How much time do you have up your sleeve? O_o
Perhaps quite a bit?
Ok I finally figured it out, but I think it's quite...um inaccurate. I put in the exact results from VCAA that I got from last year for my 3/4, it came out with a 41 when I got a 44, so it's sort of right, but sort of off the mark..
Hmm. That's interesting, what subject?
I gave it a shot with what I had back when I did yr 12. It seems to overestimate study scores by a fair bit, e.g. for spesh it gave around 6 higher than the actual (putting a B+ in one of the entries and 47, I don't think so). I had a look at the code behind it, and not sure too much what you've done differently. Mao's old Study Score estimator was always really close to the scores I actually got. Besides making it easier to enter things and and putting the grade distributions on the side and Paul's stuff, I don't really see anything new. I know you put effort into it, it's just that it's feels like it's taken other peoples work and ideas and just made things look fancy, and then skewed the actual function of the thing.
EDIT: It really overestimates on the lower end... really don't want people to be getting a false hope here.
I know, thanks and btw removed it from my original post. I knew it was off because I only have 2 sets of data to work with (my methods and physics scores) which are kinda high.
And actually, b^3, mine is completely different (which probs explains why it's so off) because I used a whole lot of normal distributions. I attempted to calculate the Standard Deviation of the total weighted marks (which vcaa saddly doesn't give on it's website). I think this is where I'm off because when I put in my actual marks from the statement of marks, lot of the SS predictors underestimate my methods and physics scores (by 3-4 points) so I skewed my version a bit to match my scores.
Also, adding Mao's and Paul's methods was more of a side note for myself to see what my predictor gives compared to theirs.
(ps I pm-ed you

)
PS: what subject gave 47 for B+s??? I just tried it and the max I got was 38 in the spesh predictor... if there's a number in the "score" column then the calc uses the number not the letter grade. Probs should have made that clearer at the start...