The data we are supposed to run on is his researched data from 2012? No way man. I never trust out dated data as fluctuations and outliers skew any data set
You can use the grade distributions from the most recent year to make calculations as accurate as possible
Obviously, nothing will be perfect - it is only a prediction, after all! This thread can be a great resource for the people that do wish to make predictions, though, and provides just one of the methods of study score prediction that are out there.
If someone does not receive a prediction that is as high as they desire, perhaps it can act as motivation to go and study more and work harder in order to realise their goal.
For what it's worth, I remember trying this back in 2015 for Revolutions, and it predicted the exact study score I achieved for it
Once scores were released last year, I used the predictor and got very similar numbers to what my actual study scores were.