January 2015 in review

This is the fourth of my month-in-review posts. My posts for the past three months (in reverse chronological order) are here, here, here respectively.

Some overall background

I had some time off my job in December 2014, due to end-of-year leanness in work, and I used this time to draft many Open Borders: The Case blog posts. Work has reverted this month to a level comparable with November or the first half of December, and is likely to increase to unprecedented levels in the next few months. My high apparent output on various projects in January is largely a result of material I put in the pipeline in December, as well as due to a shift in priorities away from relaxation and entertainment and towards online publication.

Wikipedia editing

General background information: My site page about my Wikipedia contributions

I created a number of Wikipedia pages this month, including:

  • Chartbeat, a real-time analytics service.
  • AirPair, a service that connects people who need coding help with those who can help them for one-on-one guidance.
  • ZenPayroll, a company that facilitates payroll, and is particularly cost-effective for small companies.
  • Zenefits, a company that facilitates group health insurance and human resource management for companies.
  • AdChoices, a program for advertisers to offer users choices on whether they personalize the ads they serve to users.
  • Labor Condition Application, an application whose approval needs to be submitted for H-1B, H-1B1, and E-3 petitions.

I got interested in these topics in the course of research or reading I was doing in relation to personal projects, my job, and bureaucratic formalities.

About half of these pages were made during BART rides, at the rate of about one page per ride (the ride for me, between the Ashby BART and the 16th Street/Mission Street BART, lasts about 30 minutes). Although I don’t have Internet access on BART on my laptop (my phone has its own Internet access but can’t create a Wi-Fi hotspot), I opened up a bunch of related links just before departing, and pieced them together during the ride. (See also my efficiency tips page).

I also updated some timeline pages on Wikipedia:

I did the first three updates on Sunday January 25, over a stretch of time of approximately 1.5 hours. The next two were done during BART rides, using the modus operandi described above. The last one (the timeline of PayPal) took several BART rides, because I had several years of information (2009-2014) to fill, and PayPal is a lot bigger and more important than Airbnb and Pinterest.

I also made some changes to the Optional Practical Training Wikipedia page in the course of research I did for my blog post about Optional Practical Training.

I had toyed with the idea of making Wikipedia pages on many other topics, but ultimately decided against it due to lack of time. Some of the pages I shelved aside were:

  • Company pages: Instacart, SimplyInsured
  • Timeline pages: LinkedIn, Uber, Zipcar, Zynga, Instagram, Secret (app)
  • Visa pages: H-1B1 visa (a variant of the H-1B visa), the N visas

Given the increasing pace of my work, as well as my other commitments, I don’t know if I’ll get time to pursue any of these possibilities in the near future.

One of the reasons I like using the BART rides for Wikipedia editing rather than other tasks is that it’s relatively mechanical and doesn’t really require me to be online. Another candidate for BART rides is Open Borders: The Case blog posts, but these generally require some Internet access insofar as drafting the posts raises new lines of inquiry that need to be answered through a web search. I’ve also used BART rides for coding work related to my job, but that often requires more concentration and I generally do it only if I really want to get it done or I’m not too fatigued.

January impact: Over the month, Wikipedia pages I’ve created over my lifetime got a total of 121,366 views. You can access the full details here. One of the pages I created, namely Park Yeon-mi, has been edited beyond recognition, and it also had a large number of pageviews (9,588 views over the month). Therefore, it may be fair to exclude it from the total, bringing the total down to 111,778 views.

For pages I have significantly edited but did not create, the total pageview count in the month of January was 100,228 (you can access full details here).

February forecast: I expect a fairly similar number for pages I created, but a spike for pages I edited but didn’t create, largely due to Facebook’s anniversary driving traffic to the timeline of Facebook page. But that spike will be only a few thousand pages and may be overwhelmed by other factors.

Controlling for the shorter length of the month, I expect the February number for pages I created to be modestly better than the January number, because some of the pages I created at the end of January have been getting a decent number of pageviews. My February point estimate is 117,000, my 80% confidence interval is 105,000-130,000, and my 95% confidence interval is 95,000-150,000 (revise all numbers downward by about 10,000 if the Park Yeon-mi page is to be excluded).

Subject wikis

General background information: My site page about the subject wikis

For a few hours one weekend, I spent some “relaxation” time browsing and making edits to Groupprops. I also fixed some errors in Groupprops and other subject wikis reported through the online error reporting form. You can see the error fixes in the site’s error log.

Once I’ve accomplished some specific goals with my work on Open Borders: The Case, I intend to devote more systematic time to the subject wikis. This time will probably be directed at Calculus, Market, and a nascent machine learning wiki, but I will probably keep returning to Groupprops, if only to revisit some of my organizational choices there when seeking inspiration for the newer wikis.

January impact: I’ve put the numbers for Groupprops in a table for easy perusal and comparison. Numbers are from Google Analytics.

Metric January 2015 December 2014 January 2014 December 2013
Pageviews 55,121 77,725 41,989 60,232
Sessions 26,717 36,147 21,466 28,769
Pages/session 2.06 2.15 1.96 2.09
Pageviews in sessions with at least 5 pageviews 19,408 28,641 13,166 20,793

As you can see, January numbers are generally lower than corresponding December numbers, but there has been robust year-on-year growth and the January-to-December ratios have been about the same between the two years. (If you’re wondering why the January number is lower than the December number, this has to do with the nature of the academic year. Traffic to Groupprops is higher in the months that people are more likely to be studying and reviewing group theory in their college mathematics courses).

February forecast: February 2014 saw 57,925 views. Assuming the overall trend of a 20-35% year-on-year increase continues, the February pageview count will be in 69,000-80,000 pageview range. This is my 80% confidence interval. My 95% confidence interval is 55,000-90,000 pageviews.

Open borders

General background information: My site page about open borders

Open Borders: The Case had its third highest traffic month (the two highest traffic months were November and December 2014).

A number of my blog posts (most of them drafted in December) were published on the site in January. Each post took about two hours for primary drafting and then 1-2 hours of editing and improvement spread out over several weeks:

I also revamped some of the site pages, such as double world GDP and place premium.

Open Borders: The Case will publish its own monthly review. I’ll link to that from here once it is published.

Server downtime

The server where I host vipulnaik.com, the subject wikis, and Open Borders: The Case had intermittent troubles through the month. While I’m still not sure of the core causes of the troubles, it seems like one possible culprit was logging to large log files that was slowing the system down. I made a number of server configuration changes that should hopefully address the downtime. Estimated traffic loss due to the misconfigurations is about 1%.

Bureaucratic activities

  • I wrote three recommendation letters for students applying for scholarships and internships. These are students I taught in 2013 or earlier while at the University of Chicago.
  • I finished my US taxes for 2014 and intend to file them next Saturday. I have it ready for filing but want to give it a closer look before I do the final sending, because errors are hard to correct. A few unexpected issues that arose: it took me time to figure out how to handle 1099 contractor income (but I finally figured it out). Second, just as I thought my taxes were done, I got a W-2 from the University of Chicago for a payment they made to me in January 2014, that had been associated with work I had done in December 2013. I needed to incorporate that information into my tax return.
  • I started corresponding with the company lawyers regarding my H-1B petition.

Other websites and activity

Just for the record, here are pageview counts (total for the month of January) of my other websites:

  • My personal website (vipulnaik.com) where you’re probably reading this review: 2,432 views according to WordPress, 2,351 according to Google Analytics.
  • My Quora stats (total views of my questions and answers): About 16,400 views.
  • Cognito Mentoring, a website made collaboratively with Jonah Sinick (see here): 1,012 views according to WordPress, 951 according to Google Analytics.
  • Cognito Mentoring Info Wiki: 1,946 views according to Google Analytics.

Awesome stuff by people I know, that I might have influenced or might be able to influence

The people I know have been up to some exciting stuff this month:

  • Issa Rice, a former mentee of Cognito Mentoring (read his review here) has been making progress on the Cause Prioritization Wiki, that he aims to make the web’s go-to resource for effective altruists (and others) interested in understanding and comparing top causes to invest time and money in. I take credit for being the one to originally suggest cause prioritization as a promising domain for him to look into, and encouraging/nagging Issa to execute on his vision in the area. The wiki recorded 418 views in Google Analytics. However, this does not filter out Issa’s own views.
  • Katja Grace officially lanched the AI Impacts website earlier this month, and has been busy blogging there. I played no particular role in creating or funding the site, but Katja has been keeping me updated of progress on it, and I’ve offered the occasional suggestion. I did suggest to her that this was a more promising thing to focus on than other alternative activities (although she likely would have chosen to do it anyway). This recorded 1,673 views in January since Google Analytics tracking was launched on January 12.
  • Sebastian Nickel, one of my closest Facebook friends, officially launched a blog section of his website.

Personal: diet, sleep, exercise, and entertainment

For more on my daily routine, see here (password-protected, but feel free to let me know if you’re interested)

My sleep patterns were a bit disrupted. I tend to wake up based on sunlight, so the fact that the sun rises later in January meant that I’d tend to wake up later. This in turn led me to sleep later. It didn’t help that in the last week of December, I often stayed up late during the holidays as I was pondering some matters.

The late waking up meant that on about a third of the weekdays this month I didn’t have time for my morning jog. However, towards the end of the month I’d started getting the situation under control, and I hope to make my morning exercise a regular and reliable fixture of my existence once again.

Diet patterns were good — I cooked enough that I didn’t need to eat out except once on Saturday and the company lunches. Although I still went out on occasional dinners with other people, this was a matter of choice rather than of running out of cooked food at home. On the other hand, I did end up having more expensive hot chocolates in San Francisco than usual.

As had been the case in December, I didn’t buy or watch any new movies or read any new books. I did buy one new MP3 song. I also kept some old movies and music bought earlier running in the background when I worked on weekends and occasionally in the office.