I have been working with Bandit on platforms to assist him with learning positions.
Strictly speaking, platform work is supposed to be shaped. The initial "four on the platform" behavior is shaped, and then the platform becomes "magnetized" as the dog is heavily reinforced for putting four paws on it.
Then the platform is placed in position and the dog learns to come into position, though a shaping-with-props process, the positions are put on cue, and the platforms are faded.
As I have been working with Bandit on the platforms, we have encountered a bit of a difficulty. Bandit does a beautiful job coming into front position on the platform. No surprise there - it is most natural, even if he is coming into position from beside me, to move into a position where he is facing me. But we have run into problems when I have moved the platform into heel and side position.
I have done the step where he remains on the platform and I move around and click/treat so he can get the feel for proper position. I have worked a gradual approach to the platform.
However, we consistently reach a point as the criteria increases and he is approaching the platform from a more difficult angle, where he gets stuck.
He will get onto the platform backwards so he is in heel or side position, but facing behind me (which would be fine if we were working on that, but we aren't!). Or, he will start to avoid the platform altogether and start to circle me.
If I were seeing some degree of enjoyment on his part, in working to figure out what I want through the shaping process, I would continue with it. I might break it back down a bit, and raise criteria more slowly. However, I was clearly see frustration on his part. Any learning that was happening was not the learning that I want to happen.
When I see frustration, I know it is time to reconsider my approach. The last thing I want to do is for Bandit to associate platforms with frustration. We had that experience with some pieces of Agility equipment, and it took some work to un-do the damage that was done.
So, I decided to help Bandit by giving him a clear hand cue to indicate that he should head to the rear of the platform. Once he did that, I could remove the hand and he would get into position.
Success!!
Here is a video of our first session doing this. There is some front work first.
After sharing the video, I was asked by a friend, "Why did you use the hand? Why didn't you shape it?"
That is an excellent question!
The reason is simply this: after careful consideration, taking Bandit's learning style and training history into account, I deduced that giving him a quick hand target was the best way to help him learn.
It was not so much a matter of him getting the idea quickly, but of him getting the idea without dealing with undue frustration.
Don't get me wrong - frustration is a reality in training. But I do feel very strongly that it ought to be minimized, and if it is going to be detrimental to the training process, it should be avoided if at all possible. In this case, I could see the potential of creating an actual issue, and I did not want to do that. Add to that, I knew of a way to convey what I want to Bandit in a way that is very clear to him and has the potential to facilitate learning effectively. So, I chose a training tool (hand target) that is very clear to Bandit, and he got the idea very quickly.
I expect that within two - three more sessions, he going to be getting onto that platform from any direction without me having to give the hand cue.
I plan to have a section in Bandit's Freestyle routine where he will heel with me, twirl, heel, make a large twirl, and then heel with me and do a twirl transition into center position.
It's a pretty basic sequence. We are working on building the "big" twirl.
Once he is very fluent around this "barrel", I will switch it out for gates that form about the same size circle. Then we will fade it down with staggered targets or something.
A while ago I felt that Bandit and I were reaching something of a plateau. He had made a good bit of progress with everything that we were working on, but it seemed that he had stalled out in many ways.
I really thought that I was going to need to find a way to ask more of him, but in the end I made the decision to ask less.
And the result was astonishing! He has literally smashed that plateau into pieces and we are flying forward once again.
One example of this is our heeling. I realized that we had reached a point where I needed to start getting the food out of my hands when we were working on heeling skills. Any time I tried working with him without food, however, his nose was on the ground at once.
I decided to break down the criteria into smaller pieces, and to go back to a 100% rate of reinforcement.
I set Bandit up in heel, in a sit. I had food in my hand. I cued him to heel and we moved forward about 4 feet. At first I kept that food right on his nose, and he got the treat at about the 4 foot point.
We repeated that a few times and then I moved the treat up away from him so it became a visual target, not a scent target. Success!
Finally, I did a repetition with the visual target and reinforced, and then did a repetition with no food in my hand.
BIG SUCCESS!
Before long I was able to start asking Bandit to heel between a few signs in Rally class (not all) with no food in my hand.
He is doing a phenomenal job. We did something similar with inside turns. He was just not understanding them, so I worked with him around the four corners of a chair. After doing that a few times, I noticed major improvement on his left turns.
It is counter-intuitive. Sometimes when it seems like I need to up the challenge, what I really need to do is break it down.
The day will come when Bandit successfully completes the entire Figure 8 Waltz for the Cyber Rally-O Dance Divisions. First, though, he needs to master some simpler skills.
The plateau has been dispensed with. We are once again moving onward and upward!
Bandit and I are currently taking a distance Agility class through the Fenzi Academy. We have really only gotten started.
The first thing that we did for the class was identify our "base cues". These are the cues that I will take the time to train with him as his foundation.
The base cues that I have chosen for Bandit are:
"Go", which will mean "run out and take whatever is directly ahead of you and keep running forward"
"Fly", which will mean, "go out to take something and come back to me"
"Out", which will mean, "go out and stay 'out there' taking equipment, or take the far side of a discrimination"
"Here", which will indicate the close side of a discrimination.
"Switch", which will mean "flip away from me as I flip toward you to change sides"
I am also considering right and left directionals, but I am going to focus on the more basic cues first.
I like having this very specific plan for Bandit's training.
In the class we are training our base cues on cones. This video is Bandit's "baseline" video for his cone work.
He did a send out with me doing a front cross and a send out with me handling it as a wrap. Both of these could be "fly" or "out" skills, depending on the context of the course. And then we worked on a "pass by", which sets us up for a "go" or "out", again, depending on the course.
Bandit's drive and speed amaze me. He locks in on the task (flying around the cone) and he just throws all of himself into it!
I am not used to that at all!
I am looking forward to working through the rest of this class with him!
Bandit and I are at a point now where I want to begin to take his heelwork training more seriously. Not "seriously" in the sense of making it "serious" for him, but "serious" in the sense of giving him more structure and a higher level of criteria in our heelwork.
The first question that I needed to consider was: what picture do I want to create with his heeling?
I came to the conclusion that I want the picture to be very similar to the picture that Dean and I created in his heyday. Bandit's style is a little bit different from Dean's, but I would like for Bandit to have the same level of understanding of movement in position that Dean has.
Here is a video of Dean doing some pretty good heeling. This was actually not even his best, but it is very good. Keep an eye on how he moves his body when he is on the inside curve.
It is clear that Dean knows how to move his entire body in order to maintain heel position, regardless of the direction that he and I are moving together. He can read the inside curves and adjust accordingly, and he can drive a bit on the outside curves. You don't see it in this video, but he can also move in a straight line and maintain his position.
I want that for Bandit.
In this video, it is obvious that Bandit and I have gotten a good start. He has a basic idea of moving in position, and he has a lot of enthusiasm, and his movement is very pretty.
He drives very nicely on an outside curve, but he doesn't have any idea what to do when he is on an inside curve.
And, even though this was taken last August, I can't say he and I have done a whole lot of work on it since!
I have considered this, and I believe that Bandit needs to learn how to move his entire body on an inside curve. We have done pivot work with a pivot disk, but I don't believe that Bandit is generalizing from the pivot disk to heelwork.
I am going to do some more foundation clicker work with him, and focus in on allowing him to choose to make a left turn and move his rear end in. We will do this with a 90 degree edge, such as a chair.
I am looking forward to getting into this work with him. It is my goal, by the end of May, for Bandit to be able to move with me 10 feet in any direction and maintain heel or side position!
