Hacker News Reader

Comment by pg | Hacker News Reader

There was plenty of driving before the interstates. In the unlikely event this explanation is correct, it should be easy to test: compare the graphs of car and hat sales.

View on HN

Just like how the lack of pirates caused global warming, right? ;) http://en.wikipedia.org/wiki/File%3aPiratesVsTemp_English.jp...

That graph makes zero sense. Is the background gradient supposed to be the pirates? Otherwise they are not shown at all. ;)

The x axis is number of pirates.

To be precise, the x axis is time. It's simply labelled with the number of pirates at that time.

It is common to have time on the x-axis, but that doesn't seem to be the case here. The dots on the line are equally spaced horizontally, but the number of years between each dot varies, sometimes it is 20 years, sometimes it is 40 years, so I don't think the horizontal location represents time.

I'd call this a parametric plot, useful when you have two quantities which vary over time, and you want to visualise how they relate to each other. There are some good examples of that on the Ladder of Abstraction demo, http://worrydream.com/LadderOfAbstraction/ The "Abstracting Over Time" graphic is a parametric plot of the position of the car, where x is horizontal position, y is vertical position, and distance along the line is time.

Oh, I somehow missed that half of the intervals were only 20 years.

But it's not a parametric plot either, because the number of pirates goes up and then down.

I guess it could plausibly be a normal graph with the X axis being some silly function of time; or a parametric plot with the X axis being a stupid function of pirates. But really I think aw3c2 was right: the graph makes zero sense.

Here's a better version of the plot: http://i.imgur.com/aXQei.png

This type of plot can be confusing when the relationship isn't strong, but I think it works well enough here.

edit: there's also this plot type: http://i.imgur.com/jgMFS.png

It's a parody of religious right wing magical thinking, so it's not supposed to make sense.

Is this going to end up similar to a pirates vs. global warming graph?

As hat wearing went down car sales went up?

Well duh comparing the graphs doesn't provide the answer. But it provides good negative evidence if hats worn declines before interstates, for example. And if the graphs are similar, its a piece of the puzzle towards the answer.

It's probably not a waste of time creating a graph that might increase understanding and create better questions?

Not to be too funny, but humans are great at finding patterns in pictures that don't exist.

http://static.flickr.com/54/139092366_ce5b410228_o.jpg

Suddenly hatless men found that they were getting sunburned. They needed cars in order to get quickly from one place to another.

The driving before the interstates happened on dirt roads, country roads, or paved roads in crowded city thoroughfares, with top speeds of less than thirty miles an hour. Automobiles in the '20's sat tall, with lots of interior room (even the low-end Model T had plenty of vertical space); sedans in the thirties began the trend towards lower profiles which continued up through the sporty road cruisers of the fifties. Those sleeker, lower profile designs would have been pointless without highways.

I don't think it's an entirely unlikely explanation.

True, but I am forced to say, correlation != causation ;)

It would provide evidence, however.

-/-> you mean.

Kind of like rising CO2 and global temps.

well, i don't have sales data at hand, but googles ngram viewer has this to say http://books.google.com/ngrams/graph?content=hats%2Ccars%2Ch... (no correlation, at least in books)

I downvoted you by mistake, sorry.

I'd say it's harder to find correlation in Fashion than to predict the market.

Everyone i know in design and marketing everyone blabs about how lack of marketing killed hats.

But no one i know also happens to have breakfast talks everyday with a anthropologist phd studying fashion.

It also doesn't explain why the 5 million[1] people who ride the subway in New York City every day don't wear hats.

[1] http://mta.info/nyct/facts/ridership/index.htm

That style is very trend driven would argue otherwise. Once you reach a threshold suddenly it's unfashionable to wear hats or floral dresses.

There's actually some really interesting literature about how copyright would significantly hurt the fashion industry, because it is dependent on a large degree of 'copying' styles to create trends, which is key to creating planned obsolescence to keep people buying new clothes.

Edit: A nice TED talk on this, starting at 3:57 http://www.ted.com/talks/johanna_blakley_lessons_from_fashio...

Reply on news.ycombinator.com