| Author |
 Topic  |
|
Beanmimo 
"August review site"
|
 Posted - 03/30/2007 : 12:06:52
|
quote:
Originally posted by M0rkeleb
quote:
Originally posted by Beanmimo
So the only minor minor flaw is that very ealy on Leonidas is told he cannot face the Persians as it is Auguat, I presume it would be too hot at the end of the summer to wage battle.
But the name of the month August was first used five hundred years later in rome by Augustus Caesar.
As I recall, the reason for going to war had nothing to do with weather but because of some kind of religious observance.
Ah it'd been a long time since i studied that era. Thanks |
 |
|
|
Beanmimo
"August review site"
|
Posted - 03/30/2007 : 13:58:31
|
quote:
Originally posted by MisterBadIdea
Well, they also spoke Ancient Greek back then, not English. So I think that if they can use English words (a great deal of which have bases in Latin, not to mention much newer languages), I don't think it really matters.
As i said it was only really a minor glitch that they could have easily avoided.
And again I loved the movie. |
|
|
|
Shiv 
"What a Wonderful World"
|
Posted - 04/01/2007 : 12:41:44
|
My appetite is whetted. I just saw a TV review which made me want to see it even more...but my online DVD service doesn't have it yet I'm renting the 1962 film instead |
|
|
|
Sean 
"Necrosphenisciform anthropophagist."
|
Posted - 04/04/2007 : 02:17:47
|
quote:
Originally posted by demonic
It was doing well in the 250 earlier this week but it dropped out very quickly when the negatives and middling votes started coming in. I don't really get the system personally. Anyone want to explain a "true Bayesian" estimate in layman's terms?
I just had a look at the Wiki entry on Bayesian probability and am none the wiser. 
But I did test the formula (shown at the bottom of the Top 250 page) to see what it does. A movie needs 1300 votes in order to be considered for Top 250 inclusion, and the formula skews the score back towards the average vote for all movies on the site (6.7) for those movies with a low number of votes. So, a movie that has received only 1350 votes will have it's score 'dragged' back towards 6.7, this is done to reduce the probability that a small number of people could skew the score towards an extreme level. But, by the time a movie has received plenty of votes (e.g., 100,000) this skewing will become insignificant. Hence the Top 250 score for a movie with only a few thousand votes may be 0.3 or so lower than it's movie-page score, whereas the score for something like Pulp Fiction with over 100,000 votes doesn't vary from it's movie page score. Overall, the purpose of that formula is to 'stabilise' the Top 250 somewhat.
But, that forumla won't have a huge effect on the Top 250 score. The non-counting of 'non-regular' scorers will have a much more significant effect, I'd guess. This is presumably in addition to the filtering they do right from the beginning on scores (i.e., filtering out the scores of perceived 'score wreckers') or those who've only scored a handful of movies and can't yet be relied upon.
They don't publicise precisely what kind of filtering they do on raw data in order to stop vandals from trying to wreck the system with 'tactical scoring', as they state here. So, overall there's data filtering done initially prior to their "weighted average" being calculated for each movie, then more filtering done before a movie can be considered for the Top 250, then finally it's adjusted with the formula mentioned above to smooth the final score further to allow for different levels of voting and give a little more weight to those movies with more votes. Get it? |
|
|
|
Topic |
|
|
|
300
