Thank you for reading! I completely agree that the prior might be adjusted. However, disregarding low ratings would bring a survivorship bias into play. So, having no additional information I think it is fair to use Uniform distribution as the first naive belief. In the end the purpose was to demonstrate the framework :)