Building a Stable EA Together (Looking for Backtest ForwardTest)

You should tell people what to expect. They are testing for you and suddently the EA has expired. What will be the end? I have seen so often, that the OP just stop releasing when he had enough inputs and the tool is ready for sale. Then the people who has used time on the case just get a long nose.

Hi @RiskFighter,
I understand your concern — and to be fair, that situation does happen quite often in this space.
Just to clarify, there’s no intention to “disappear” or suddenly turn this into a sales thread. The reason the earlier versions had expiry was mainly to avoid outdated logic being used in changing market conditions, which can be risky if left unmanaged.
What I’m doing here is more of a continuous development process — testing, observing behavior, and improving the approach step by step. I prefer to keep things transparent and share the evolution rather than release something and leave it unmanaged.
I appreciate everyone who has taken the time to test and give feedback — it’s part of building something more robust over time

Interesting approach on the collective testing idea. One thing I'd suggest before running forward tests: make sure your backtest setup avoids common pitfalls that can make results look better than they are.
A few things to watch for with EA backtesting:

1. Tick quality: MT4's "Every tick" mode still uses interpolated ticks, not real market data. Results can differ significantly from live.
2. Spread handling: Fixed spread backtests hide a lot of pain. If JG-3/JG-4 scalp or trade during news, variable spreads will eat into those results.
3. Close[0] vs close[1]: If the EA reads the current bar's close to trigger entries, you're getting look-ahead bias. Only the previous bar's close is confirmed data.

For anyone wanting to validate the logic visually before running it live, tools like Backtrex let you map out entry/exit rules as blocks and see exactly where trades would fire on historical data. Useful to spot if the EA is entering on conditions that wouldn't hold in real-time.
Happy to help review backtest reports if you share them. What pairs are you focusing on?

Hi @backtrex,

Appreciate the detailed input — all very valid points

Especially on tick quality and spread handling — those are exactly the areas where backtests tend to give a false sense of stability if not handled carefully.

On my side, I’ve already moved away from relying too much on backtest optimization for that reason. The focus now is more on how the logic behaves under live conditions, particularly with variable spreads, execution delays, and changing volatility.

Also agree on the close[0] vs close[1] point — avoiding any form of look-ahead bias is critical, otherwise the results become misleading very quickly.

At this stage, I’m more interested in forward behavior and consistency rather than how clean the historical curve looks.

Currently focusing mainly on XAUUSD.

Thanks again for sharing your insights

Hi everyone,

After several rounds of backtesting and forward testing, we’re happy to share that both JG-3 and JG-4 are now in their final optimized versions.
What’s new:

1. Improved entry filtering (better drawdown control)
2. More stable execution across volatile sessions
3. Optimized for XAUUSD

Current Status:

1. Forward test results are stable
2. Risk management has been tightened
3. Suitable for real-account testing (with proper lot sizing)

IMPORTANT NOTICE
This version will be available only until April 30.
After that:

1. EA access will be discontinued
2. We will transition to a new model (please reach out for detail)

Next Phase
If you want to continue using the strategy after expiry, please DM me.

We truly appreciate all the feedback and participation so far.
Feel free to share your forward test results or questions.

Let’s grow this together