with SL=TP=n, probability of order closure is (n-spread)/2n around 50% if n>> spread. When the SL/2 level is reached, half of the position is closed . Then closing the whole position by SL will give less loss than closing the same position by TP .
No, because part of the TP will be closed with half lot.
Probability of first reaching the SL/2 level before reaching TP: 2/3. That we reach TP immediately: 1/3
If we reach the SL/2 level, then with 3/4 probability we will have a stop loss and with 1/4 probability we reach TP
I.e. there are 3 possible outcomes
probability of winning
1/3__________ +1
2/3*3/4______ -0.75
2/3*1/4______ +0.25
Price of the game = 1/3 - 6/12*0.75 + 2/12*0.25 = 0 excluding spread
P.S. Of course, all calculations are for SB. In reality a partial close may make sense using certain price patterns.
A sum of about a hundred bets was determined for the game. In the course of the game there were both "strikes" (drawdowns), and different bets were used (you can call it some kind of Martin's 1 to 3, very rarely the odds reached 5).
We were in deficit a little (in the beginning), most of the time we were in the plus.
Maximum balance reached about +3.5 depo. Then there were ideological contradictions within the GT. As a result, I moved away from the IS and downgraded the account to +0.5 securities. Then a return to ~ +1.5 depo. And termination of the experiment. Played ~ 30000 games. Every day logs were kept. ))
.......
Many will say, yes you had a martin, so it helped. Do the math for yourself. Our result would have been between (minus) -6 and -7 starting deposits. Martin would have only accelerated the defeat.
Conclusion?
Martin does not accelerate the loss. He is capable of stretching the game considerably if lucky at the start. I.e. it is sensitive to the t-count.
Here's an example generated in Excel. Eagle, player always bets on the same thing (eagle for example). Initial capital is 10000. First bet is 100, then 200 in case of failure and so on by doubling. When we win we start again from bet 100. Margin Call is not taken into consideration, i.e. leaving in minus is possible.
For illustration of the chart, if we get broke we start again from 10000. Here are a couple of generations:
We can see that there were successful series, where the deposit increased 50 times. But the average price of the game is still 0. Because there are a lot of unsuccessful starts and 10000 plummeting, as well as plummeting in the end of any successful series.
Yes, if you stop at a certain peak, you'll be in the black. But one cannot predict in advance when one will stop and after how many failures this long successful series will come.
I'm not saying that martin is nonsense, but in order to use it in any modification there must be appropriate row properties (antipersistence), and if they are not detected, it's a game of luck. Yes, it can get lucky. And stopping after 3 doubles, or any other tricks they will not make MO positive, although they will affect some equity characteristics. For example, will make it less likely to drain at the start, but reduce the probabilities of super lucky series, etc.
Finding a system on a knowingly random series, without using its specific properties, is akin to finding a perpetual motion machine. imho
I completely agree with your last post.
Only for the sake of purity of the experiment, your Excel model should definitely include:
1) Zero - i.e. every 37 game is unprofitable, no matter what bet is placed on it. But this loss is "noticed" only by Equity. For martin it is as if there was none, it works according to its own scheme.
2) And Martin's restrictions - let there be 4 steps {100;200;400;800}.
3) And at the same time it is desirable to know the value of maximum and average series before the margin call.
Please, do it. And compare.
I completely agree with your last post.
Only for the sake of purity of the experiment, your Excel model should definitely include:
1) Zero - i.e. every 37 game is unprofitable, no matter what bet is placed on it. But this loss is "noticed" only by Equity. For martin it is as if there was none, it works according to its own scheme.
2) And limitations on Martin - let it be 4 steps {100;200;400;800}.
3) And at the same time it is desirable to know the value of maximum and average series before the margin call.
Please, do it. And let's compare.
It's easier to do it in MQL by the script.
If you specify the conditions:
1. A margin call is considered when there is not enough money for the minimum stake (100).
2. After losing a bet of 800, we start again at 100
3. if the deposit is not enough to double the stake, we again start from 100.
4. initial deposit =10 000
Then average duration of play to margin call = 1690 with 100.000 margin call trials. Converges to this number even with a much smaller number of trials.
To estimate maximal game duration you need much more statistics (number of games to a margin call)
So for 1000 margins it is 15000-20000.
For 10,000 margins it is 33000.
At 100,000 margins it is 35,000.
Therefore converge approximately to 35000. The maximal increase of the deposit is 8 times.
And by the way, the maximal increase does not change much. At 1,000 margins it changes by a maximum of 6 times. At 100 it's 5 times. At 10 it is 3 times. Of course, the smaller the number of tests, the greater the spread. And when playing up to 1 margin may well disperse and 3 - 5 times. But most likely you will withdraw much earlier.
to margin call = 1690 with 100,000 margin call trials. It converges to this number even with a much smaller number of trials.
Much more statistics (number of games to margin call) are needed to estimate maximum game duration.
I'm a bit confused...
You have one trial - from zero trades with balance 10000 units to the last trade in this series n(i) with balance < 100. Is it correct?
With five trials we have, for example, n(1-5)={1000;1500;7000;400;2200}, then n_average = 2420, n_max = 7000.
But you're counting something else under n_max.
Wouldn't it be better to show it in Excel, with graphs?