We are on our regular walk with the dogs. One of the things that we do while hiking with them is practicing coming when called. If we want reliability, we must practice. If we want reliability, we must also pay the dog.
Deuce is really, really reliable at recall, and that makes me smile with pride. Once I figured out that playing tug was very reinforcing for him I began to use it as a reward for coming back. In no time his recall went from 80% reliability to almost 100% on every hike. How cool is that, people?
This morning I asked John, who walks the dogs many times without me, to play tug with Deuce at a particular spot. The reasoning behind this is twofold. Unfortunately for me, I cannot play as much tug as I want with my dogs due to an injury to my right hand. In addition, John plays regularly with him on walks, but kind of on the same part of the hike. Today, I am asking him to induce Deuce to play in a spot that he normally does not.
I explain to John that unpredictability is important in training dogs. Yes, indeed I can make the same argument about being predictable of course, but today I want to focus on being unpredictable.
Unpredictability adds another dimension to training and the acquisition of behavior because it makes it fun.
In our case with Deuce, this is a dog that is so interested in tugging, that he walks and runs as he turns towards us in the event that we might just pull the tug out for him or ask him at long distance to “take it”.
Ah, magical words for this dog. I explain to John that if he only pulls the tug on the same spots over and over again the recall might just fall apart. You see, dogs are excellent at flow charts: When this happens, then this other thing takes place, etc. Before long your dog is kinda training you instead of you being able to control the consequences which are at the crux of getting behavior.
In addition, when we use unpredictability as explained above, by default we are using a thinner schedule of reinforcement which in lay talk means that dog is only getting reinforced in some occasions.
Depending on which schedule of reinforcement used which instant of behavior is getting reinforced.
For example: I might reinforce with a game of tug for the return from Deuce that fall under 3 seconds say. Or only when he runs at a certain speed, etc. I can also choose to reinforce a certain % of all the times he comes when called as another example of a thinner schedule.
Unpredictability in reinforcing works wonders because apparently, there is an element of frustration and I would argue hope for future reinforcements. Think of betting at a casino. Those folks are working just like my pup at a thinner schedule of reinforcement and the behavior (coming or betting) goes up.
Of course, I am interested in reliability because coming when called can be a life saver for my pups but besides this reliability I love playing little games with my dogs. I enjoy engaging with them in manners that brings them joy.
Have you ever realized that we too are motivated by the expectation of something we want hard?
Say a trip. Have you noticed that often the anticipation of the trip with all the planning and day dreaming with expectation is actually more fun than the trip? This is what I am talking about here!
As we walk on our hikes I relish knowing that Deuce is attentive because he so much wants to have an opportunity to go for the tug and this game that we are playing makes the walk so much more enjoyable for the two of us.
So here is my advice to you. Next time you are thinking of reinforcing your dog, think about what (or how) you want your dog to perform that given task. See if you can observe closely and begin to reinforce for more accuracy, cuteness, etc. Now, do keep in mind that making sure the dog can perform first the behavior is mandatory for a good outcome. As a form of example: If my dog has not yet learned to lie down, how then can I request he lies “sphinx” style with both hips equally tucked and in perfect symmetry?
If my dog has yet not learned to come to me when he is 10ft away under certain specific conditions (wild life, other dogs, people etc.) how then can I expect for him to come at 20ft or even at a great distance under those same circumstances? You get the picture?