Why do trading bots fail?

Why do trading bots, which are designed to automate trading decisions and take advantage of market inefficiencies, often fail to deliver the expected results? Could it be due to the complexity and volatility of the cryptocurrency market, making it difficult for bots to accurately predict price movements? Or could it be a lack of proper backtesting and optimization, leading to subpar performance? Additionally, could the use of outdated algorithms or a lack of adaptability to changing market conditions also contribute to their failure? It's crucial to understand the reasons behind trading bot failures in order to improve their performance and maximize returns.