1 600 USD
Demo indirildi:
8 969
Yayınlandı:
2 Ağustos 2025
Mevcut sürüm:
6.5
Uygun bir robot bulamadınız mı?
Freelance üzerinden kendi
robotunuzu sipariş edin
Freelance'e git
Freelance üzerinden kendi
robotunuzu sipariş edin
Bir alım-satım robotu veya gösterge nasıl satın alınır?
Uzman Danışmanınızı
sanal sunucuda çalıştırın
sanal sunucuda çalıştırın
Satın almadan önce göstergeyi/alım-satım robotunu test edin
Mağazada kazanç sağlamak ister misiniz?
Satış için bir ürün nasıl sunulur?
Alım-satım fırsatlarını kaçırıyorsunuz:
- Ücretsiz alım-satım uygulamaları
- İşlem kopyalama için 8.000'den fazla sinyal
- Finansal piyasaları keşfetmek için ekonomik haberler
Kayıt
Giriş yap
Gizlilik ve Veri Koruma Politikasını ve MQL5.com Kullanım Şartlarını kabul edersiniz
Hesabınız yoksa, lütfen kaydolun
A visual demonstration of the training and forward testing periods.
Red indicates the data the model has seen during training.
Green indicates the data the model has never seen before — not even during validation.
MODEL: Scalper PRO5
Scalper PRO5 model shows growth over the 3 years leading up to the training period, and maintains good robustness after training.
Before 2017, the model does not demonstrate strong predictive power.
It should be noted that the StopLoss levels were adjusted based on current market conditions, which likely had some impact.
------------------------------------------------------------------------------------------------------------------------------------------------
Model:MIRROR
Mirror Model. Perfect out-of-the-box test results across any period — even without SL/TP optimization.
As can be seen, most of the data was used for training, including 2025 data (in mirrored form).
There is no forward testing period.
High likelihood of overfitting.
I do not recommend using it without thorough testing on a demo account.
------------------------------------------------------------------------------------------------------------------------------------------------------
Model:MAD TURTLE
Turtle Model shows the most stable performance overall. It uses a very short training period, while maintaining a consistent 5-year performance prior to the training zone.
Notable facts:
The model performs better on new, unseen data than on the training set — this suggests that it hasn’t merely memorized price structures, but has actually learned to identify patterns suitable for real trading.
Interestingly, the model selectively demonstrates stable performance during specific historical periods — for example, from 2009 to 2012, and from 2006 to 2008.
This indicates that the patterns utilized by the model have a long-term nature.
Overall, this is the best model among all.
The first trades
(The signal was not artificially boosted beforehand)I have the model running: Turtle and Scalper PRO;
Unfortunately, I forgot to change the magic number to Saclper, and this model closed positions earlier.
Important! Don't forget to change the magic number for each model. In future updates, I will add an automatic verification feature, there is already such a feature.
Hello Gennady,
Another impressive EA from you!
I was just doing some backtesting and had a few questions:
Why is it Buy Only? Is that just due to the current hype around gold – which only seems to go in one direction – so more of a precautionary measure?
During my tests, I didn’t notice any drop in performance when using Buy and Sell.
Also, why does your "Scalper_PRO5.set" file have the same name as your second strategy? That’s a bit confusing, since I can apply the set file to any strategy.
Hello Gennady,
Another impressive EA from you!
I was just doing some backtesting and had a few questions:
Why is it Buy Only? Is that just due to the current hype around gold – which only seems to go in one direction – so more of a precautionary measure?
During my tests, I didn’t notice any drop in performance when using Buy and Sell.
Also, why does your "Scalper_PRO5.set" file have the same name as your second strategy? That’s a bit confusing, since I can apply the set file to any strategy.
Alex greetings!
Let me answer in order:
1) If you look at the gold chart, you’ll see that for the past 1.5 years we’ve had a very strong upward trend (which is why so many strategies only take buy trades). Normally, of course, it’s much easier to trade with the trend than against it, but in the case of gold the trend has been so strong that a sell strategy simply cannot find good entry points.. especially considering that trading is done on H1 and positions are held for several days. If this were micro-scalping, it probably wouldn’t matter as much. But in our specific strategy, enabling sell trades has a negative effect.
It’s also worth noting another factor in machine learning: ML doesn’t behave like linear strategies, such as RSI > 90 = SELL and symmetrically RSI < 10 = BUY. In our case, the model is a three-class model — it learns to sell, learns to buy, and learns to filter; these are three different classes within the model. When trained on XAUUSD, the model always has a slight bias toward BUY. I’m not sure why, but I’ve only observed this effect on XAUUSD. So, it may be that the model is simply better at buying than selling.
To make the model fully mirrored, you’d have to use a binary training method — but I prefer more composite models, so think of this as a ready-made solution. However, this absolutely does not mean that the model can’t sell — in fact, a SELL signal is also a signal to close a BUY position.
In the settings, you can find the parameter:
close with counter signal
If you disable this parameter, trades will only close via SL/TP, and you’ll see how important SELL signals are for BUY positions.
2) Scalper_PRO5 - in the set file there’s a small error, after loading the set file, you need to manually enable the Scalper PRO5 model (I’ll attach the corrected set file below). But keep in mind that for this model you need to activate the bonuses, otherwise, the default model will be launched.
Thanks/
Please send me the set file.
Hello.
Please send me the set file.
Hello, I replied in pm
Weekly Updates:
Fixed static handle bug — previously, in some cases the model did not switch correctly and the default model would launch. For example, when switching from the default model to Scalper5, only the name changed while the actual model remained the same. This issue has been resolved.
Added and tested the split-lot option. This feature divides one lot into multiple smaller ones (e.g., split 1 lot into 5 × 0.2). For example, instead of opening a 0.1 lot trade, the system will open 5 trades of 0.02 each.
This was designed for prop firms, signal providers, and PAMM accounts.
⚠️ Important: this option can only be used with single-position strategies (e.g., Turtle). It cannot be used with Scalper5, since that model may open multiple positions.
Trading Results:
Turtle closed 1 Stop Loss and 1 profitable trade.
Scalper PRO5 is still holding 1 position from Friday and has added another trade along the trend.
Positions remain open.
Hi, i have purchase the EA, please send me the set files or setting for me. Thanks.
You only need to select the model. If your deposit allows, you can run all models simultaneously or just the ones you prefer.
If you need any help, I’m always glad to assist.
您好，对的，您完全正确理解了这个智能交易系统名称的含义。
hi i just bought this EA.. how to set this.. why dashboard in right below doesn't show any moving signal.. different from backtest auto moving signal..
Giao dịch có lợi nhuận. Tuy nhiên, có một giai đoạn Scaper Pro 5 giữ vị trí khá lâu mặc dù tín hiệu Mua chỉ là 2. Nếu đóng mức giá đó, rút lui sẽ cao hơn.
Instructions for launching the Turtle EA
Where to start?
Backtest and choose the models you like best.
If you cannot decide and your deposit allows, you can launch all models at the same time — this will not be a mistake.
Recommended minimum deposit for each model:
Turtle H1/M5: 150 USD
Mirror: 150 USD
Scalper: 200 USD
Omega: 300 USD
Short description of the models:
Turtle H1 (M5):
Main trend-following model.
Likes: strong trends.
Dislikes: trend reversals, flat phases.
Weakness: tends to enter at price peaks.
Strength: handles downtrends adequately.
Trades with 1 order.
Mirror:
Complex experimental model.
Has shown good real results after release.
Likes: unknown.
Dislikes: unknown.
Strength: flexible model, can extract profit from various structures.
Weakness: may experience "hallucinations" on certain structures.
Trades with 1 order.
Scalper:
Classic scalper with trend adaptation.
Likes: flat markets, moderate trends.
Dislikes: strong trends.
Strength: sharp entries, logical trades, quickly corrects mistakes.
Weakness: no strong negative traits.
Trades with up to 5 orders.
Omega:
Classic scalper with an aggressive character.
Likes: flat markets, trend reversals.
Dislikes: any kind of strong trends.
Strength: sharp entries.
Weakness: inability to adapt to trends, aggressive position holding.
Trades with up to 5 orders.
Hi,
Please advise on the set up.
How many charts do i need to open and attach the EA?
FAQ for Turtle EA
Where should I start?
Depending on the size of your deposit, you need to attach each model to a separate XAUUSD M5 chart (in fact, the timeframe doesn’t matter).
Each model goes on its own chart.
Below is the list of models and the margin required for stable EA operation:
Turtle H1: 150 USD
Turtle M5: 150 USD
Mirror: 150 USD
Scalper: 200 USD
Omega: 350 USD
So, if your deposit is 100 USD, you can only run Turtle or Mirror.
If your deposit is 1500 USD, you can run all models, each one on a separate XAUUSD M5 chart.
Do not change the settings!
No need to enable/disable BUY or SELL (see below for more details).
Do not change the Magic Number (each model has an extra digit automatically added to avoid conflicts).
In short, the only thing you need to change is selecting the model.
You can also change the Trade Comment so later you can see which positions belong to which model in your order list.
How to distribute risk?
Turtle H1: 0.02
Turtle M5: 0.02
Mirror: 0.03
Scalper: 0.02
Omega: 0.01
As you can see, Omega has 4x fewer lots than Turtle because it’s the most aggressive model.
Scalper has 2x fewer lots than Turtle because Scalper can open up to 5 positions, while Turtle opens only 1.
(Descriptions of each model are provided below.)
These are just recommendations. If you’re an experienced trader and can run your own tests, feel free to set your own risk allocation.
Frequently Asked Questions:
❓ Which model do you recommend?
👉 If you can only run one model, I recommend Turtle H1.
👉 If your deposit allows, run all models, but with risk distribution.
❓ Which models are you running on your signal?
Turtle H1: 0.04
Turtle M5: 0.04
Mirror: 0.03
Scalper: 0.02
Omega: 0.02
❓ What is PERFORMANCE mode?
👉 In this mode, all models are launched together.
❓ Can PERFORMANCE mode be used with other models?
👉 There’s no point — it would just duplicate models.
❓ Why is SELL disabled for most models?
SELL and BUY are separate sub-models in each model.For example, Scalper BUY and Scalper SELL are two different models inside the Scalper system.
Machine learning is not the same as simple linear strategies with mirrored rules, like MA10 > MA20 = BUY // MA10 < MA20 = SELL.
During training, the model may learn to BUY and SELL differently.
Conclusion: If a model buys well, it does not mean it will sell well. The difference can reach 10–30%. This effect is very pronounced on XAUUSD. On EURUSD, for instance, such a bias does not exist. Why? I don’t know — maybe due to the different structures of growth/decline.
The fact that SELL is disabled does not mean the SELL model is useless.
The SELL sub-model closes BUY positions. You can disable “close with counter signal” and check the importance of the SELL model.
Conclusion: even with SELL disabled, the model still does its job. It was exactly the SELL model that closed positions at the very top on 09.09.
In the Omega and Turtle M5 strategies, there is a significant BUY bias designed to speed up position exits. This is regulated by the parameters:
%threshold BUY
%threshold SELL
For example, in Turtle M5: %threshold BUY = 47, %threshold SELL = 44.
Conclusion: if SELL is enabled on such a model, it will enter sells too early.
In fact, BUY models do understand that the trend is not always upward — they attempt to adapt.
Run a Visual Test and carefully watch how the models behave.
Yes, of course, buying in a downtrend looks foolish, but the problem is that trend determination is extremely difficult.
The fact that the trend went up or down yesterday does not mean it will continue today. It only means the trend was up or down several days ago. Predicting where it will go today is extremely difficult.
If you think it’s simple — “just enable BUY in an uptrend and SELL in a downtrend” — you can be sure it only seems simple. Trend determination is an incredibly complex process.
Conclusion: Knowing when there will be a downtrend is extremely difficult. Even if you guess correctly once, most likely you will be wrong the next time.
The models try to maximize gains in an uptrend and minimize losses in a downtrend. Look at Turtle: all week it held its positions during the uptrend, for a long time, even when it seemed like time to sell — but today, Turtle opened a BUY and immediately closed it with a small profit. This shows the model understands well that the trend is changing.
Overall conclusion: I don’t know what will happen with XAUUSD — whether it will rise or fall. I only rely on numbers and tests. Officially, SELL is still not recommended to be enabled.