Google+ and Facebook react to the ant throwdown

A few days ago, as an experiment, I uploaded the following image to Google + and Facebook:

A battle between Rhytidoponera victoriae and Amblyopone ferruginea (Victoria, Australia)

I was interested to compare differences in community engagement across social media platforms. Why? I post my new photos to Google+ and don’t do much on Facebook (which I hate with a passion, for a variety of reasons), but if Google+ continues losing steam I might have to suck it up and go back. Hence, a test image.

The post went up on the evening of March 10th. Here’s what happened: pandamonium over on Google:

The Google community totally screwed the comparison by engaging too much with the image. The post got listed under “what’s hot” and next thing I know a bunch of random non-followers were plussing and commenting all over the place. That’s not what usually happens. As facebook doesn’t have an equivalent to a global “what’s hot” feed, this is not a fair test. In essence, the data point succumbed to random internet weather patterns.

I promised folks I’d report the results, though, and promises being what they are here’s a graph:

Google Wins

20 thoughts on “Google+ and Facebook react to the ant throwdown”

  1. @Jason g. Goldman I think I’d count this as a single trial within an experiment. So it’s time for more trials and some stats 😛

  2. Did you notice a big difference in the nature of the comments? (like, was it more positively received in one forum than another? More general “cool!” comments somewhere vs. more ento/photo/technical comments somewhere else? I guess I’m trying to figure out a way to gauge the types of viewers and their interests…)

  3. If you wish to test some hypothesis numerically, I will agree with several others that you need to run more test examples. Random selections from the photo pool would be best and you simply run enough until N yields the desired variance, eg p = .95 according to formulas like those here:


    Otherwise, who cares ? Anecdotal conclusions are just like weather, always interesting and changing.

    I vote for an animated gif of that battle next time with a WWF soundtrack thang – you KNOW you could do it. Alex !

    AntWrestleMania Saturday Night Cage Matches !! Think of the sitetraffic —

    Oh the Humanity !!!

  4. Alex

    Sorry to hear that you are so negative about FB. Quite often I put your articles on my wall. I hope that this is not frustrating to you. If so, of course, I will not put it there in the future.

  5. I think that before you try any kind of actual experiment, you need to develop some hypotheses and criteria to evaluate them. A single random post isn’t very scientific.

    1. David – how could you ?

      Why should a single observation not be very scientific ? We have an entire … erm … science called “Climate Science” based on single observations, so they can’t be ALL bad, can they !!

        1. WOW, Alex

          You really don’t know how weather station data has been collected for the last ~200 years ?

          In the beginning, there was ONE liquid in glass thermometer at a station generally read ONE time a day …. mostly, they missed some … reading accurate to the nearest 1 degree with equipment having precision and limit of observability of .5 degree C. This was how temperature was recorded until more modern methods were introduced in the late 20th century (1970’s).

          That is n = 1 …. a sample size of one, and non-random.

          Here are the instructions for US GHCN stations (see page 11)

          Read more at about how about 2/3 of the GHCN stations were abandoned because those stations’ inaccuracy greater than 2-5 degrees C

        2. Riiiight, Bob. All weather data is taken from just one measurement.

          What do you call it when you compare measurements from tens of thousands of stations?

          Multiple measurements.

          I’m not going to debate climate science further here. This is not a climate science blog, and I’m not going to provide a platform for crazy.

          If you want to spread that kind of ideological off-topic crap here, I’ll delete it.

          [update: on second thought, at some point I should open up a separate climate change post/thread]. I just don’t want that crap polluting my bug threads unless it’s actually relevant]

        3. I will cover this briefly, but you can discuss it with a statistician sometime or do some reading to flesh it out. Most important of all is to understand the weather is chaotic and variable and the population of temperatures changes in time and space.

          First, was yesterday’s population of temperatures in one city (call it max-min for brevity) exactly the same as today’s ? No
          Was yesterday’s max-min exactly the same in Urbana as Bloomington’s ? No
          So, we are not measuring the population of all places on all days when we measure the temperature of one place. All places and time-ranges potentially have different max-min populations. I will let others argue about the proper dimension of our time/space ‘container’ but obviously there is a problem here.

          When we attempt to measure the daily temperature of one place we also know there are local variations (sun/shade/grass/asphalt/wet/dry/etc) but we do not know exactly the population variance. That is why it is good science to take random, replicated samples which let us estimate the variance of that particular population. We can determine based on estimated data that 25 – 70 random samples will provide an 95% probability that the true min-max values have been obtained.

          However, we usually have ONE sample, not 25.

          Each station is a different max-min population, which is the point of HAVING more than one.

          Sorry if you think basic sampling is ideology, Alex, but it’s just the basis of the scientific method, imo.

        4. By all means delete them if you like, Alex. They were a response to your “call bullshit” comment anyway. Hopefully you understood the point that when you want the average daily temp of Urbana, you take n > 1 measurements each day in Urbana, not one time in each of 10,000 other places.

  6. That is a true action shot if I ever saw one! Looks like they are wrestling back and forth, amazing capture.
    The damaged antennae on the Rhytidoponera is a nice extra detail, I hope no clone tool has been invlolved? 😉


Leave a Reply