Training Dog Agility when Visually Impaired

On average I have 6-10 degrees of functional vision.  In certain lighting conditions I have less.  With my glasses on I have 20/20 central vision.

In doing dog agility, or training dogs, I have to make some modifications.  And often get creative.

My #1 tenet is: Train it until I trust it.

And my #2 is: Trust the dog.

When I’m out on course, depending on where I am in relation to obstacles and my dog at any given time, I likely will not see her.  So I have to strategically chose handling paths, that put me into the spots where seeing her is additionally helpful.  And where I can hear what she is doing, thankfully I have extremely good hearing.

In training Zora, I’ve learned what her path and footfalls sound like when she is doing her contact behavior properly.  I’ve learned what she sounds like as she races through a tunnel.  I’ve learned what she sounds like as she digs in on a tight turn.

To meet my #1 tenet: Train it until I trust it, I set very specific criteria that I then ask the dog to be able to repeat in any given set up.  I use a lot of external tools to help me maintain the criteria during training.  Some of the tools I use allow me to know if Zora did the behavior I’m seeking without my having to see it.  Some of the tools I use allow me to better learn what her body sounds like when she is doing the behavior I am seeking.  Some of the tools I use allow me to move strategically while building her confidence with me at a distance.

For example:  Mark Buckets.  I’ve built a strong affinity to racing to them and putting her front feet on them for Zora.  I can send her to one, from a path of other obstacles.  Giving her this end place of value, means I can practice moving to other spaces in the sequence that give me the data I need while she drives along the correct desired path to eventually hit her mark bucket.

Or hit it type boards.  I duct tape clickers to stuff.  The i-clicks are excellent for this.  Last year a client actually gave me an official Hit It board they were no longer using, but before I had that, I duct taped clickers to a lot of stuff. Taking a light weight board, duct tape a clicker to the underside, dog touches the board, runs over the board, moves on to the board, weight of dog causes clicker to press against the ground.  Sound tells me dog is in contact with the board.  I place the board where I want the dog to be.  And now I have an external way of maintaining criteria independent of vision.  Useful for teaching contact behaviors, knowing when the dog has stayed on path (as if they are on path say coming out of a tunnel or off a jump they will move over the board), and even when teaching things like weaves.  For ones where dogs will be moving at speed, I use a textured board for traction.  For other things I have light weight pieces of plexiglass with i-clicks taped to the underside.

Speaking of weaves.  Zora is the first dog in many years that I didn’t use an external accountable way (ie guide wires or x-pen cages) to teach weaves.  For her it was because I felt she wouldn’t learn weaves well using those methods.  Though using wires or cages is certainly easier for me, again external non vision based accountability for the behavior I’m trying to get consistently.  With her, I did a modified version of the 2×2 method and used a Treat N Train.  The treat n train allowed me to position it to better reinforce the exit path I wanted to condition for Zora, and again meant I didn’t have to place the food myself.   The 2×2 method of progressively building each in-out weave motion pattern allowed me to also position myself to be able to see the pattern as Zora learned each step.

I train my directional cues progressively, first beginning on the flat without equipment and then building up equipment and distance.  I have very specific set criteria and definitions for what each cue means, and what behavior exactly I want Zora to do with each cue.  I train them to reliable.  Out means I want a foot fall change away from me.  Here means come in to the inside path nearest me.  Switch is turn away from me.  Go is move on a straight line.  Tight is turn in toward me as a wrap cue.  It is the rare discrimination challenge on course that I actually see her do it, but because her out and here cues are trained to reliable, it’s actually rather rare we don’t complete the discrimination correctly as indicated on course.

My obstacle cues include not just the obstacle but the path onto and off of the obstacle I want the dog to take.  For example my ‘walk it’ cue includes training in the approach to the dog walk, the performance of the dog walk itself, the contact behavior, and the exit off the dog walk.  Training those pieces onto one ‘walk it’ cue means Zora has her job and it’s one trained to a degree that I can learn to trust the sounds her feet make as she performs it to check she’s done it as trained.  I can hear if she hits the down ramp on the wrong foot fall pattern and as such know she’s likely to over stride the contact and add in a ‘steady’ to get her to check her stride if needed before she hits the contact zone.

As often leash runners at trials place the leash on the ground, or on a chair, Zora has a ‘find your leash’ cue.  Which means run to it and stand on it.  She doesn’t grab it with her mouth.  She doesn’t play with it.  It’s simply a ‘find it’ cue.  I also made her a leash that bright orange and more likely to stand out against the green grass. I’ve also learned as we finish the course to move myself in a way that encourages leash runners who are still holding the leash to shove it into my hands.  At this point, there are a number of folks in my local trialing area who also know I may not see where they put the leash and they are good about handing it to me.

Another thing I do, both when training and trialing, is use video a TON.  In the actual moment, there is a high chance I will not see what Zora just did.  Video replay helps me immensely.  I learn what we need to train further.  What exactly happened at any given spot on course.  What went well.  What didn’t.  Why x, y or z happened.  And to clue me in further to how Zora response to my positioning and whether what we have been training is having the results I am seeking or if we need to go back to the drawing board.

I’ll also let you in on a little secret.  I don’t teach my dogs to make eye contact.  None of my dogs have a look at me cue.  Sure I teach it to my students, I have a really good way of teaching dogs to actively seek out eye contact and through that students can then use advanced eye contact work to additionally move their dogs in space.  But it’s not necessary for successful dog agility work.  And I don’t use it with my own dogs.  I do work hard to be consistent with my body cues, always a work in progress of course, and using my head positioning, but eye contact itself isn’t part of my training criteria.

When we’re out on our practice field training agility at home, Tom also has a job.  I didn’t teach it to him, but I do reinforce it.  He helps keep me from falling over equipment I forgot was there.  Tom’s guide dog training made him an expert at keeping tabs on me in space, and he’s become really adept at body blocking me when I’m training Zora and fully focused on what I’m doing with her, having forgotten where I’ve placed all of the jumps, or tunnel, or the dog walk.  Tom prevents many injuries when I’m training agility with Zora.

When trialing, I chose my venue in part due to risk of injury in other venues.  In NADAC the course design is most often such that I have a path free of obstacles.  The couple of times I ran AKC agility years ago, each time I took out a jump.  Once I smashed into the pause box placed in the freaking middle of the course right in my path.  It was painful and unpleasant.  I’ve not run AKC agility since.  And as much as I’d prefer to run agility outside at a trial, reality is indoor trials are easier and safer for me.  The footing is reliable.  The lighting is reliable.  Outside the lighting can change moment to moment, time of day to time of day, meaning what I can and cannot see can fluctuate greatly.

Many people who run dog agility, micro manage many things in their dog and compete despite having many incompletely trained behaviors.  Unreliable contact responses.  Dogs not actually jump trained.  Lack of start line stay.  Lack of understanding of directional cues.  And so on.  That may work for them, and that’s great for them.  It doesn’t work for me.  I need to train to a level I can trust my training and trust my dog to do her job reliably in order to successfully compete.