How big is this anaconda?

March 13, 2019 • 5:00 pm

by Greg Mayer

The anaconda in the video posted this morning is real, but it is certainly not 15 m long. Alert readers went digging, and found clues (posted in the comments) as to what the story is. I’ll post on that later, but for now, here’s another anaconda, this one a specimen I saw on display at the San Diego Natural History Museum. How big do you think it is?

Anaconda on display at the San Diego Natural History Museum, 16 January 2019.

This illustrates at least two phenomena, both of which have been problems in determining how big giant snakes can get: the difficulty of estimating size, and the effects of skin stretching. I’ll post my measurement of this snake along with the reveal on the “15 m” one later.

33 thoughts on “How big is this anaconda?

    1. Of course! Good point… Now, are measurements done along the line of the body, as I would suppose, or tip of nose to tip of tail? Well good luck with that if it IS a writhing 15m snake!

    1. There have been a few cases in which a freshly dead snake was measured, and then the skin was measured. The skins are generally said to be 20% longer than the snake. I can recall one such case off-hand (because I read about it while preparing the first anaconda post): Arthur Loveridge of the MCZ measured a 2-3 m rock python, and found the skin to be 21% longer.

      By looking closely at a skin, the source of the stretching is evident. The body scales of most snakes slightly overlap one another, with soft, unscaled skin in between. When skinned, the scales are pulled away from one another, and you can see the soft skin between the scales, which add to the length. If you enlarge the photo by clicking on it, you can see the between-scale skin, especially laterally on the body. This stretching, and the amount of skin between scales, may not be uniform along the whole body, so it would be difficult to try to reconstruct the length of just the scaled part of the skin without. So, I think Loveridges’ method is best for estimating the effect, but of course knowing what usually happens doesn’t tell you the effect size for your snake.


  1. from the pattern on the skin it does not appear stretched. Moreover, it looks like a full skin, not a shed mould. It must be east to measure ot.

  2. The skin appears to be about 16-18 feet long to me. I imagine before it’s death, it was perhaps 15 feet.

  3. I remember seeing a 25-foot python skin on the wall of the “lodge” we stayed at in Canaima, near Angel Falls in Venezuela in maybe 1971. The live version had apparently been found swimming near the children, and had to be dispatched. I have no idea how stretched it was, but it was hugely wide as well as long. Not something I’d want to run into on land or sea.

  4. I don’t know how big that anaconda is, I only know it “don’t want none unless you got buns hon.”

  5. Using Greg’s USN “NAVY” cap as a measure that skin is 20 ft long, but Greg works in metric so let’s call that 6 metres – this squares with the wheeled footstool dimensions I’m familiar with too.

    From Wiki I learn this:

    “A shed skin is much longer than the snake that shed it, as the skin covers the top and bottom of each scale. If the skin is shed intact, each scale is unwrapped on the top and bottom side of the scale which almost doubles the length of the shed skin”

    So I’m going with 3.5 metres for the snake [that’s 11’6″]

  6. Here in Far North Queensland (coastal) – we have to deal daily with amythestine pythons – which have a particular love for flying foxes. The largest recorded, that I’m aware of, was 10 meters long. We regularly get 3 – 4 meter snakes on our baby cage. Prior to European settlement here – Aboriginal families lost children to these snakes. Luckily they are not particularly aggressive – although I’ve received bites from them.

  7. I haven’t bothered to measure my pet ball pythons in years! But maybe I should measure them again.

    So how to measure a living ball python? Well, I could take a tape measure and hold it next to Julius Squeezer’s head and try to measure her down her spine. (Yes, “her”. Julius is a girl.) That’s hard. She’s fidgeting around.

    The other way is to put a yard stick, um, a folding one that measures six feet rather than three since my little snakes were somewhere between four and five feet, if my memory serves me, and stick it along the wall. It’s “Julius! Sweetheart! Let’s see if you can stretch out your full length! C’mon, HoneySnake, lemme see how long you are…”

    I love cats, which is pretty normal. I also love snakes, which is pretty weird.

    1. Baby buy an Anaconda
      baby buy a thing to wear
      She go walking she go crawling
      she go sliding through the air…

      Sisters of Mercy song Anaconda!

  8. Hey Greg it looks like they simply switched the aspect ratio of the ordinal video (0.550 aspect ratio, or about 1:2, taller than wide) to a 16:9 aspect ratio (wider than tall). Spotted by infiniteimprobabilit in the other thread.

    1. Actually, the switching aspect ratios theory is still kind of a guess since we don’t know what precise video the stretcher stretched (there are more than one floating around), and my results from the videos I used aren’t precisely lining up, and the results can be achieved by just stretching a video instead of switching aspect ratios. Like infiniteimprobabilit says, “The net effect is the same, though.” Anyways you get the idea and at this point I’m probably babbling lol.

      1. OK – lets get pedantic. 🙂

        There are black bars top & bottom in the stretched video. I’ve coloured the black areas in the ‘before’ & ‘after’ videos below a nice red for clarity:

        By pausing both videos at the same timings I can see there are no crops – the full image is displayed in both. And both are sitting inside a 1280 x 720, 16:9 YouTube video frame.

        The original image area is 9:16 [portrait] or 405 x 720 & it has no bars top & bottom = 0.56 ratio

        The ‘stretched’ video is fitting imperfectly & there’s black bars top & bottom making the image 1280 x 704 = 1.82 ratio

        1.82/0.56 = a horizontal stretch of x 3.2 of the image area

        1. This is starting to hurt my brain now but I will note that the black bars were probably in the uploaded video and not added by youtube, since nowadays youtube resizes the video player to fit the videos. (So the uploaded video was at 16:9.) I’m now leaning toward stretched manually since I’m not sure if there is an easy way to tell the video editing program “Hey change the aspect ratio” and then have it stretch like that, and have the black bars.

          1. [1] The source video is THIS 41 SECOND VID

            It’s 41 second, portrait, 9:16, 405 x 720, 30fps video [405 x 720 being the standard smartphone portrait resolution]
            At 28 SECONDS a text overlay begins saying “DUMATO Nature is alive & free”

            [2] The ripper downloaded THAT video & trimmed off the last 16 seconds so as to not have the text overlay leaving him with:

            A 25 second, portrait, 9:16, 405 x 720, 30fps video

            [3] The ripper didn’t know his video formats so he told his software to upscale the video to 1296 x 720 [18:10] & he set the the speed to 25fps instead of 30fps [this happens a lot!]

            [4] When he uploaded [3] to YouTube it ‘downscaled’ 1296 x 720 to 1280 x 704 so it would fit in the YT 1280 x 720 frame – this puts an 8 pixel high bar to the top & an 8 pixel bar to the bottom. Which is THIS 30 SECOND VID

            1. Point [3]: I mean he told his software to stretch the video by exactly 3.2 from 405 pixels

              3.2 x 405 = 1296 pixels exactly
              I had a cheapo nasty, but free video editor from the pre-16:9 days with no 16:9 option, but it allowed me to scale x & y separately by ratios. I think the ripper guessed wrongly that 3.2 would give him 16:9

            2. This is what I get when I upload a 405 x 720: watch?v=3AnVBkiAHUw

              This is what I get when I upload a 1296 x 720: watch?v=DFC6sMEUVXA

              Youtube resized their player to fit that video.

              If the ripper would have downloaded the original DUMATO video they would have gotten a 640×360 or a 854×480. (That’s what I got when I downloaded it.)

              The ripper didn’t know his video formats so he told his software to upscale the video to 1296 x 720 [18:10]

              Do you have steps to reproduce that in Vegas because I’m having trouble doing that and having it come out stretched and with black bars! 😀

              1. My name isn’t “Vegas”

                The image with “1296 x 720” in white characters on a red ground. On the upload to video it is 1280 x 712 with 4 pixels of black top & bottom I think. But I will try myself to see.

              2. The Dumato video displays as a 1280 px by 720 px on my screen & it can be DL’d at the screen resolution using VLC & no doubt other progs.

              3. That’s what I get when I re-download it. 1280 x 712. Definitely uploaded at 1296 x 720 though. The other 405 x 720 one when I re-download it I get a 406×720.

  9. As it is a picture about size in animals & birds, consider that the approximate average size of a mammal is the same as a dog or cat…

    I almost wondered if the penguin had been eaten by the anaconda! 😉

Leave a Reply