My understanding is that you use each artist's six highest-ranked albums and six highest-ranked songs, with albums counting twice as much as songs...and you derive some kind of logarithmic function for the rankings, correct?
If you can explain how you do that, it would be a tremendous help to me. I'd like to be able to duplicate your process for figuring artist rankings for my next Big Project (coming this summer, once Survivor is over).
Of course, if you consider it proprietary technology, I understand.
schleuse – why is it, I picture you all mad scientist-like, laughing manically over facts and figures while hatching your next hairbrained idea? And your wife walks into the room, and says “Hey hun, whatcha’ workin’ on?” you quickly shuffle papers into a desk drawer and reply “Nothing sweetie, just go back to bed.”
It appears that I've given you the wrong information earlier, as I'm counting songs 0.6 times as high as albums, and not 0.5.
'album1' in the formula means the AM position for the most acclaimed album for an artist. If an artist has, say, only 3 albums in AM top 3000, then artist4, artist5 and artist6 are set to 3100.
It is nice to see what you can do with that formula, e.g. schleuse's best artists' years.
Henrik, I remember asking you sometime last year about an easy way to calculate the best years. Well, your formula does the trick alright. Results coming up soon
It's the natural logarithm, the inverse function of the exponential function !
Or if you ate maths, just consider it is an excel function which give you a number if you give him an other one
It appears that I've given you the wrong information earlier, as I'm counting songs 0.6 times as high as albums, and not 0.5.
'album1' in the formula means the AM position for the most acclaimed album for an artist. If an artist has, say, only 3 albums in AM top 3000, then artist4, artist5 and artist6 are set to 3100.
What if an artist never released any real albums? (Robert Johnson, for example)
It appears that I've given you the wrong information earlier, as I'm counting songs 0.6 times as high as albums, and not 0.5.
'album1' in the formula means the AM position for the most acclaimed album for an artist. If an artist has, say, only 3 albums in AM top 3000, then artist4, artist5 and artist6 are set to 3100.
What if an artist never released any real albums? (Robert Johnson, for example)
Well, how 'bout that... the origin of the Mad Scientist nickname!
(That was over two years ago? Christ, I'm getting old.)
You and me both. I'm still amazed that, over a two-year period, I always had at least one labor-intensive project going on.
I cannot imagine fitting that into my schedule now, although occasionally I'll look at someone like Matt Schroeder or listyguy and think the grizzled vet needs to show the young'uns he's still got his fastball. Maybe one of these days...
Let's say I wanted to aplly this formula to my favorite songs and albums lists. How owuld I go abot doing that? (I am planning on taking a statistics course nest year, but until then, I'm in the dark.)
+ if an artist have less than 3 albums in your list, you should put something around "total number of albums you ranked +100" for the missing values.
Same for songs.
I personaly had a top 200 albums and top 500 songs and I think I used "250" and "550" as settings but I guess "300" and "600" would have not changed much the results.
I know that part, my question is how to do "in". Can you do that on a graphing calc?
For example, the formula applied to the Beatles would be (roughly): ln(ln(10+6)) + ln(ln(10+11)) + ln(ln(10+17)) +
0.6 * (ln(ln(10+6)) + ln(ln(10+33)) + ln(ln(10+42)))
simplified:
ln(ln(16)) + ln(ln(21)) + ln(ln(27)) +
0.6 * (ln(ln(16)) + ln(ln(43)) + ln(ln(52)))
It's Ln not In, as in Natural Logarithm (I guess it is LN and not NL because it is a german mathematician who named it). We learn about it in "première" en France I think, which is your 11th grade.
Anyway, if you put
= ln(ln(16)) + ln(ln(21)) + ln(ln(27)) +
0.6 * (ln(ln(16)) + ln(ln(43)) + ln(ln(52)))
in Excel, you get 5,556915564
It appears that I've given you the wrong information earlier, as I'm counting songs 0.6 times as high as albums, and not 0.5.
'album1' in the formula means the AM position for the most acclaimed album for an artist. If an artist has, say, only 3 albums in AM top 3000, then artist4, artist5 and artist6 are set to 3100.
If you want I could upload an Excel file I also use that has a sheet to list your albums, a sheet to list your songs and finally a sheet that takes data from the first two sheets and does all the artist calculations.
If you want I could upload an Excel file I also use that has a sheet to list your albums, a sheet to list your songs and finally a sheet that takes data from the first two sheets and does all the artist calculations.
Thanks, but I still need to make my albums list. Maybe before the next poll.