The gaming industry produces vast amounts of data that can inform on many different kinds of research problems. Unfortunately, this data is rarely made available for research purposes and when it is, there are often limitations presented by the researcher's need to protect the data source's identity. In an effort to solve this problem, a Markov chain based statistical process for producing synthetic slot floor event data that is realistic and statistically similar to a provided real data set is proposed. Due to the uniqueness of this statistical process, two new methods of transition probability matrix estimation are introduced. This statistical process and the new probability matrix estimation methods are tested on an anonymous data set. The process is able to replicate event data and resultant session data distributions well, but falters in approximating a realistic time series.