During the handler education lesson last week we worked on reading sheep and doing efficient feeding with two groups. Remember that lesson was about me learning to read sheep, sort in an intelligent manner and do the work without the dog, so I can better do the job with the dog. Part of that process is only picking up half the flock and putting them up, so i don’t have to do the extra step of sorting them out into two groups manually. Which sounds easy enough until you remember that not only do stockdogs naturally want the group things together, but i’ve actively encouraged that behavior in my dog. While the two groups of sheep do self separate, they don’t tend to self separate very far. If she comes in too tight on the outrun for one group, then they will recombine into one mass flock again and i end up gate sorting anyway. (read: waste time and energy, probably while standing in the rain)
The original problem with picking up one group, is teaching her to just take one group and ignore the other. Working against all her instinct and all my shaping.
When teaching a new behavior, it is usually good form to break the behavior down to its small parts. This allows you to raise the chances of getting it right, rewarding the right behavior and repeating the right behavior without extra behaviors to unlearn later.
- Stop on balance
- Come to handler
- Walk into pressure
- Leave Stock Behind/Focus on desired stock
As individual behaviors, these are no big deal. Put them together and one runs into the Feet Moving, Brain Stopping vortex. This vortex stops you from taking the easiest path: send the dog out, stop the dog once it has gotten just to the eye of the sheep you want and bringing those sheep in. It is an outrun, right? Sure.
My assignment was to take my 20 sheep, make a hole between two groups and call my dog through with some enthusiasm.
That is not exactly what i did. It is what i tried, but i have discovered that Molly generally does new things with caution. Enthusiasm comes later and usually includes ignoring me because she understands the job.
The best way to make a visual is to imagine a rectangle. All things being equal, Molly and i are at the long sides and the sheep are going from short end to short end. If we’re smart and fast then the sheep are still moving and we can use existing motion to find and work a hole. Really they’re not moving, but they are spread out a bit. Since we’re just starting to figure out what is going on, i find a little self-made-sheepie-hole where the group is starting to separate. I walk in to the hole and use my pressure to make the hole big enough to comfortably call Molly to me. All while stopping Molly from ‘helping’ by bringing the group back together, but that’s beside the point. So i walk into the hole, face the part of the group that i want and I am *supposed* to just let her run though the group. But there isn’t a lot of running going on. There is a lot of caution confusion.
Change the training scenario to match the dog.
Instead of working for the run through, i worked for her to walk to me while i’m standing in the hole facing the group i want. After a few failures, i took her away from the sheep and sharpened up her walking to my side or bum. After that little bit of groundwork the expectation seemed to be more clear and the rate of success was higher.
Once she started to walk to me, we at first walked together and pushed the group away. I’ve found it is helpful to show her once or twice what i want her to do. ‘this is what i want. This behavior, this speed.’ I try not to do it more than twice because after that she starts to think that is the entire behavior as opposed to a template. With the template in place, i started to then quickly get around the sheep so she was positioned in a fetch and she’d bring that group to me.
The first time we tried the fetch, she started to look back at the group we left. I had to make a quick decision: use the opportunity now to teach a proper Look Back, or stick with the plan and finish the shed? I decided to stick with the plan as the situation was not unique and could be replicated later. Besides, attempting to install two behaviors at once can be confusing. Instead, I went with a command she is familiar with and matched it with the behavior i wanted: Leave Them; followed quickly with Walk up to encourage the fetch. I rewarded the whole series with a walk about with our group that included some turns that required covering the group. Comfortable behaviors with clear and agreed upon expectations.
By the end of our little 1 hour session, our completion rate was about 50%.
Not bad for the first day. Most of our incompletions were where she didn’t come in and instead put the group together. This seemed to happen when the hole was not big enough. There were very few outright rejections to move and those all happened when i was too close to the fence. The next step will be me standing not so much in the hole and teaching her to walk into the pressure. But also to get some confidence going. I do want her to get some speed and thus some power going. Not every group of sheep we ever work will be as amiable.
As a side note: In the process of training my dogs to do *everything* i find myself coming back to the interview with Joy and Maggie Sisler regarding Jay Sisler that is the Working Aussie Source Stockdog Library. It’s not only a great article from a breed history point of view, but it reminds us all that training theory does ebb and flow. Much like in couture, Everything that is old becomes new again. Jay found a process that worked for him and his dogs. If you need a visual that understanding your dog and how to communicate with your dog is important, look at what Jay accomplished by taking the time to find a common language and understanding how his dogs learned.