- Joined
- Feb 23, 2018
- Messages
- 16
Hey folks, I've been seeing windpants get more popular for a little while now and wondered whether they actually were. I decided that one possible way to measure this would be windpants-related google searches as a portion of all simple fashion-related searches on google. The underlying assumption is that if people see people wearing windpants or see them in the media, they're more likely to search about them. I've included over 30 keywords, all of them mean specifically the nylon type of track pants, not the cotton type. Here is a 12-month moving average of a transformed version of the result:
Don't worry about the dimensions, it's a more complicated calculation than I want to go into here and the result is dimensionless. The early part of the Canada series is noisy and I don't think it reflects the true popularity. I believe that the US series is accurate. Here are the growth rates for the US, both the basic series and its moving average in orange:
Being YOY and a moving average, the dates don't truly line up with the trend, but they are accurate to within 6 months. The dimensions here are meaningful. By this analysis, windpants were 20% more popular in the fall of 2016 than they were in the fall of 2015. I don't think that the 30% spike in 2006 is real, it is probably an artifact of the noisy data from 12 years ago (it comes from the unexplained dip in winter 2005-2006).
I think that these two charts line up really well with my subjective experience. I feel like windpants have been becoming more popular since about 2015, but aren't nearly back to their early-2000s level. What do you think?
Don't worry about the dimensions, it's a more complicated calculation than I want to go into here and the result is dimensionless. The early part of the Canada series is noisy and I don't think it reflects the true popularity. I believe that the US series is accurate. Here are the growth rates for the US, both the basic series and its moving average in orange:
Being YOY and a moving average, the dates don't truly line up with the trend, but they are accurate to within 6 months. The dimensions here are meaningful. By this analysis, windpants were 20% more popular in the fall of 2016 than they were in the fall of 2015. I don't think that the 30% spike in 2006 is real, it is probably an artifact of the noisy data from 12 years ago (it comes from the unexplained dip in winter 2005-2006).
I think that these two charts line up really well with my subjective experience. I feel like windpants have been becoming more popular since about 2015, but aren't nearly back to their early-2000s level. What do you think?