1. Oystein

    Oystein Active Member

    This article is not intended to "debunk" any specific claim, but to present information on the Facebook page of AE911Truth - as per yesterday, August 28th 2019.

    They today have over 500,000 "Likes" - quite a feat in a "9/11 Truth" world that nowhere attracts more than 25,000 followers (as measured by signatures, members, "likes").

    AE911Truth created their Facebook group on February 28, 2009 - 10.5 years ago:
    Architects & Engineers for 9/11 Truth (on Facebook)
    I have been keeping a tally of the "Likes" to this page since about December 2014. I have started to tabulate the "Likes" number almost daily on April 1st, 2015. Before that, I have occasional records - here is a graph of my 13 datapoints between February 2009 and April 1st, 2015:

    FB AE Likes 20090228-20150401.

    You see that, in the first 3 1/2 years, they got 88,725 signatures - about 70 per day - and after that, the curve was steeper at about 120/day. Probably what you could expect, seeing that Facebook was growing in popularity.

    Note that from March 13 to March 14, 2015, they lost 10,456 signatures in one single day. I have no idea what happened then. An intervention by Facebook? Of a general nature, or targeting AE911Truth and similar pages? One can only speculate.
    A similar, but not as big, short-term loss occurred between April 13 and 16, 2017 (-2096).

    I am maintaining a spreadsheet (with LibreOffice Calc), where I manually jot down the date and the number of "Likes" at that moment, and compute the resulting "Likes/day" and "annual growth rate in %". This occurs at somewhat irregular intervals:
    • Between April 2015 and January 2016 I recorded the Likes almost daily (I skipped the occasional day when e.g. internet or computer was down, or a week when I was on vacation). Average time interval was 1.06 day, maximum interval 8 days.
    • Between February 2016 and Mid-March 2018 I slagged - I had intervals between datapoints averaging 3.5 days, with a maximum of 15 days.
    • Since Mid-March 2018, I am back to an "almost daily" routine, with average intervals of 1.05 days, and max 5 days.
    Note that I do not put down the number at the same time on the clock every day! I usually do it late in the evening here in Germany, say 10 to 11 p.m., but sometimes it's after midnight, sometimes late in the afternoon.

    I keep a separate table for each month, to which I copy the records for each 1st of the month - or the nearest date, if I skipped the 1st. Again, I compute "Likes/day" and "Annual %".

    Here is a simple graph showing how the number of "Likes" has developed, month by month - total number:

    FB_Likes_20190828.

    My monthly tallying began actually on January 1st 2015. I include the 1-day dip of March 13, 2015.
    You see already that the curve was rather steep, with varying gradients, between August 2015 and October 2016 - and since then, it's been going stepwise: Several months of shallow gradients and short 2-months steps of high gradient.

    This graphs shows the resulting "Likes/day" on a monthly average basis:

    FB_LikesPerDay_20190828.

    Note that the curve before 2015 is of little value, as the time intervals vary greatly. The years from late 2014 to September 2016 show great ups and downs - which I have never been able to correlate with anything, but since October 2016, patterns emerge among the randomness. A closer look at the fine data reveals what I mean. The next graph shows the actual "Likes/day" as finely granulated as I have them, i.e. almost daily for much of the scope:

    FB AE LikesPerDay 20160101-20190828 Annotated.

    I want to point out the following features:
    1. The 9/11 anniversaries: There are very high, very short peaks at the same dates every year: Between about September 10 and September 14. I'll show you detailed graphs for each September below in thumbnails. This one's obvious: Both the general interest and AE911Truth's own marketing efforts peek around 9/11 each year.
    2. There is also always a sharp peak at or around the 1st day of the year. I have no explanation for that.
    3. Most of the time, the curve meanders randomly about some low baseline, which occasionally even dips slightly below zero for periods of time. These periods are mostly confined to a range between -50 and +100 Likes/day.
    4. And then, there are two periods, each 65 days long, where suddenly, from one day to the next, the low random walk explodes to a whole new level. Each time, the first full day of such a period sees the local maximum: +1,899 on 12/18/2017, and 1864 on 10/02/2018. After that, these periods fluctuated at levels between 250 and 600 (Dec 2017/Jan18) and between 600 and 900 (Oct/Nov 2018). Then, as suddenly as the frenzy started, it drops from top to zero from one day to the next.
    These thumbnails show the Septembers of 2015, 2016, 2017 and 2018:
    FB_LikesPerDay_September2015. FB_LikesPerDay_September2016. FB_LikesPerDay_September2017. FB_LikesPerDay_September2018.
    In these years, the 4 days from the 11th to the 14th of September had the following total new Likes:
    2015: 3,194
    2016: 2,367 (only 11th to 13th)
    2017: 1,109
    2018: 1,537

    Same for the New Years of 2016 through 2019:
    FB_LikesPerDay_NewYear2016. FB_LikesPerDay_NewYear2017. FB_LikesPerDay_NewYear2018. FB_LikesPerDay_NewYear2019.
    These graphs are not as obvious as the Septembers. Some explanations:
    • 2016: The y-scale goes to 3000, to capture an extraordinary peak on January 15th. Its 2256 Likes were the record high for a single day I recorded so far, and it was followed by 1,639, 966, 835, 719, 701, 474 and fading into the around-300 range. I'll have to go back later and figure out what that was. Anyway, the peak for the beginning of January at 560 comes on Jan 3rd, and it looks small to that scale, but is a local maximum.
    • 2017: Unfortunately, I only have a data point for the first 7 days of January 2017: That was +1,774 for 7 days. This could well hide a sharp, genuine peak. It's still a local maximum, the last time that this value of 253/day had been exceeded before was 70 days prior, and the next time this value would get exceeded was the next September 11th.
    • 2018: This is in the middle of a 2-months period of very high traffic, but still comfortably a local maximum, exceeded only by the first day of that high-traffic period.
    • 2019: A clear, lonesome peak, distributed over 2 days. It's always possible that an event that occurs entirely within a day in, say, California, will be registered by me as being spread over 2 days - time zones!
    I can roughly estimate how many "Likes" those January 1st peaks added to the general trend at the time, by subtracting the local baseline:
    2016: +500
    2017: +1,000
    2018: +800
    2019: +850


    Now for the most interesting feature of them all the two 65-days periods of crazy increases.
    First, I'll thumbnail these two periods so you can appreciate the suddenness of their beginnings and ends:
    FB_LikesPerDay_Frenzy201712-201802. FB_LikesPerDay_Frenzy201810-201812.
    In both cases, the fact that there is a first "intermediate" day before the maximum peek aroun 1,900 can be attributed to the fact that, whatever happened there, began during a day, and that first day may have captured only some hours of the sudden activity burst. Same goes for a single step down at the end.

    Rather obviously, some activity was "switched on", and 65 days later "switched off".

    Here is the effect this had in a more broad picture - I segmented the era since October 21st, 2016 (when an period of low activity began) into 5 segments, bounded by these 6 dates / values:
    10/21/2016 : 407,995
    12/15/2017 : 420,402 - 420 days of low activity, +12407
    02/18/2018 : 448,033 - 65 days of high activity, +27,631
    09/30/2018 : 454,056 - 224 days of low activity, +6,023
    12/04/2018 : 502,667 - 65 days of high activity, +48,611
    08/28/2019 : 507,510 - 267 days of low activity, ongoing, +4,843

    FB_LikesPerDay_Phases201610-201908.
    In that era, which was 1,041 days long, they added 99,515 Likes - that's 96/day.
    Of these, 49% came in the second hot phase, 28% in the first hot phase, and the remaining 23% in the three cold phases. So 12% of the days got 77% of the "Likes".
    The hot phases had annual growth rates of 43% and 77%, respectively, while the cold phases had rates of 2.6%, 2.2% and 1.3% annually.

    We could subtract from the "cold" phases the peaks of Jan 1st and Sept 09 to arrive at a "base" performance:
    1st low phase: subtract 1,100 and 1,000 = ca. 10,300 in 420 days = 24.5/day
    2nd low phase: subtract 800 = ca. 5,023 in 224 days = 22.4/day
    3rd low phase: subtract 850 = ca. 4,000 in 267 days = 15.0/day

    Now actually, for the last half year, since mid-February, the number of Likes has stagnated: +/- 0! There are shorter periods where the number steadily increases at rates between 1% and 4% annually, alternating with periods where the number steadily decreases at rates around 2% annually. It looks like this:

    FB_Likes_20190101-20190828.

    It is peculiar that lately, the direction (up or down) has been maintained for periods of 25 days and more before turning to go the other way with equal steadiness. This doesn't appear to be random, but I have no explanation at this time.



    tl;dr

    Ok ok, that was lots of borring statistics. here is the quick takeaway:

    1. The AE911Truth Facebook page has existed since February 2009. It has received more than 500,000 "Likes" since then
    2. I have monitored this development from day to day since April 2015
    3. In 2015 and until October 2016, the rates at which "Likes" came in were going up and down chaotically, but generally fast, mostly at rates between 100 and 500 per day.
    4. Since November 2016, Likes have, for most (88%) of the time trickeled in at under 25/day on average, with periods of sustained daily losses interspersed.
    5. There are notable peaks, adding on the order of 1,000 Likes, each year around the first day of the year, and around the 9/11 anniversary. While these peaks are very visible, they are only borderline significant, often contributing less than 5% of a year's "Likes"
    6. Occasionally, however, there comes a period, very suddenly, and goes as suddenly 65 later, when massive numbers of Likes flow in - often more than 500/day - these account for more than 3/4th of all new "Likes" in the last 3 years.

    Personal Conclusions
    • It is my conviction, though I have no independed evidence, that these 65-day bursts are caused by direct action, most likely on the part of AE911Truth, or affiliated parties. Such action might be the purchase of major ad campaigns from Facebook (you pay money so Facebook shows your page to millions upon millions of people that Facebook thinks might like your kind of page). It could also be that AE911Truth (or someone acting on their behalf) purchases the services of bot farms (borderline legal - usually a breach of Facebook standards) that induce live accounts, or fake accounts, to "Like" AE911Truth. These campaigns are time limited, and may have certain goals in mind. It is noteworthy, for example, that the era of "low" activity started in October 2016 shortly (4 weeks) after they had surpassed 400,000, and the most recent "hot" phase ended just 3 days after reaching 500,000. It alone had added close to 49,000 Likes in a little over 2 months.
    • I do not accuse AE911Truth of wrong-doing. It is certainly legitimate and normal to invest in advertising to increase visibility and followership on social media.
    • I am however arguing that AE911Truth sees very little organic growth - members telling other members, etc.
    • Lately, it seems the baseline even seems to have been turning to negative growth.
     
    • Informative Informative x 7
  2. Crusher

    Crusher New Member

    By your own admission you have nothing and so everything you've said can be dismissed.

    Yet you imply wrong-doing above "It could also be that AE911Truth (or someone acting on their behalf) purchases the services of bot farms (borderline legal - usually a breach of Facebook standards) that induce live accounts, or fake accounts, to "Like" AE911Truth." Even though you have no evidence for it.

    You haven't even provided any evidence for this. For example, compared to the hundreds of thousands of facebook pages, how does this stack up? are there similar trends in all facebook groups? [...]

    [paraphrasing removed]
     
    Last edited by a moderator: Sep 2, 2019
  3. Mick West

    Mick West Administrator Staff Member

    I have a small FB page for Metabunk, just under 6K likes. I only post around once a week, so there's not a lot going on. Facebook only gives you two years of stats.
    Metabunk 2019-09-02 08-56-35.
    Even when I've not posted for quite a while, I don't see the likes decrease.

    I think the AE911 data is pretty interesting, even without direct comparables. I've heard from several people that the movement is running out of steam. Anecdotes, to be sure, but they were some reasonably high profile people who just decided to get on with their lives (while still having lots of suspicions about 9/11)

    They often point to 2006 as the heyday of 9/11 Truth, which matched the data from Google Trends
    Metabunk 2019-09-02 09-11-37.

    Currently (Sept 2, 9AM PDT) they are at 507,487 likes, 508,420 follows. Tomorrow is the start of their latest great media campaign with the release of the Hulsey report. It will be interesting to see what effect it has.
     
  4. Crusher

    Crusher New Member

    Still not enough of a sample, and still a waste of time without it.

    [off-topic material removed]
     
  5. Mick West

    Mick West Administrator Staff Member

    Absent the rigorous data you desire, what do you personally feel about the state of 9/11 Truth? Have you noticed more or less interest in recent years?
     
  6. Oystein

    Oystein Active Member

    No. This is FALSE, and the opposite is true: I do present a lot of data that is consistent with the claim I make, and my claim rests on that data. That is not "nothing". I do however readily admit that I have "nothing" to independently confirm it.

    I cordially request that you refrain from misrepresenting what I write in the future.

    No. This is FALSE. I do NOT "imply" wrong-doing, I merely list it as one of the logical and plausible possibilities, alongside a perfectly okay possibility, which is equally plausible (running paid ads with Facebook, or elsewhere). I have no evidence to decide either way, so I let both possibilities stand side by side, and explicitly refrain from accusing them of the borderline legal deed.

    I cordially request that you refrain from misrepresenting what I write in the future.

    This is FALSE, again.
    I have presented a lot of evidence. This is perfectly obvious.
    You may validly argue that this evidence is insufficient, and suggest additional evidence to be presented to close the gaps. Very well. Just please refrain from misrepresenting what I posted, and what I didn't.

    I claimed that there is little organic growth most of the time. My data is sufficient to show this: For 3 years now, "Likes" have, for most (88%) of the time, been increasing by 2.6% and less annually, with a downward trend, and recently even more than half a year during which nothing was gained.
    This could only be interpreted in a positive light, i.e. other than "little organic growth", if it could be shown that Facebook membership had been declining in recent years, particularly in the USA (I am taking total number of users in the USA as the first obvious proxy of overall growth on Facebook).

    I have plenty of other data about support levels of 9/11 Truth groups, and yes, there is a general trend of stagnation in numbers, decline in interest and activity.

    However, my main reason to post this article at this time is that I felt I now have enough data to make a case that most of AE911Truth's Facebook "Likes" come in during short and sharply delineated periods, transitioning from almost no growth to rapid growth, and back, in just a day or two. This is indicative of them (or someone) switching on and off major campaigns.
    But I have not seen any campaigns either on their own Facebook presence, or in terms of media presence, touring campaigns, newsletter blitzes, or whatever you could have to generate organic growth. The only mechanisms I can think of to switch on/off such events is paid campaigns on Facebook - either legally or unethically.

    Or do you know of any other possibilities?
     
  7. Oystein

    Oystein Active Member

    Here is another graph for AE911Truth:
    They still have the old homepage design, which has, at the bottom, a counter for the "Content View Hits", which currently stands at "20351037". This counter has been counting, quite amazingly, since at least June 06, 2017, when it stood at 41255. It was out of operation for about 5 months in 2010, and from time to time, the number drops by a few thousands. I keep a tally on that count, too, and have plotted the number of daily "Content View Hits", averaged per month:
    20190901_AE911T_PageHitsMonthly.

    You see there was continuing growth in traffic from the beginnings until the 10-year anniversary. There have been clear local peaks every September, marking the anniversaries.

    From spring of 2015, when they had launched a new design, there was again an upward trend - until more than a year ago, and now for more than a year, traffic has oscillated near the lowest level they had in the last almost 9 years.
    The anniversary peaks have been unobvious in the last 2 years.

    It seems to be holding up pretty well compared to the decline in interest in 9/11 as a search term than Mick documented, but certainly it is not increasing.
    Now this data is difficult to evaluate - during this time, there has of course been a trend away from classical homepages and toward social media presences. I cannot tell how two the launches of new web designs affect the counter. So I'll be careful not to read any doom into this data set. But we can confidently state that the data reveals no boom for AE911Truth.

    I am by the way putting out these data now, with a few days to go before the usual 9/11 anniversary hoopla, to wedge in a prediction:

    The annual 9/11 peaks of interest, both on the homepage counter and the Facebook likes, will be less pronounced than in previous years, and perhaps even barely noticeable. (Unless they do launch another campaign at Facebook to boost Hulsey - but then we won't see a "9/11" peak but rather a "block" of high Likes/day, starting with >1000 in 1 day.)
     
  8. deirdre

    deirdre Moderator Staff Member

    which is implying.

    are there big 'share' posts around that time? if another big FB group shares a particular post with their large group, you might get a lot of new likes.
     
    • Agree Agree x 1
  9. Oystein

    Oystein Active Member

    Sorry, no.
    If I observed that you ate at Joe's Village Grill, and the menu lists Hot Dog and Burger as the only two items, and I'd list as the possibilities that you ate Hot Dog OR Burger, I'd not be implying that you ate Burger.

    a) No, I found no noteworthy activity of "sharing" coinciding with those high "new Likes" times.
    b) When some event of the sort you describe occurs, a single posting that generates high traffic and interest, what you see is a fast build-up to a peak, and then a decaying slope. That's not what I observed and documented above, where you have a solid block of a little over 2 months, and then a SUDDEN drop back to base level.
    Amidst the data, you see plenty such short-lived peaks as I just described. More often than not, I have not seen a particular posting on their own site that I can clearly associate with such peaks, so we may guess that some promotional article elsewhere may have generated these.

    Again, we see sudden onset, and then sudden death, of those phase - towers of "Likes" standing solitary like mesas in the Utah desert.
    Those are the result of campaigns.
    I just can't tell yet the nature of these campaigns.
     
  10. NoParty

    NoParty Senior Member

    Is your argument that any information that is less than entirely conclusive (yet?) is "a waste of time" ?
     
  11. deirdre

    deirdre Moderator Staff Member

    To me the analogy is: You saw me eat at Joe's. You know I have no money and you are contemplating how I got the burger.
    One of the plausible and logical possibilities you list is that I stole the burger. That is implying I might have stole the burger.

    There's nothing wrong with implying unethical behavior from AE911. Just saying you don't have to jump all over the other guy for pointing it out.
     
  12. Oystein

    Oystein Active Member

    No. Just no. Simply, objectively and unequivocally FALSE. Sorry.

    The "I have no money" thing has no analogy in my AE analysis - they DO have money.

    I am - read and behold my words, and don't ever misconstrue them again, please - NOT implying that they went the unethical route of buying fake likes. I am implying that they bought ads ("ethical") OR fake likes ("unethical").

    This is a matter of strict logic: Implying "A OR B" never ever is equivalent to implying "B"!
     
    • Disagree Disagree x 2
    • Agree Agree x 1
  13. Crusher

    Crusher New Member

    [comment: lying about what another moderator removed up thread: removed. ]

    There is nothing to back up any claims other than a claim similar to "The ae911 facebook like count is growing, sometimes loses a few thousand". There is no argument there. The suggestion is to get a larger sample size then present data to see if we can find patterns. Not doing so is a waste of time.





    I respect if your claim that the ae911 facebook like count is growing, sometimes loses a few thousand. It's not about your data, it's but your use of it to imply things that the data doesn't imply.


    Here's the example. You still haven't demonstrated what 'little organic growth' is. So it looks like you'll first need a subset of facebook groups, analyze the data, get a baseline, compare the results and present your findings. With all the data here it doesn't demonstrate how it's 'organic growth' at all. That's not to say you're wrong, just that you haven't demonstrated it with a larger sample set.



    So why did you hold it back from the findings?



    Ok, nice hypothesis, but now you need to prove it with more data.


    Honestly, i think you need to gather a lot more data before you start coming to conclusions. this is like an assumption on an unproven assumption. I wouldn't assume to know of any other possibilities, but i could think of at least one, such as people seeing their information and liking it. So at this point i would bet there's quite a few possibilities you probably haven't thought of yet.
     
    Last edited by a moderator: Sep 3, 2019
    • Funny Funny x 1
  14. Oystein

    Oystein Active Member

    Any specific suggestions what data I should gather?

    No. An assumption (in logic) is something conjured up from nothing. My conclusions are post-data.

    People see and like AE911Truth's information only 12% of the time? And then suddenly, from one day to the next, stop seeing and liking it? Implausible.
    What you describe explains many of the peaks - particularly the 9/11 anniversary peaks, but very likely many of the other. And of course the low-rate growth is best explained as a people (a few world wide each day) seeing and linking what they see. This is trivial.
    The question is: What causes this "seeing and linking" to jump up so dramatically, remain high for a while, and then drop down to almost zero as dramatically? Add-campaigns would potentially explain that many people see the page who have not seen it before, such that they might like it. "Buying likes" campaigns however would potentially explain why accounts (whether it's people of bots) "like" (hit the "Like" button on) the page - but they would not actually "see" it - i.e. read, understand, process, and actually like it.

    I am open for suggestions! That's part of the reason I published this.
     
  15. Crusher

    Crusher New Member

    The suggestion has been provided numerous times:


    tl;dr (and in case you missed it): get more data
     
  16. Oystein

    Oystein Active Member

    Your suggestions do not respond the quote. The quote, "I am open for suggestions!" was originally uttered by me in the context of this:

    You wrote:
    This does not answer the question I asked in context. I had asked:
    I am bolding and red-fonting the key word.

    You make very general and vague suggestions, like:
    • Comparing to "hundreds of thousands of Facebook pages" - really? If I compare to only one-hundred-thousand Facebook pages, that's insufficient then? How should I choose which "hundreds of thousands of Facebook pages" to tackle?
    • "not enough of a sample" - can you specify which sample size you'd accept? Or can you provide an objective argument why my sample size is not "enough"?
    • "more data" - which data?
    So you see, your objections and suggestions are not helpful because they lack specificity.
     
  17. Oystein

    Oystein Active Member

    The Hulsey Draft was announced on the AE911Truth Facebook page close to 20 hours ago:
    https://www.facebook.com/ae911truth...6Rf0QUOqT8yrbJ30xDAh7fUnhBtGlJcpj-w&__tn__=-R

    It has, as of this writing
    • been shared 935 times
    • been commented 276 times
    • received just over 1,000 reactions (Likes, Yawns, Laughters...)
    Within the last 24 hours, the page still lost 11 "Likes", a little more than in the 3 days before, when there was no similarly popular post.
    The most recent post that has, as of today, accumulated higher activity numbers (1.7K Shares, 344 Comments, 1.7K Reactions) was on August 20th, 3:45 a.m. (Central European). On August 21st, I observed a loss of 13 Likes, which was about average for the week that sourrounded it.

    So this shows that popular posts with those kinds of numbers of engagements do NOT alter perceptibly the development of net "Likes" gains.
     
  18. Mick West

    Mick West Administrator Staff Member

    I think that indicated a large number of stagnant followers. AE911 was simply not showing up in their feed because they did not have many popular posts. So where a viral post comes along more people see it, then since they have long-ago moved away from that kind of thing, they are prompted to unsubscribe.
     
    • Like Like x 1
  19. Oystein

    Oystein Active Member

    But -11 is not a remarkable loss. July averaged -10/day, August -20/day, and the last 2 weeks before today averaged -14. There is no effect - neither positive nor negative.