How was your first EA developed? - page 4

New comment
 
Osmar Sandoval Espinosa:

For me was very hard, it took like 2-3 months, the thing that helped me the mos wast replicating a pattern that I watched in the market:

  • High of the day + M15 sweep / Low of the day + M15 sweep

Firstly backtesting it a lot and then taking screenshots of the pattern, then started coding and learning mql5.

After that, managing risk, fixed lots or percetange, BE or trailing, etc...

It was quite a journey...

How was your first EA?

Honestly, mine didn’t start seriously at all.

I thought I had a solid idea, jumped into it, mixed random code together… and the result was a complete mess. First backtest wiped out fast enough to be a reality check.

That’s when I realized I don’t actually understand what I’m building.

After that I started focusing more on learning how things work — going through MQL5 resources, checking logs, observing behavior, and understanding risk instead of just expecting results.

Still learning, but that phase probably taught me the most.

 
Ryan L Johnson #:

I stopped checking in on TV years ago. Once I saw that coding a "hook" to a paid 3rd party automation website was the only way to automate trading, I declared it a kid's toy.

I even gave TS a go at one point. Nope─it printed duplicate Renko bars. So much for integrated features. It's a sad day for TS when free custom utilities here blow TS's integrated features out of the water.

Hmmmm, I think that nowadays they have PineScript that can also be used to create indicators and strategies (as far as I have seen in the reddit community),
 
Boris Sedov #:

I started my journey with MT4 back in 2005. At that time, I knew a little programming, but there was still so much I didn’t understand.

I don’t quite remember what my very first EA was, but one of the earliest ones I do recall came together pretty quickly. It was a simple test of an idea — trading based on indicator signals with a very tight stop loss and high risk. Classic "search for the Grail" stuff.

I spent countless hours in the strategy tester back then. Days and nights just tweaking parameters, watching how things behaved, chasing that perfect curve.

Very interesting. :)

How lond did it take you to feel secure about your EAs? In my case like 6 months.

Years passed. My interests shifted, and now I work on completely different things — professional trade copying solutions, modern GUI, architecture that actually makes sense for real trading.

But looking back, those early days were genuinely exciting. A time of experiments, mistakes, and endless curiosity.

Okay, aside from that, do you trade with EAs or not anymore?
 
Vishal Verma #:

Honestly, mine didn’t start seriously at all.

I thought I had a solid idea, jumped into it, mixed random code together… and the result was a complete mess. First backtest wiped out fast enough to be a reality check.

That’s when I realized I don’t actually understand what I’m building.

After that I started focusing more on learning how things work — going through MQL5 resources, checking logs, observing behavior, and understanding risk instead of just expecting results.

Still learning, but that phase probably taught me the most.

Interesting!

What did you tried to build at first?

 
My first EA was not really about strategy, but about fixing problems.

I started from simple tools:
- break even
- risk limits
- partial close

Only later I worked on entry logic.

In my opinion, most EAs fail not because of entry, but because of poor risk management.
 
Osmar Sandoval Espinosa #:
Hmmmm, I think that nowadays they have PineScript that can also be used to create indicators and strategies (as far as I have seen in the reddit community),
I just did a quick search and it looks like TV/Pine still uses alerts combined with 3rd party webhooks as a factory hodgepodge auto-trading workaround.
 
My first EA was honestly overcomplicated.

I tried to include everything — multiple indicators, filters, different entry types… it looked good on paper but in reality it just overtraded and broke down in different market conditions.

The biggest shift for me was simplifying it and focusing more on behaviour rather than entries. Things like when not to trade, how it manages positions, and controlling risk properly.

Once I stopped chasing perfect entries and started building around structure and discipline, everything improved a lot.

Still learning with every version but that was probably the biggest lesson early on.
 
Marco Savia #:
My first EA was not really about strategy, but about fixing problems.

I started from simple tools:
- break even
- risk limits
- partial close

Only later I worked on entry logic.

In my opinion, most EAs fail not because of entry, but because of poor risk management.
What have you found that works the best for risk management?
 
Ryan L Johnson #:
Ryan L Johnson #:
I just did a quick search and it looks like TV/Pine still uses alerts combined with 3rd party webhooks as a factory hodgepodge auto-trading workaround.


Lol, yeap I also found the same.


I thought that you could connect your account to TV and trade...

 
Keegan Ellis #:
My first EA was honestly overcomplicated.

I tried to include everything — multiple indicators, filters, different entry types… it looked good on paper but in reality it just overtraded and broke down in different market conditions.

The biggest shift for me was simplifying it and focusing more on behaviour rather than entries. Things like when not to trade, how it manages positions, and controlling risk properly.

Once I stopped chasing perfect entries and started building around structure and discipline, everything improved a lot.

Still learning with every version but that was probably the biggest lesson early on.
Nowadays which strategy do you trade?
1234567
New comment