Wednesday, January 06, 2010

Best of?

Dealing with statistical anomaly's all day I tend to have a problem with people who publish list like The Best Music of 2009 or the Best Movies of the last decade.

It implies that you have surveyed the entire population of music and movies in that particular time period and are giving an informed opinion.

Chances are, if the category is that broad, you have not.

I could see if it was something like The Best Local Bands of Medium Sized Town USA or something like that but how many people are really qualified to pick the Best of these huge categories?

Whats wrong with just saying, "Best Movies I Saw in the Last 10 Years?"

This is why I am a huge fan of raw data. For example if you only saw 12 movies last year and you are giving me your top ten out of those 12 then chanced are there are just some ok movies in there because the sample size is so small.

To frame up my recommendations

I read 50 books a year my population size goes back to 2003 for a total of 350 books.

I watch 50 movies a year going back the same period for a population of 350. So when I give my top tens they are out of number which may or may not mean anything to you because when it is all said and done it is just an opinion.

I just think the readers should be informed to how well that opinion was made and based off of what.

1 comment:

Thomas said...

My Top 5 favorite movies of 2009 (Inglourious Basterds, Up In The Air, A Serious Man, District 9, (500) Days of Summer) are skewed to the high-quality side as I avoid obvious garbage like Transformers 2 and Wolverine.