2/19 Sometimes I borrow the power of ChatGPT, Episode 29 ~ Various EA developers ~ - EA Beginner Course - Navi+ - GogoJungle | Auto Trading - Market Analysis

FX Auto Trading, EA developersReiwa’s Double-Ihere.

Since last year, I’ve been writing this series aiming for daily posts.

We’ve probably continued for almost a year now!

If you read what I’ve written so far,

I’m confident you’ll come closer to better EA operation.

And since I’d like to take a short break,

I’ll continue posting with the help of ChatGPT.

Occasionally I’ll post irregularly, as before.

EA Developer’s Way of Thinking – Reflections on Loss Count and Logic Building

Hello, I’m Reiwa’s Double-I, an EA developer.

Today, I’d like to discuss the “importance of loss counts” and “how to build logic” in EA development,
sharing my and other developers’ thoughts while progressing a thoughtful discussion.

By gathering questions and ideas that many EA developers feel,
I hope new discoveries and inspirations will emerge!

First, about the idea of loss counts.

Why is the “loss count” important?

Regarding the topic of “loss counts” mentioned by Mr. T,
there is a view that when the number of losses is small, the EA may fail to handle unexpected market moves in actual operation.
This can lead to overfitting the EA or lack of reproducibility, so it’s a point to be especially careful about.

An EA with few losses may overfit to certain market environments in backtests or optimizations,
and when facing different market conditions in real trading, it may become weak.

For example, if an EA trades 100 times and wins 90 times (win rate 90%), there are only 10 losses.
Continuing operation with only those few loss patterns increases the risk of not coping with other market conditions or loss patterns.
If real trading experiences more losses, it’s highly likely you’ll think, “This pattern was unexpected!” and fail to respond.

On the other hand, an EA with many losses allows testing various loss patterns in advance.
By understanding which patterns are prone to losses and taking countermeasures, you can develop a stronger logic EA.

However, too many losses can also make mental management during operation difficult.

This depends on each investor’s risk tolerance.

EA Logic Building – Developers’ Styles

Next, I asked other developers how they assemble EA logic.

【Reiwa’s Double-I’s Logic Building】

In my case, I often start by considering the decision logic (settling/exit logic) first, then search for entry logic.
I look for a place where I think, “If I settle here, I’ll win,” and build the entry points from there.

This approach means I decide the winning conditions first, and then set entries to satisfy those conditions.
There are times when entry comes first, but fundamentally I prioritize the exit (settlement) that leads to victory.

【Other Developers’ Views】

H-san watches candlesticks for a long time, discovers zones and patterns from them, and shapes the EA logic accordingly. He observes the chart and develops ideas from the feeling of shaping candlesticks.

T-san creates EA with a theme first, and builds logic aligned to that theme to clearly define the final goal, such as “Create a high-win-rate EA” or “A discretionary-reproducing EA.” The theme keeps the logic from drifting.

S-san gathers external information and inspiration, then builds EA logic based on that. He studies “millionaire trader methods” from investment magazines and translates them into EA.

Learning from Diverse Styles

What I feel here is that each EA developer has their own unique style.
All methods are effective, but all developers share a common stance: taking notes, continuously observing the chart, and valuing flashes of inspiration.

I also make it a habit to take notes whenever ideas arise, and those notes often become the foundation of new EA logic.
In everyday moments, bright ideas can appear, so it’s important not to miss even a small idea.

Summary – Loss Count and Logic in EA Development

**“Loss count”** is not only about win rate or the profit factor; it’s an important element to examine how losses occur.
EA with few losses may be weak to unexpected markets, so it’s crucial to test with diverse patterns.

Also, while the logic-building approaches differ by developer,
a common trait is the willingness to gain inspiration and turn it into a concrete form.
Some leverage discretionary trading experience, others pursue research from external information, but all have their own style in development.

In EA development, having diverse approaches while establishing your own style might be the shortcut to success.

With that, please continue to have a good trading life!

━━━━━━━━━━━━━━━━

【From Starting EA to Intermediate – Flow Summary】

First, mastering the installation of EA is important.

--------------------------------------------------------------------

MT4 EA (Expert Advisor) and Indicator Installation Procedures

https://www.gogojungle.co.jp/post/1/8212

--------------------------------------------------------------------

※GoGoJungle Video Version※