Monday, November 24, 2014

Chp 537. Mizoram and Pornography

So apparently Mizoram has the highest rate of online porn viewers in India, according to a research by Quartz published today, with data they claimed to have obtained from PornHub

Link - Quartz: Everything you wanted to know about how India watches porn in one map and five charts

And here are their findings -

According to the PornHub website data above, Mizoram has the highest “pages viewed per visit” in India at 8.47 per visitor which is higher than both the national and global average.

Though I am not saying this to defend or deny this bizarre "achievement" by Mizoram (hey, if we are indeed the top most porn viewers in India, then that's all cool brah, peoples do what peoples do) but I'm just curious to know how they reached this conclusion.

First of all, let’s take a look at PornHub and its ilk. For those of you who may not have heard of it, it is a popular pornography website where people can upload and share adult videos, and visitors can view them directly online (known as “streaming”). And that is where my first skepticism comes from. Internet penetration in Mizoram is not only low but the speed is pathetic most of the time. And it is a known fact that in such places where internet connection is bad, it makes more sense to download porn rather than watch them online.

Most of the people in Mizoram that I know who watches porn, do so by downloading them via torrent and not through online streaming sites such as PornHub, YouPorn, RedTube, xHamster etc because it apparently takes forever to buffer on our slow internet speed (which also has a frequent habit of disconnecting).

Yes I know you can download videos from PornHub too, but that means you have to register, which many aren’t willing to do. Plus most of the videos are low quality and if you want to download the ones with better quality, you have to become a premium member (yeah I too did my own "research" to find this out). So that is why torrent clients are preferred at regions with bad internet coverage such as Mizoram. And that may only show that many people watch porn in Mizoram, but this statement also contradicts the very premise of this “Quartz research” since their data is derived solely from only those who visit PornHub and not any other sources.

Secondly, I don’t know how accurate their data is. I know in most cases, one can easily find out which state a visitor is from, but sometimes IP addresses can be unreliable, especially when it comes to the North East. Here in Mumbai, we develop and publish mobile games, and I have downloaded a lot of our games on my friends and relatives’ phones in Mizoram. But those instances never show up on our Google Analytics report! Even when we use our own customized in-built analytics tool, some data from the North East are still faulty. That can be because the region is still developing or due to lack of infrastructure in the NE region (people who visit my blog from Mizoram sometimes show as Kolkata or Guwahati on my blog tracker), so I’m a bit surprised to know PornHub managed to get the absolute correct info (not that I’m saying that is impossible to do so). 

I really hope they are not establishing their data from the registration form where people can state anything they want, whether it is about their sex, their sexual orientation, their birthdays or their LOCATION. After all, how many people actually reveal true information at such websites?

A part of me can’t help but suspect the data is taken from this registration form only, otherwise how can they possibly know whether the user is a male or a female?

In India, a quarter of Pornhub’s users are female, which is 2% higher than the worldwide average of 23%

But then again, I can be mistaken too and there may be other ways in which they managed to obtain such information, so let’s leave it at that.

The North East still lacks behind the rest of India when it comes to online activities, which made me doubt PornHub’s data furthermore since the Northeastern states have some of the highest rates according to their findings. Just take a look at the ongoing ISL football tournament. If you visit their website during a match, you will see an Indian map with two different colours of the two teams currently playing. Those colours are representations of the tweets from respective fans, and you will see close to 3000, 4000, 5000 number of tweets during any match for most of the bigger States. But look at Mizoram or any other NE states and you will see just a handful of tweets. A region that is hardcore crazy about football and hardly any tweets? That's because internet is only just starting to pick up over there.

Having said that, the fact that they used "average duration of visit" and “pages viewed per visit” to come to this conclusion doesn't really hold water. Suppose two guys from Mizoram happened to land at PornHub separately in spite of the bad connectivity and they both visited 8 and 9 pages during their visit respectively, that means Mizoram scores an 8.5 as is given in the graph above. And then say 10,000 people from another state with fast internet connection landed at PornHub and they visited different number of pages ranging from 1 to 10, and the average value comes to 8, it gives Mizoram a much higher rating in spite of the fact that there were just two people visiting PornHub. But to a layman, that may sound like there are more number of people from Mizoram watching porn than other states, even though that is not the case. The population of Mizoram is less than the total number of people living here in Andheri East in Mumbai! There is a big difference between highest volume and highest percentage, and obscure statistic like this tends to mix up the two.

That is why Census India sometimes categorizes India into three different sections while comparing/ranking different regions. They are - “bigger states”, “smaller states” and “union territories” since they know the findings will be skewed or unfair if all States/UTs are compared together using the same yard stick.

By all means I am not trying to shine a positive light on Mizoram just because it is my state. If the research was about the highest percentage of people who watch pornography (the sources being online streaming as well as downloads from torrent, darknet, xxx forums, HDD to HDD transfer etc etc) and suppose Mizoram was on the top of that list, then there’s really nothing I can say about that. I will openly accept that result. I am not a prude :) 

But since this particularly report is based purely on the statistical data obtained exclusively from PornHub, I feel the result is inconclusive. Apart from the fact that streaming is not popular in regions with slow internet connection, the sample space taken also raises a question mark. I mean, technically, it may be correct (after all, even a small number of people has a high rate if the base is also small) but overall, it may not be the answer you seek and it tends to throw a wrong perception. It’s kinda like how your anti-virus detects your “keygen” as a virus even though it is not. “False positives” we call them. It would be more interesting to know the exact number of distinct pings from each state rather than just the rate if we were to truly compare and rank different states. Just my opinion.