To prevent a particular user from skewing the average I would suggest calculating average ratings of the beer for each user, and then average those values (versus averaging all the ratings independent of what user entered them).
This way somebody who really likes American Light and logs it all the time as a "5" won't skew the fact that most people only drank this once and consider it a "1".
To prevent a particular user from skewing the average I would suggest calculating average ratings of the beer for each user, and then average those values (versus averaging all the ratings independent of what user entered them).
This way somebody who really likes American Light and logs it all the time as a "5" won't skew the fact that most people only drank this once and consider it a "1".