Update on the video of an Australian Terrier used as an example of ETS in 2010

The video below was uploaded to YouTube on Apr 21, 2010 by Clean Run Magazine as a companion for the article in the May 2010 issue of Clean Run: What Is Early Takeoff Syndrome?”

When I first watched that video back in 2010 it was very upsetting for me to see an Australian Terrier (or any dog for that matter) trying so hard to succeed yet failing so miserably. Based on the description of the video on YouTube,  I think it is safe to say this video is intended to be an example of a dog who has ETS.  Back in 2010, I didn’t question that assumption and felt terrible for both the dog and handler. However, after recently watching the video again, I was surprised to see many other potential explanations for this dog’s jumping failures based on my increased understanding of the challenges dogs face with agility-style jumping.

1)  Design of the Jump Grid.

A) The dog had nothing to drive towards after the last jump: No toy, No manner’s minder, No hoop, No target with food.  In addition, the handler decelerated while her dog was jumping the final jump, so the dog was correct to think he should land short and turn towards the handler vs. jump in extension and continue to drive forward.

B) The last jump being a spread jump created a double challenge.  Many “normal” dogs take-off early for the final jump on a course, a sequence, or in a jump grid that does not have another obstacle (or reward point) beyond it.  One possible reason is the lack of object overlay or *parallax which help dogs determine the location of objects in space.  See link at bottom for more information about parallax and canine vision.

C) The jumps were not evenly spaced and the spacing appeared to be awkward for a dog this particular size and structure.   I will elaborate further on this topic in a bit.

2)  Procedure.

The dog was being asked to repeat a jump grid in which everything remained the same EXCEPT for the double and triple jumps bars, which were moved from front to back between each repetition.  Most dogs quickly develop striding patterns when being asked to perform a jump grid multiple times in a row.  That is actually one reason people think jump grids are useful because repeated striding patterns help dogs build muscle memory.  However, watching this video you can see that the overall jumping “picture” remained the same with 4 jumps positioned in the same locations for all repetitions.  The only things that changed were the locations of the bars in the double and triple jumps (and the height of the bars for reps 3 and 4).

Below are  3D renderings of the jump grid featured in the video as seen from the dog’s perspective and bird’s-eye views.  The spacing in the digital rendering is 10′ between jump #1 to the front edge of the double (jump #2), 9′ between jump #2 and jump #3, and 8′ between jump #3 to the front edge of the triple (jump #4).  The dog’s view is based on the dog’s set point in the first rep.   This spacing is based on the number and length of the handler’s steps as she ran along side the grid as well as the dog’s number and length of strides between jumps.

In the bird’s-eye view, jump #1 is to the right and jump #4 is to the left.  NOTE: jump #1 stanchions should have been white vs. red (Oopsie) but it is not relevant to the topic at hand.

ETS_birdseye_view_bars_front copyETS_dogs_view_bars_front copyETS_dogs_view_bars_back copyETS_dirdeye_view_bars_backComparing the dog’s view of the jumping “picture” with bars in the front vs. bars in the back (as seen above), you can see how similar they look so I don’t think it is reasonable to expect the dog to notice such subtle differences in the “geometry” of this jumping “picture” as he approached the grid on the 2nd rep.

I think it is also important to notice how uneven the jump grid became when the bars were moved to the back edges of the spread jumps as seen in the bird’s-eye view image above.

I think it is far more likely that he thought the jump grid was the same (as grids like this usually are) and thus patterned the same striding as in the 1st rep.  Below are screen saves that show how his take-off spot was nearly identical in reps #1 and #2.

Double_jump_take-off_front_bak_comparisontriple_take_off_front_back_comparisonOn rep #2, as seen in the screen saves above (Bars In Back), you can see that the dog took off from the same spot for the spread jumps as he did in rep #1 (Bars in Front).  On the video you can also see that he made a Herculean effort to clear those bars (which likely caught him by surprise by not being in the same place as they were in rep #1).  His grand effort to clear the double on rep #2 suggests to me that he could see where the bars were  but was “tricked” (by rep #1) into thinking the jump grid was going to be the same.. as jump grids tend to be.

The following is from a post that explains why dogs and people sometimes project what they think is there vs. see what is really there.


paris_springDid you say “I love Paris in the Springtime”?  If you did, you are not alone.  That is what most people say when reading that text out loud but if you look carefully, you will see something surprising.  It’s a perfect example of how we see what we “think” is there vs. what is really there.  Here is a link the full post: https://artanddogblog.wordpress.com/2013/06/13/some-of-the-thinking-behind-hgr-games/

Granted the dog in the video takes-off earlier than what most people would consider ideal (including myself), but in the first rep, he looked confident and his jumping looked easy for him to do.  If you watch his approach towards jump #3, you can see how he shortened his last stride in order to find  his “preferred take-off spot.”  Another thing worth noting is that he landed approximately the same distance from all 4 jumps which further leads me to believe he is able to see where jumps and bars are in space.  I do not think this dog has a “vision defect/ ETS.” I am more inclined to think he has developed a habit of taking-off from a particular spot and landing in a particular spot due to a combination of structure and lack of effective jump training.

Below are screen saves which show how similar his take-off and landing spots are on rep #1.

dog_takeoff_and_landings_bars_frontHere is my elaboration on the dog’s jumping style relating to the awkward spacing of this jump grid which I referenced earlier.  I have 2 Australian Terriers.  In addition to doing a lot of jump grids with my dogs, I have set up jump grids for four other ATs .  Based on my observations of these six ATs, 7.5 to 8′ appears to be easy spacing for one stride between jumps and about 12′ spacing is easy spacing for two strides between jumps.

The spacing in the grid seen on the video was either deliberately set with descending  spacing (approximately 10′, 9′, and 8′ between jumps), OR the grid was not carefully constructed.  Regardless, the spacing was awkward and uneven yet I think the dog did a decent job navigating through the grid on his first rep even though his “preferred” take-off and landing spots were unconventional.  My gut feeling is this dog had not been trained to look for horizontal bars to determine where jumps are in space and thus focused solely on the visually dominant, bulky spread jumps in addition to expecting the grid was the same on rep #2 as it was on rep #1 (as grids usually are).

I do not see any indication this dog has a “vision defect” that is causing him to take-off earlier than ideal.   In fact I see so many other potential contributing factors in regards to this particular jumping “test,” I am far more likely to think the design of the “test”  is to blame and not the dog.

To reiterate, the overall jumping “picture” remained the same, the jumps were set with descending and awkward spacing between them, and the dog likely lacked the type of advanced jump training needed to be able to navigate over jumps with difficult spacing and changing the  location of jump bars (but not the jumps themselves) between reps.

IMO any combination of those factors could have contributed to this dog crashing and burning over the double and triple jumps when the bars were moved from front to back in the second rep.

The following illustrations take into account  20/70 vision that “normal” dogs have and how nearly impossible it would be for a dog to see that the bars had been moved from his starting position.


ETS_dogs_view_bars_back_lens_blurThere is yet one more aspect of this jumping “test” I think needs to be considered–how a handler’s movement can affect a dog’s performance.

3) Handling.

The dog started the 1st rep standing (vs. sitting) a few feet in front of the 1st jump.  If the handler’s intent was to set up her dog at an easy “set point” I’d say the dog was about one foot too far from the jump.  Regardless, the handler began the 1st repetition with an easy lead-out and the dog took off early but jumped with reasonably ease and landed comfortably beyond each bar.  While the dog was in motion, the handler pointed briefly to a spot just beyond each jump and then dropped her arm until pointing to a spot just beyond the next jump. Coincidentally (or not), the dog landed about the same distance from all jumps, which also happened to be on or near the spots the handler pointed to.

The handler also decelerated when her dog was jumping the triple (jump #4) on all reps.  Based on how the dog stopped and looked at the handler after completing the grid, I assume the reward was always given from the handler’s hand vs. being thrown or placed well beyond the last jump, which would have encouraged the dog to continue forward with strong momentum vs. decelerating and curling back towards the handler.

In the 2nd repletion, the handler did not take a lead out. Instead, this rep began with the dog and handler running together.  I think the dog expected the grid to be the same and thus attempted to repeat the same striding pattern as in rep #1. His take-off and landing spots were very similar to rep #1.  You can see how while airborne over the double, the dog realized the bars are were over his “preferred landing spot” and how he tried to extend his jumping arc but failed.  Then after crashing and burning, he picked himself up and took off closer to #3 as a result. He then shortened his jumping arc to prepare to jump the triple (jump #4) from his “preferred take-off spot.”  And just like the double jump, the dog took off and landed in approximately same spot (as rep #1), landing directly on top of the bars.

In rep #2, the handler did not point to the landing spots like she did in rep #1.  Instead she clutched her hand into a fist and pumped her arms when running along side the dog.   She also turned slightly in toward the dog and decelerated when he was jumping the final jump (the triple), which further encouraged the dog to land short and turn towards her.  I think it is safe to assume this handler did not know that she should have continued to run forward.. full steam ahead.. well beyond the last jump… especially since the last jump was a triple with “moving bars.”

In rep #3, the bars were set at 4″ (Bars in Front).   The handler pumped her arms while running but in this rep, her motion looked more relaxed, her stride lengths were even, and her pace and motion was steady.  After crashing and burning on rep #2, the dog appeared to have learned this was “no ordinary jump grid.”  As I already mentioned, a vast majority of the jump grid “picture” remained the same.  In this rep, the dog appeared to be focusing fully on the bars to determine where they were in space, and adds a stride before the double jump.   IMO, this dog learns fast and has great resilience. What a great little dog!

In rep #4, the bars were set at 4″ (Bars in Back).  The handler and dog “start the race” together.  The handler’s energy was much higher than it was in rep #3.  The dog responded by running ahead and “winning the race.” However with NO incentive to continue driving past the last jump, the dog turned his head to look at the handler while jumping the triple (jump #4) and as a result he knocked the bars.   One thing I noticed is the dog’s confidence appeared to have been shaken by this point and his striding, take-off and landing spots were all over the place.  IMO, the dog was just winging it as best he could.  I also have to wonder if the handler was worrying when the bars were back, and if that unconsciously affected her handling… creating a self-fulfilling prophecy of sorts.

So while I agree with the statement in the description of the video on YouTube that bars set at the backs of spread jumps can cause visual confusion for dogs, I don’t think it automatically signifies that a dog has a “Vision Defect / ETS.”  I am more inclined to think this dog was not trained to find closer take-off spots or to focus on horizontal bars vs. wings and stanchions as a means of determining where obstacles are in space.

If the dog in the video was my dog, first of all, I would be THRILLED.  He is smart, courageous, athletic, and 100% willing to give it another go, even after being inadvertently tricked into crashing and burning.  If he was my dog, I would take a few months off from jumping and start fresh, which is exactly what I did with my dogs two years ago and it worked brilliantly.

Below are videos of a few jumping games specifically related to spread jumps.

(video below) This set-up teaches dogs to step into the “trap” of double and triple jumps.  It begins with one jump with a single bar in the back and the dog starting at an easy “set point” distance in relation to the bar.. not the front of the wing. Using a mat on the landing side of the jump encourages dogs who tend to land close, to extend their jumping arcs so they land a bit longer.  Obviously a dog needs to have the foundation skills (of targeting a mat) in place for the mat to be effective.

(video below) The game eventually progresses to widely spaced, fast jump grids like this one.

*Parallax is a term used to describe how objects appear to be moving at different speeds, depending on their distances.  For example, when you are driving a car, distant objects appear to be moving very slowly.  As a result, you can easily read street signs, see other cars, buildings, mountains, etc.  But the closer those objects become, the faster they appear to be moving (across the lens of your eye) so by the time you pass a road sign, it appears to be flying by.

Here are two very short videos of object overlay and parallax in action.  These animations  simulate the dog’s view as he is running through the jump grid.  I did not include the up and down motion of jumping as that would have taken a long time to set up but I think it is interesting to see how much more movement the spread jump wings generate as they move across the lens of the camera… compared to the straight-on jump bars.   It’s also interesting to compare the view with “bars on front” with “bars on back.”

(above) Bars in Front

(above) Bars in Back

Dogs see motion far better than humans and it doesn’t matter if the motion is occurring out there in the real world or if the dog is moving which causes static objects/ images to move across the lens of the dog’s eyes. This might explain why dogs need to be trained to focus on jump bars… as they are moving less across the lens of their eyes than the uprights.. in particular 3d spread jump wings.

One more relevant point is accentuated by fact that the spread jumps (in the original video and in the animations) are red and blue yet both have solid white horizontal supports which happen to match the solid white jump bars.   I believe this contributed to the overall visual confusion of the jumping “picture.”  I also think horizontal jump supports are far more likely to cause visual confusion for mini-dogs than with medium and large dogs, due to the proximity of the supports to the jump bars with mini-dogs.

For more info on canine vision go to: https://artanddogblog.wordpress.com/2013/07/10/when-the-emperor-with-no-clothes-happens-to-be-an-agility-fanatic/

There are also many other posts related to jumping issues and solutions on: https://artanddogblog.wordpress.com/

Enter the word “jumping” in the search field on the right hand side of the home page to see a list of posts related to jumping issues and solutions.

And finally, if you know anyone who might benefit from reading this post, please share the link.  I think it is important to provide people with different perspectives so they can draw their own conclusions.  https://artanddogblog.wordpress.com/2014/01/13/update-on-the-video-of-an-australian-terrier-used-as-an-example-of-ets-in-2010/

1 thought on “Update on the video of an Australian Terrier used as an example of ETS in 2010

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s