Saturday, September 12, 2015

Forgotten Movies of Greatness: Gattaca

When I saw Gattaca back in 1997 I called it sci-fi film noir (a characterization I stand by). I also called it fantastic: a really compelling story, rendered by solid actors, gripping direction from newcomer Andrew Niccol and sharp visuals (costumes, art direction and visual effects). Most critics reacted favorably as well; it garnered an 82% "fresh" rating from Rotten Tomatoes.

Roger Ebert called it "one of the smartest and most provocative of science fiction films, a thriller with ideas."

 I told people it was one of the best 10 movies of the year and deserved 4 or 5 Oscar nominations (it received one for Art Direction, but should have been in contention for Costume Design, Original Score and Original Screenplay at least).

Ultimately very few heeded my recommendation. The movie made only $12.5 million of its $36 million budget. Aside from a Rand Paul speech in 2013 it's been relegated to obscurity ever since. Which really sucks because the movie is good.

Friday, September 11, 2015

Song of the Day: "Molecules"

Above: molecule
Even the most cynical among us probably digs that light fixture.

Song: "Molecules"
Artist: Atlas Genius

Wednesday, September 9, 2015

Song of the Day: "Downtown"

Macklemore (& Ryan Lewis) "Downtown" lyrics
Is Ryan Lewis the new Art Garfunkel? Discuss.

Song: "Downtown"
Artist: Macklemore & Ryan Lewis

Measures of Central Tendency, the Golden Girls Way

Statistics, they're just a bunch of hooey and they always lie and, aside from keeping me from doing manual labor for the last 25 years, they're no good to anyone.

Or are they?

Probably not, but in the slim chance someone might derive value from them, here are the 3 measures of central tendencies, as illustrated by the Golden Girls' sex lives (data from the inter-webz so you know it's right).

Total Number of "Dates" per Golden Girl
Mean: The first measure of central tendency is the arithmetic average of a set of numbers. It is calculated by adding up all the numbers in a data set and dividing the sum by the number of numbers you just added up.

Whathefucknow you may have just said... it's the total of the numbers divided by the count of the numbers. It's routinely called the "mean" or "average" and it's a perfectly valid way to summarize some characteristics of a group.
165 + 43 + 30 + 25 =263
263 divided by 4 is 65.8
What to watch out for? Outliers: individuals or individual cases with an unusually large or small characteristic. Using the Golden Girls as an illustration, the number of "dates" Blanche had is more than the number for the other 3 combined. While her number is laughably small for a pre-Will & Grace* gay guy, she is an outlier among these women. Removing Blanche's number yields an average of 32.7.

Outliers exist in non-bounded data (income, sex partners, miles driven in a lifetime). They don't in bounded data (0-to-10 scales, how many days a week someone rides public transportation [0 to 7], how many hours you typically go online in a week [0 to 168, the total number of hours in a week]). Although for the hours online in a week, a number like 168 means that person didn't sleep or shower or do anything else be use the internet, so scrutinize your data and feel free to question the validity of certain responses.

The most popular outlier in our society these days? The 1%: the ultra wealthy whose concentration of wealth tips the scales of any analysis of income in America. I offer this for context and not to make any specific point on policy.

Median: The midpoint of the data when you sort it. It helps to have an odd number of cases so you have an actual midpoint. When you have an even number of cases (like we do here) use the average of the 2 middle numbers. In this case 43 and 30 (43+30=73, 73/2=36.5) the median is 36.5.

The median is a good contrast with the mean since the median isn't distorted by outliers.

Fun fact: If someone is trying to cite a statistic and mentions the medium household income or medium age, that person is an idiot.

Mode: The most common data point. This statistic is cited less often than the prior 2 but that doesn't mean that the mode isn't regularly referenced. If you've ever heard someone say something like "the average resident of Dallas is a 46-year old white female" you might wonder how non-numeric characteristics like gender and race can be averaged. They can't; 46 can be an average age, but white and female can't be means or medians. It's the statistic to use when you have categorical data, but it doesn't work well for this scenario.

Here we have 4 data points and they all occur equally (these are the numbers 165, 43, 30 and 25--they each occur once). However if we take a look at the data from a different perspective, say we asked each of these 263 men [assuming none of the women had a man in common] who of the 4 they slept with (or slept with first, to deal with potential overlap) the mode would be Blanche. It would be mentioned in a summary as "the modal sex partner among the Golden Girls is Blanche" and everyone would nod knowingly.

Come back soon when I attempt to explain null hypothesis testing using the Muppets.

* Will & Grace was by no means a groundbreaking television show, but its late 90s/early aughts airing, just after the culturally relevant Gay 90s (from Angels in America to Don't Ask Don't Tell, the rate at which serious shit happened (and not all of it good) for the 'mos and 'beans et al between 1990 and 1999 makes it a point of inflection for American society. The tendency of gay and/or closeted "straight" men to have furtive anonymous sex in places like mall restrooms was a response to a sense of semi-invisibility and lack of normalcy leading up to that time. Courageous kids tried to take same-sex dates to their high school proms before then, but that was considered outrageous, inappropriate. After Will & Grace it's considered common. I don't credit Will & Grace with this (its existence is evidence of it) but using the OJ Trial or Tiger Woods's first Masters win as a reference point seems really confusing.

Monday, September 7, 2015

Song of the Day: "Pressure Off"

Once upon a time there was a band named Duran Duran and they were quite popular. And seemingly every time I completely forget they exist they release a new album. Often these albums are good.

With this single they've reunited with the awesome Nile Rodgers (producer of their biggest hit, "The Reflex," along with other tracks and albums and such). Mark Ronson and Mr. Hudson also receive producer credits on the album, Paper Gods, potentially a good sign beyond this track.

Janelle MonĂ¡e joins the "wild boys" on this track. Unlike the fun. song, I can actually hear her here.
Their limited US dates bring them to the Washington State Fair this month... wonder if it's worth a trip.

Song: "Pressure Off"
Artist: Duran Duran

Sunday, September 6, 2015

Song of the Day: "I've Done Bad Things"

If you saw This is 40, Judd Apatow's alledged comedy about miserable people being miserable and making each other miserable you might recall a fictitious band called Graham Parker & the Rumor as a key plot point (namely the band destroys Paul Rudd's record label financially). In an amazing coincidence that band was played by members of an actual band called Graham Parker & the Rumor Actually the movie is something of a love letter to Graham Parker & the Rumor, including their first reunion in 30+ years as a part of the movie a month before their reunion album, Three Chords Good, was released. Had the movie been good it might have been construed as an actual love letter.

Mystery Glue is the follow up to Three Chords Good and both received similar acclaim (they both received a Metacritic score of 80); both albums were better received than This is 40. It really is a difficult movie to watch, let alone enjoy.

Luckily you can enjoy the albums without trying to watch the movie. So do that.

Song: "I've Done Bad Things"
Artist: Graham Parker & the Rumor