Optimization Guru – Analytics & Forecasts – 21 January 2023
Recently, many users ask me about the difference between MT4 and MT5.
The difference between MT4 and MT5 is huge, it’s just an abyss, especially if you are a software developer.
Most of these users are still using MT4, just out of habit, or they have previously bought a lot of systems in MT4 and they simply have no other choice.
MQL Market does not allow you to change the platform, even if you have already purchased the system in MT4.
Developers don’t have those privileges either, we just don’t have the rights to do so. We are not MQL administrators.
MT4 does not allow you to receive quotes directly from your broker, it is simply not possible.
All you download in MT4 is MetaQuptes quotes, which are not even brokers.
It is possible to get paid quotes from Dukascopy, but this already guarantees you failure during testing in real trading.
Your broker does not use these quotes in real trading, Dukas is just one of the possible quote providers.
I already wrote earlier in my blog about it.
MT4 is an outdated platform, so much so that it only allows one check at a time, so optimization is extremely difficult and long for MT4.
Even if you have the most powerful processor, MT4 will still only use 1 thread. This is a platform architecture that did not even anticipate such possibilities in the future.
Such testing and optimization can take several days. For any optimization, you need to know all the settings of your system very carefully.
If your system is for MT4, this is already doomed to failure, because. testing accuracy will be low and finding good settings is either by chance or using huge resources such as time.
That is why I recommend everyone to move away from MT4 towards MT5. MT5 allows you to use all the threads of your processor.
For example, my server is able to have 70 free threads. Each thread takes up about 1-2 GB of RAM during testing, so you will need at least 70-140 GB of very fast RAM to just run 70 tests at once.
Optimization in MT5 involves searching through all the settings that you specify in the required range.
For example, you can specify how SL should change from 100 to 500 points in 50 point increments.
Thus, you get 10 iterations at once during optimization in MT5.
The more iterations you use, the longer your testing will take.
The more RAM and CPU threads you have, the faster you can complete the optimization.
After the completion of the optimization, you get the results of all the settings that have been reworked and you can choose the best and most convenient settings for yourself.
Let’s say you have already optimized and want to choose the best settings.
Usually it is a dilemma between the accuracy of the settings and the maximum profit.
Someone cares about accuracy and wants to sleep soundly with very low chances of taking any loss.
Someone does not pay attention to losses and only the final profit is important for him.
Even the largest investment companies cover losses, it is impossible to trade constantly and receive only profit.
Also, when choosing settings, you should definitely understand the ratio of average profit and possible loss.
For example EveEA yesterday made a profit of 540 points for GBPNZD and this is also displayed during the check on real quotes in MT5.
At the same time, GBPNZD has a maximum theoretical SL of 1000 points. Those. the ratio is about 1 to 2.
EveEA constantly monitors the broker and has expiration time protection and follows the market, accompanies both profit and loss.
Very few systems based on MQL are able to accompany the loss and minimize it.
So SL is just protection, in real trading the average loss is only 300-350 pips.
Thus, the ratio changes slightly and is approximately 1 to 1.
Therefore, when choosing settings, it is very important to take this ratio into account.
Most MQL systems have very low profits and huge SL.
They over-hold positions and do not follow the market, they do NOT accompany losing positions.
Therefore, from the point of view of the probability of getting 20 points of profit, the chances are much greater than of making a loss.
It’s just a theory of probability.
Therefore, if during optimization you get high SL and low TP settings, these are bad settings, although in numbers they can have a huge ProfitFactor.
Also, always pay attention to the number of transactions that occur in 1 year. If there are too few deals, then it’s just a history fit.
Almost all systems on the MQL Market have been re-optimized. This means that if the market changes a little, in this case the system will be broken.
Over-optimization is when you just overdid your settings and made them too good, too precise.
Therefore, I recommend using MT5 for both developers and users.
All our systems are open. This means that all settings are available to you right out of the box and you can create something of your own.
It’s always good to stay away from the crowd and have a slightly different set up.
This will allow you to hide from the eyes of the Market Maker and be a little out of the way.
To enable this mode, you simply need to disable the ONE CHART MODE= False trading mode on one chart.
Thus, the system will simply use the current settings for the current currency pair.
Also, I never recommend checking and optimizing systems on VPS.
During testing, MT5 downloads huge amounts of data and quotes from the broker, it takes up a lot of space on the SSD (HDD).
Therefore, the free space may simply run out for the remaining active MT5 terminals on your VPS.
For any questions, I will always be happy to help in Telegram
Comments are closed.