Via API using external personal platforms (acting like a regular user placing trades on the TradingView chart).
Using the Eithcap broker, where participants could choose any supported trading platform.
Our team chose to use MetaTrader 4 (MT4) due to its advanced tools and flexibility, allowing us to create custom adaptations — even for functions incompatible with TradingView — by masking variables with code blocks that are accepted. This method allowed us to extract capabilities that TradingView itself doesn’t offer. Of course, this required a highly skilled team that fully understood the task.
The theoretical objective was simple:
To extract non-standard, dynamic information from the market so the algorithm could decide what to do in real time, placing pending orders in certain areas and cancelling them if the scenario changed.
While implementing this in standard MQL4 wasn’t difficult, the real challenge was teaching the algorithm how to avoid certain scenarios and how to reassess when conditions change — essentially a digital brain, similar to AI, but focused on short-term pattern recognition.
However, for that, it needed to store data externally, which required the creation of APIs to consult outside the MT4 environment.
Until a brilliant idea emerged: we could take a snapshot of the analyzed region, convert it into binary data (0s and 1s), and store it in a .txt file inside the platform’s Files folder. This way, the algorithm could access the data without relying on external servers.
This innovation eliminated delays caused by data fetching from external servers, which could cost valuable seconds — even a single second can be the difference between executing or avoiding an order. Everything had to be precise, and every step was a new challenge that demanded a solution never created before.
Fortunately, our coding experts don’t see these challenges as problems — but as opportunities to overcome. Passionate about their craft, they contributed voluntarily, driven by the desire to create something the world has never seen.
Two weeks later, the first prototype was ready in MQL4 — around 14,000 lines of code. The team nervously awaited whether MetaTrader 4 would accept the massive file or crash.
After countless compilation errors and 9 straight hours fixing just 8 errors, the code was finally running — a true milestone, and ready for tournament submission.
Then came the intellectual phase of the project:
The financial experts fine-tuned the code for secure risk management and sustainable balance growth.
After 48 hours of deep analysis, they requested specific adjustments, which the coders quickly delivered after 16 hours of intense work. The strategy was finally refined and optimized.
The goal wasn't profit maximization, but rather:
The lowest possible drawdown
The highest number of consecutive wins
While most competitors chased high profits (risking their entire equity), our algorithm aimed to grow gradually, with near-zero risk, steadily climbing toward success with unmatched performance.
After 23 consecutive trading days (excluding weekends), the competition provided each participant with a $500 demo account, allowing trading in the following assets for performance monitoring:
XAU/USD
NASDAQ
EUR/USD
Many accounts skyrocketed from $500 to over $10,000 within the first 10 days — impressive gains. However, in the days that followed, they not only lost all their profits but also blew the entire account.
Meanwhile, our algorithm advanced steadily and quietly, without drawing much attention.
During the entire competition period, it maintained a 100% win rate, completely eliminated risk, and consistently progressed.
As a result, it won the award for:
Best Performance
Top Security
Highest Risk-to-Reward Ratio
I’ll continue updating this article whenever possible, but for those truly interested, the file is already available for acquisition.