Cryptocurrency markets, with their extreme volatility, pose a significant challenge to accurate price prediction. While manual trading often relies on subjective assessments, algorithmic trading, which combines technical analysis and machine learning, has emerged as a powerful tool for cryptocurrency price forecasting and portfolio management. This study compares the effectiveness of Bricks charting technique and Japanese candlesticks data for predicting one-step-ahead daily logarithmic price returns in the Bitcoin/USDT pair market. By integrating an GRU model and transforming OHLC data into graphical bricks, we aim to enhance forecasting accuracy. The performance of both approaches is evaluated using Mean Squared Error (MSE), Mean Absolute Error (MAE), and coefficient of determination (R2). Our findings demonstrate that Bricks charting technique significantly outperforms Japanese candlesticks data, consistently achieving high R2 values. This suggests that Bricks charting technique is a more effective representation of cryptocurrency price dynamics, allowing the GRU model to capture underlying patterns and make more accurate predictions. While their effectiveness may vary across different timeframes and cryptocurrency markets, this study highlights the potential of Bricks charting technique as a valuable tool for algorithmic traders seeking to improve their forecasting capabilities.