It’s been almost 3 months since the initial lockdowns and here are some of my thoughts on the pandemic.

The whole “this wasn’t the apocalypse that we were looking forward to” meme referring to zombie apocalypse and nuclear holocaust “fans” was sort of amusing.

Simply put, this is one of the few calamities that I have actually prepared for, albeit accidentally.

Much of my work and recreation doesn’t require getting out of the house, and I’ve learned to reduce my need for social interaction. My habit of getting premium equipment and backup alternatives also meant I don’t need to go out to buy things often.

I’ve even got random stuff that ended up being more useful in the situation. For example, that expensive mask I got for running – I switched out the filter for a carbon-filter one which I happened to buy and never got to use until the Taal Volcano eruption.

Another is the rechargeable beard clipper I got for my beard; it came with hair guards equivalent to No 5 and below. I was able to give myself the pandemic buzzcut that’s fashionable nowadays. /s

I can’t comment on the disease itself since I’m not an Epidemiologist. I am however, a software developer and I deal with large data sets all the time.

And pretty much all of the data shown on the news is useless from a data analysis standpoint.

Don’t get me wrong, death statistics have to be taken seriously, but the usual “Tested Positive, Recovered, Deaths” that’s been plastered all over the news is insufficient to make any informed decision from a data analysis perspective. Here’s just a small fraction of the problems with this “data”:

Tested Positive – most common flaw: the total number of tests performed is not mentioned, obscuring the infection rate. Another obvious flaw: calculating the percent of the country’s/municipalities’ total population infected/tested is left to the viewer. Other problems include not showing the breakdown of how many were tested via contact tracing (hints how the infection really spreads), and the test kit false positive/negative rate (hints the correct combination of test kits to use).

Recovered – the time of recovery (minimum, maximum, average) is never mentioned even though this can tell the public the actual severity of the infection. Also, the criteria for recovery is vague – what tests are done to categorize a patient as recovered.

Deaths – the criteria for counting a death is inconsistent among sources, basically there’s no strict rules whether to combine people who died from the disease and people who died with the disease. Patient privacy is important, but to make sense of the death statistics, we must know how these people died so agencies can make better treatment protocols, as well as decide which portion of the population to impose further quarantine measures on.

I could go on and on. And yes I’ve seen the site, but no it’s still not enough data.

You might say, “it’s ~500,000 tested and ~26,000 confirmed cases; there’s no way the public can handle that data, let alone a single software developer without a ‘data analyst’ title.”

To be frank, that amount of data is pretty small especially in this age of cloud computing. In fact, 26k data points is just a lazy Tuesday morning for me – finding the right terminology and tone for the report would actually take me far more time.

That’s just a software dev speaking. Think about what a real data scientist can do given full access to all of the data.

So why are these government institutions and news agencies disseminating such poor quality data?

I’m going to defer to Hanlon’s Razor here and assume at least 50% of it is due to plain incompetence, with the rest to further the narrative or agenda they wish to promote.

Except for US and US-aligned mainstream media.

That’s all definitely narrative pushing, seeing how quickly news about the pandemic plummeted due to the riots. And it’s going to be worse as we approach the November elections.

What will the near future hold for us?

I don’t know. I’m not an Epidemiologist.

Share →

Leave a Reply