This dissertation evaluates attempts at electoral engineering in Turkey to understand the regime's stability and longevity from 1980 to 2018. The primary focus is on determining whether electoral engineering functions as intended and why electoral authoritarians make significant mistakes despite having access to data and resources. The study aims to understand the extent of these failures and the circumstances under which they occur by conducting simulations and comparing their results with actual election outcomes. The dissertation comprises three chapters, each exploring different aspects of elections and electoral politics in Turkey. The first chapter focuses on the formula proposed by Flis, Słomczyński, and Stolicki (FSS formula) for predicting seat allocation in Jefferson-D'Hondt elections. I propose a "regional correction" to this formula to improve its accuracy in nations with regional imbalances, using data from Turkish Parliamentary elections. I show that dividing a country into three regions that are similar to each other based on a machine learning algorithm, k-means clustering, provides significantly more precise estimates for simulating elections. The second chapter applies this method to analyze the June 2018 Turkish Parliamentary elections, revealing that most parties made flawed decisions in designing the 2018 electoral reform in Turkey. I show that the incumbent governing political party, AKP, lost approximately 27 to 28 seats in the parliament due to the reform they implemented. The third chapter expands the scope to include electoral engineering efforts from 1980 to 2018, demonstrating a high failure rate in political parties' attempts at optimal electoral strategies. By analyzing electoral engineering attempts such as initiated reforms or instances when political parties had ample opportunities for electoral engineering but they did not. Overall, I show that political parties were not able to make better decisions compared to random decision-making on electoral engineering. This research highlights the complexity of electoral engineering and the frequent missteps by political parties, offering insights into the potential outcomes of various hypothetical scenarios in Turkish elections.