When it comes to positive reinforcement based clicker training, there are four major ways of getting behavior: shaping, luring, targeting, and capturing. There are other ways, including physical modeling, but they have limited applications and are rarely used by most force-free trainers. In this post, I’ll address shaping, and in the coming weeks, I’ll follow-up with the others.
What is shaping? Shaping is a method of building behavior “from scratch” by clicking successive approximations towards an end behavior. Often a trainer will create a mental or written “shaping plan” that lays out the steps it might take to get the final desired behavior.
One popular way to introduce shaping is when teaching a dog to get on a mat. The goal behavior is that the dog lies down on the mat for an extended period.
In the beginning, the trainer will start by clicking and treating their dog just for looking at the mat. This is usually followed by sniffing the mat, brushing up against it, or putting a paw on it, all of which can also be clicked. After the dog has received several rewards for interacting with the mat in this way, they will start to experiment just a little, often by purposefully putting a paw or two on the mat. Click!
From there, the trainer waits for the dog to put three or four of her paws on the mat before clicking and treating. When the dog demonstrates an understanding of “paws on the mat = treat”, the trainer patiently waits for the dog to offer a sit or a down. Some dogs offer sitting right away and later relax into a down, while others flop over into a down right away.
It’s very important to understand that the trainer remains relatively quiet during this process and does not prompt the dog by talking, saying cues, patting the object, or touching the dog. All of these things mean nothing to the dog! He can’t comprehend English and he doesn’t know the desired behavior yet. At best, it’s luring which must be faded later; at best, it confuses the dog and slows down the learning process.
The dog must figure out what behavior earns him a click and a treat. (Praise and happy talk after the dog has interacted with the mat and received a click is just fine, but you should stay quiet while the dog is figuring things out.)
Instead, the trainer should manipulate the environment to set the dog up for success. When I teach group training classes, I tend to work on matwork as the last exercise, when the dogs are comfortable and are a little tired. I have students sit on their mats during class and only take them out when it’s time to work on mat training, and I instruct them to be prepared before putting the mat down. They need to be ready to click the instant the dog starts looking at or sniffing the mat, which is often before it even hits the floor!
I recommend that students to practice matwork in the evening when the dog is already considering taking a nap. Sleepy dogs are more likely to lie down on an object than wound-up, excited dogs. The flip side to this, of course, is that if you are shaping a fast movement-oriented behavior like agility obstacles, tricks that involve movement, or coming when called, do it when your dog is awake and full of energy.