Recall Training: Dealing with Distractions
If you joined us yesterday on the Dogster Guide to Behavior & Training, you were asked to create a distraction hierarchy for your dog as a framework for introducing distractions to recall training.
It may not be a bad idea to further split this list into "stimuli that my dog can interact with safely" and "dangerous" stimuli. If the "stimuli your dog can interact with safely" are things that your dog is excited, rather than frightened by, you can actually incorporate these distractions into your reinforcement sequence (fresh bowl of water, gopher hole to sniff, toy to play with, friendly other dog, etc.) through application of the Premack Principle, which asserts that "less probable behavior (recall) can be reinforced by more probable behavior (squirrel chase)." Essentially, Premack is the "If you eat your broccoli, you can have your ice cream" principle.
In doing so, you would blow your recall signal, reinforce your dog heavily after (remember to sing your ABC's, as we discussed last week!), and occasionally release your dog after to interact with the distraction object. For every 10 recalls you practice with this distraction, I would allow 1 - 2 interactions with the distraction object. I don't want to allow the squirrel chase, gopher sniff, or other dog greeting every single time I recalled my dog, because then you build anticipation and expectation - your dog will begin to expect access to the distraction each time and may in fact find it punishing in future sessions if you blow your recall signal and he is not provided the interaction opportunity.
I like to follow this sequence when I'm introducing distractions to the recall:
- Practice around all ten of your level one distractions. Aim for one or two practice sessions a day, using a single level one distraction. Practice just one or two whistle blows per session. At this point, we are not asking for any distance - you are still right near your dog. You can start at a distance from the object and gradually work closer if you'd like. Practice until you are right near the object and the dog won't even glance at it when they hear your whistle blow.
- Begin adding distance - once your dog is ignoring the object, see if you can place the object between you and your dog and if you can call your dog to you coming past the object. Feel free to use a long line to prevent your dog from self-reinforcing for blowing off your recall. If your dog chooses to interact with the object and "blows off" your recall, grab his leash, walk him inside, and end the session for now. When you resume your sessions, go back to practicing near the object for a while and when you reintroduce the object between you and your dog, consider shortening the distance temporarily.
- Repeat until your dog will gladly run past any of the level one distractions in response to your recall signal.
- Remember that these distraction values are fluid and cumulative. Multiple level one distractions may combine to equal a higher level distraction. The more you practice around these distractions, the less distracting they become - once you practice around all of these level one distractions for a while, they become more like "level .2 distractions." Practice these same steps using two, then three, then four, and eventually five level one distractions before you move to level two distractions.
- Practice around these distractions in new environments. If you've only practiced in your yard, consider loading your car up with a long line and some "mobile distractions" and taking them to a large empty field, a fairly quiet local park, etc.
- Repeat these same steps at all distraction levels. Introduce distractions first, then distance. Then reduce distance and work with multiple distractions at the current level you're working at. Then increase distance with multiple distractions! When you are working at level two and using multiple distractions, feel free to add a few level ones in there for good measure!
Stay tuned - tomorrow we'll talk about trouble shooting. Why do good recalls go bad and how can you prevent these regressions in performance? Stay tuned to find out tomorrow! Until then, in case you missed our previous entries on recall training, here they are: