The article explores different ways to measure deviations in triangular distributions, commonly used in various fields. Instead of relying solely on standard deviation, the researchers compare other metrics like mean absolute deviation from the mean, median, and quantile-based deviation. They show how these measures can be easily constructed using basic geometric tools. The study reveals that mean absolute deviation from the median is the least volatile and often better than other deviation measures. This new method can help estimate distribution parameters more accurately, especially when dealing with limited sample data.