Missing Time Series

I just posted a rather uncharacteristic video on YouTube about my deletion of videos. Usually I confine myself to either instructional videos or presentations, but this was just a rambling self-explanatory monologue.

This is a text version of some similar thoughts.

In my work, my personal organizing practice is to scan for files that are ten years old (or more) and reviewing them to either delete or permanently archive as a memento of the past. The cutoff is arbitrary, but the principle is not. I believe that documents that have not been updated or touched in ten years and are no longer referred to or in use in any way represent thinking that can be discarded, or that at least no longer applies to the present. Please note that I am not talking about cultural or creative materials, only information created in a work context.

Sometimes this process generates future plans. If I see something that was not carried through on a decade ago, sometimes it is worthwhile reformulating it and taking another shot at it. If it was ten years old, it most certainly necessary to completely reassess the landscape and do a complete reboot. But most often, the old files get deleted, and I get to reclaim and reformat some brain cells by not trying to remember the details of old stuff that newer employees have never even heard of. I still retain the wisdom and experience from those past activities, but at least it is my hope that I can be more flexible and adaptable by letting go of the minutiae of bygone decades.

So this practice caught up to me when I realized that my “Time Series in R” instructional videos (and R script) were soon going to fall under my ten-year rule. It seems that I was one of the first to post on this particular statistical topic in YouTube back in 2013, because my “Times Series in R” videos are by far the most viewed videos on my channel, with one video netting 138,480 views. Those numbers are still enough to put the videos among the most viewed Time Series+R videos even today—1a topic whose interest seems to max out at around 250k views.

But—and once again there is a big but here—my time series workshops were never really popular at Rutgers. When I was giving them in person, only two or three people would show up, at most. I never got questions about time series topics. The end result is that I have not really touched anything to do with time series since about 2016. Also, the series was not very original to begin with, being a walkthrough of selected highlights from Introductory Time Series with R by Metcalfe and Cowpertwait.

I had nothing new or updated to contribute to a refresh on time series. While I view my videos as a nice addition to my main paid gig at Rutgers, I am primarily helping Rutgers students and creating content for Rutgers students, and am distributing videos via YouTube because it is both the easiest way to get this material out there and can reach a larger audience at no cost. While the likes and comments are gratifying, I do not engage much with YouTube comments, except to fix obvious issues that are pointed out, because I am not in a position to provide ongoing R support to the world at large. I am not trying to build a YouTube “brand”. In fact, for creative content, I plan to use Vimeo (coming later).

And so, in spite of their “massive” popularity, by my sheltered standards, I pulled the plug on these videos and deleted them. I have actually been posting instructional videos on YouTube since 2009, and had deleted some of those earlier, less consequential videos without hesitation. But this time, I felt guilty enough about it to record a video and write this post to explain what was going on, as proof that at one point in the now distant past, I was “big on YouTube”.