Wednesday, April 4, 2012

Retraining an AFrame

More than anything I wanted to put these thoughts down for myself.  One, to remember what I have done, and two, it helps me solidify my thoughts on this topic.

I decided about a year and a half ago (maybe more) that I needed to retrain Split from a 2o2o AFrame (AF) to a running AF.  He clearly didn't like the pressure on his shoulders and after 5 or so reps he would start limping on the front-end.  Plus, I just really like the fluidity that it gives my runs.

I tried many things and some I believe contributed to the solution and some didn't.

Split is a boy that wants to do the right thing, is very sensitive, and is a little (a lot) Velcro.   His personality is key in how I trained things.  It is hard to work with sometimes and almost always I have to adjust my plan.

The first thing I needed to do was "break" the 2o2o and have him run his AF.  I put the AF  down low so that it would appear different to him.  It took me about a week of practice every day to get him to start running it without hesitation (see doesn't take long to ruin criteria does it?).  When it was down low he would run it without a leap.  Once I started to move it up, he began to leap (it was probably less than 5 feet high).  I tried marking the correct behavior, but things didn't really improve.  He enjoyed leaping and sometimes would creep to make me happy.  He just didn't get it.

Frustration and many weeks later...

OK, so now he is leaping.  I decided to train him to hit a box on the ground, similar to Rachel Sanders box method.  I trained Split so he would run away from me, intentionally hit the box and keep running (to catch a ball or something).  Although I don't remember exactly, I think this probably took several weeks to get the behavior that I was happy with.  Once I had this I put the box on the AF. 

Fast forward many months...

When the box was on the AF he would hit it.  Trying to fade the box was murder!  Also, a gnawing feeling in me just knew he didn't really "understand" the criteria.  After all, how do you explain that to a dog.  Really, this is about muscle memory to a degree.

Frustration again, and a break from training this (after all, I wasn't getting a reward)...

While re-training Split on the AFI began to train Tangle on a running DW .  Let me just say, everyone should train a running dogwalk.  Not to get the end behavior, but because it teaches you and your dog so many valuable things (maybe I will blog on this some day).  Seriously, this is a fabulous experience but it does take diligence.

The training of the running dogwalk gave me a much deeper understanding of how to teach a dog to run a contact.  I learned details on how to progress through the process that I never would have learned retraining Splitty.  So, I put Split through running AF 101 school which Tangle was doing running DW 101.  I didn't want to retrain Split's dogwalk, just too much.  But, fundamentally all the principals that I learned with Tangle applied to Split and his AF.

AF went down low again, clicker came out, and the treat-n-train.  Another obstacle to overcome with Split!  Some days he would be afraid of the clicker sound and almost always the treat-n-train.  Back to the tennis ball!

In order to not make this a novel, basically I followed many of the steps that Sylvia uses for training a running DW.  Low obstacle, very little stimulation, low motion, lots of repetition, gradually increase all of those.  Always, jackpot and high reward for getting it right.  Training the AF was now Split's favorite game!

Again, once I raised it to about 5 feet he began to leap.  Frustration set in!  This is when I began to put a stride regulator on AF between the change of color.  There are two things that this did   1) Split's stride would ALWAYS hit the yellow 2) it allowed me to get his performance on a higher AF to the point of frequent reward--key!!!  I ran the AF this way for probably a month.  Slowly adding more motion and other obstacles.

When I took the stride regulator off I had leaping back again.  Another round of frustration! 

But once I sat back and thought of everything I learned with Tangle and running DW the emotion wained and some constructive thoughts set in (I learned this process from training Tangle's running DW).  I removed my motion, stimulation and obstacles and left the AF high.  I began to reward Split for successes and not reward for failure.  Presto, I was beginning to get different results!

To me, the key was teaching him what success was (the look, sound, and my reaction).  That is, when he did something right what would happen and when he did something wrong what DIDN'T happen.  When Split got his AF right I would throw a huge party (not too much happy sounds because he is emotion sensitive I had to increment into this), throw the ball, and give a boat load of praise.  When he didn't get it right, there was no reaction, just walk back and try it again.

Within days I began to see Split adjust his stride to "get it right"! 

So, we are now several months into this break through.  We have done a couple of trials without any AF calls.  I still work the AF 4-5 times a week in a training mode.  Each session starts with little motion and increments into full motion and obstacles.

Can I call him "retrained"?  Nope, not yet!  But have we hit a major success milestone? Yep, we have!

P.S. There are a ton of thoughts out there on what the "proper" running AF performance looks like.  Leaping over the apex, almost touching their belly over the apex, only one stride on the downside, two strides on the downside and on and on.  I now have three dogs with a running AF.  Each performance is slightly different.  The answer is "it depends" and ultimately what can your dog do confidently and reliably matters!

Split's progress to date, http://youtu.be/vZiwfyocu2Y



Split and Tangle with stride regulator, http://www.youtube.com/watch?v=QGrB8XgXS_Y

5 comments:

Rose said...

Boy, can I relate to this post. I started retraining the AF using Rachel Sander's method 15 months ago. Did too many reps and my dog came up lame. After dealing with that, I resolved to stop practicing a running AF and accept whatever I got at trial since no way will I go back to a 2o2o AF with this dog.

About three weeks ago, I started using Silvia Trkman's method. Much lower impact, but it remains to be seen if it will indeed transfer to the AF.

Mary said...

Rose, thanks for your comments. It made me re-read what I had written, that was fun and enlightening. I have learned so much about running contacts since then. I have a much more solid idea of how to train/retrain ONLY a running AF (without a running DW).

agileborderterriers said...

I did a google search on running A Frames and fading the stride regulator and found your blog. The ups and downs of my running AF journey has a lot of similarity to yours. Except that I'm still not convinced this will all work out. Oh and I am only interested in a running AF, not a running DW.

After a 2o2o that had become creepier and creeper with my border collie Zappa, I watched all the running contact DVDs available and started our running AF journey with RS's box method. We had great success with the AF up to full height and as long as the box was on the AF, it didn't matter what I did handling or speed-wise he gave me a two-hit downside. And then I tried to start fading the box. I had heard that the fading was the hard part, but my dog had been so brilliant so far that I thought things would be different for us. Ha! It completely fell apart. I cut out the middle portion of the top bar of the box and the dog gave me one hit right ion that space between the cut bar and that was it.

So I took a break and thought about it. I decided to see if I could fade the box into just one stride regulator (what would have been the top bar of the box). That transition was no problem. So then I experimented with how much I could cut that bar on each side before the behavior would fall apart. It seems that as long as the middle part of the SR is in place, he still gives me 2 hits, the second one in the yellow. But when that middle portion is too small, the behavior falls apart again. The experimentation was done in one session just to see, so I should probably fade more slowly, and indeed at this point that is the plan. But here is what I have observed...

Zappa makes up his mind what he will do as he crests over the apex. If he sees the SR, he will shorten his first stride and his first hit will be 3/4 of the way down the coloured part and the 2nd hit will be in the yellow. If he doesn't see the SR or the SR is too short to be of any concern to him, he will hit very deep into the coloured part, just above the yellow and then he leaps off. So only one hit.

I have some of my friends saying that since in order to get two hits Zappa has to decide to compress his first stride on the downside, it will not hold up when he is high in competition. Since you've been at this running AF thing for quite a while now, and your journey has some similarity to mine, I'm hoping you can give me your thoughts and experience on this. Does the decision to shorten the first stride hold up in competition with no SR??

My other question is when you started over and decided to use Sylvia Trkman’s method on the AF, I presume you stopped using a SR, correct? When I was choosing which method to use, I watched ST DVD and didn’t think I had a good enough eye to reward correctly using her method. The box method seemed easier for me to grasp. Plus ST’s mthod focuses on the DW and I am only interested in the AF.

I’d love to hear how your running AF has evolved since your original blog post.

Mary said...

agileborderterriers, I might be able to provide some insight or perhaps some things to look at. I put a contact form on the blog. Give me your email address and we can converse that way. I think email will be easier.

agileborderterriers said...

I don't see the contact form. Please contact me at agileborderterriers at gmail dot com I look forward to hearing from you. I'm crossing my fingers that your insights will help me through this!!
Sylvie