Tags

, , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , ,


RobotsAfter publishing yesterday’s post titled Who You Calling “Small Fry”?, a friend of this blog who writes under the nome de plume “Foolishness”, left me the following comment:

You can find objective website traffic rankings through alexa.com and if you register, (you can do that for free), you can do site comparisons and they will provide a convenient little graph so you can see how pewsitter,com, Father Z’s blog, and Cruxnow.com compare. :Last i checked Crux is getting more traffic on the web than either, but far less than the solid pro-life LifeSiteNews.com.

So naturally, since this blog is the Catholic blog of a Catholic blogger, who believes in a Catholic God, triumphalist doctrinal certainty, including but not limited to the Eight Commandant and “objective truth”, a follow-up post was in order.  Therefore, not wanting to spend the rest of eternity in … ahammmm, I quickly went to the alexa.com website, registered and pulled up the websites in question to check out if my initial observations were correct.

And low and behold, it appeared as if some of the initial conclusions posted in the Who You Calling “Small Fry”? (see here) were not correct!

Introduction

The part that appeared to be not correct was the assertion that the “offensive NYT/BG” website had lower visitors/page views than either the Pew Sitter website or the Father Z blog.

So I put up the three websites, side by side using the alexa.com functionality and here is the resulting graph:

alexa-rank 1 - CopyIn the above graph, the numbers are not “visitor” numbers or “page view” numbers, but rather rankings of this site relative to other sites. Alexa.com does not provide visitor data or page view data to the general public. Here is the explanation:

alexa 3I put up the above in the spirit of full disclosure.

But back to the narrative. From the above graph, it would appear that the following is the case in ranking table form:

1) “Offensive NYT/BG” website -> 

61 345 globally and 15 769 in the US

2) Father Z bolg ->

93 042 globally and 23 187 in the US

3) Pew Sitter ->

312 497 globally and 69 047 in the US

In other words, according to the popularity ranking, i.e. the higher the ranking, the more visited the website, it would appear that the “offensive NYT/BG” website has a much higher visitor rate and a much higher absolute number of page views than either Fr. Z’s blog or Pew Sitter.

Which raise a whole host of questions, which your humble blogger will attempt to answer in the below text.

Questions, questions?

In the previous post, we observed the following:

Here is a table that I mocked up from the screenshots:

Fr. Z VIIWhat we can easily observe in the above table is that within a span of a day ( actually, it was only a few hours), the Fr. Z  visitors that he directed to the “offensive NYT/BG” website were able of “completely reverse the voting outcome” of the poll at the targeted website.

If the above is correct, i.e. if the “relative popularity” of the “offensive NYT/BG” website is in fact that much greater than the Fr. Z blog (i.e. many more visitors/page views), than how is it possible that the request of Fr.Z for his visitors to go over to the “offensive NYT/BG” website and vote would have had the effect on the order of magnitude that we have observed?

Figuratively speaking, using phraseology one can find in your typical RAP composition, one could say that Fr. Z “owns this “offensive NYT/BG” website like he would own his “female member of the canine family”.

Therefore, what we can clearly identify in the above are issues with the “veracity” of data.

Three issues

From the data above, three issues become readily apparent, issues that need to be addressed.

1) One issue (and a possible explanation) could be a faulty assumption on our part. Under our methodology, we assumed that the behavior patterns of both sets of visitors will be identical. This appeared to be a reasonable assumption at the time.

If this is not the case, the explanation for this different breakdown of poll results between the 18th and 19th of January could be as simple as a difference of the voter profiles between the the visitors to Fr. Z’s blog and the “offensive NYT/BG”website. Or to be more precise, the typical Fr. Z visitor “likes to vote” in website polls while the visitor at the other website “does not”.

However, I find this explanation hard to accept. The subject of the poll was an “emotionally” charged issue, i.e feminization of the Catholic Church. One would expect that both camps, i.e. the Catholics and the Modernists would want to get as favorable a result as possible for their respective positions. Support for this contention is provided by two minor observations, i.e. the issue was so “charged” that it got Fr. Z to ask his visitors to cross over and vote, and 2) the votes cast for the “neither option” (Maybe) was miniscule. In other words, all voters in this poll had an ax to grind. So the “different” voter profile explanation between the two sets of voters can be discarded rather easily.

2) The next explanation that could explain this anomaly is that the majority of visitors on the “offensive NYT/BG” website did not vote. In order to further explain, here is a graph that represents the “relative” popularity of the “offensive NYT/BG” website against its owners, i.e. the NYT website and the Boston Globe website:

alexa-rank-2 - Copy What we see in the above graph is that the “offensive NYT/BG” website had a period of “above average” growth between the end of September 2014 and the end of December 2014. A social scientist or a business student would easily recognize the underlying function which depicts the growth stage in a business life-cycle. But what would strike that social scientist or business student as odd, is the time horizon in which this growth phase took place. From the graph above, it would appear that his “offensive NYT/BG” website went through its growth phase in three month, i.e. one quarter. In month 4, the “offensive NYT/BG” website was already in the mature phase of its business life-cycle.

Therefore, it is not unreasonable to suspect that the visitors/page views numbers that are being generated on this website are being generated by… shall we say ” “processes that do not arise from organic growth”. Therefore, it would be reasonable to assume that the visitors to the “offensive” NYT/BG website that are a part of this “non organic growth” would not care to vote in this poll. More about this in a bit…

3) The third explanation that can be posited here can be seen in the below two screen shots, The first one shows the results after the Fr. Z intervention:

Fr z IIIThis second one shows the results after three days time from the launch of the poll and two days time after the 1st Intervention from Fr. Z’s visitors:

Fr z ivWhen comparing these two tables, what we notice is the following: the Fr. Z voters are piling on and pushing the results more in their favor. In this two day time period, the Fr. Z voters were able to take the Yes result from 69.84% up to 75.03%. Please keep in mind that the number of marginal votes needed to move the results by 5% from 75% to 80% is much greater than that needed to move the 5% from 50% to 55%. And this increase in the YES column is being done by stragglers who did not visit Fr. Z’s site on the 18th and whose vote was not part of the original cross over vote. Or to put it another way, this 5+% increase was generated by visitors who clicked into a one day, two day or three day old Fr. Z post. In other, other words, this result is a blow out!

Summa Summarum

Concluding, it would appear as if the initial group of Fr.Z inspired voters that came across and voted, established the distribution of the final poll and that is the way the results stayed.

Furthermore, there was either: 1) no “motivation” for the Modernist camp visitors to the “offensive”NYT/BG website to enter the fray in order to try to change the results or 2) no “new unique” visitors appeared that could participate in the poll and counter the Fr. Z cross over voters.

If the latter is the case, which is highly probable, it would appear as if there is a “set” group of visitors who generate visitor traffic on a daily basis at this “offensive NYT/BG” website. The support for this contention is that once those that could vote in the poll cast there votes on the 18th of January, there were no more unique voters who could participate after the initial round of voting to counter the Fr. Z cross over voting traffic.

Therefore, it is quite obvious that once the original “set”group of voters cast their votes in the beginning of the poll, the overwhelming majority of the remaining voters that cast there votes came across from the Fr.Z blog. This would explain why in the two days from the 19th to the 21st, the results went even further into the Catholic camp than they were after the 1st Intervention by Fr. Z. This could have been caused by straggler visitors to the Fr. Z blog going over in the next two days and casting their vote at in the poll conducted by the “offensive “NYT/BG” HUGE media conglomerate website.

Which brings this blogger back to the “composition of visitor/voter” issue. After the above evidence is examined, it is still highly likely that the original results were accurate, under a slight correction to the definition of our assumption.

From today’s analysis, it would appear that the “offensive NYT/BG” website has approximately ≈2000 legitimate voting visitors for every ≈2900 visitors to Fr. Z’s blog that vote in his polls.

As far as the rest of the traffic to the “offensive” NYT/BG website, hard to say who or what they are.

Could be robots for all we know.

Advertisements