Before You Fade Reinforcement
Your dog should be really solid on the behaviour before you start to fade reinforcement. For each exercise, you should have gone through the following process.
- Train new behaviours in your special low distraction training area
- Move to different areas of your home
- Train in the backyard
- Train in the front yard
Train in new places
- Generalize the behaviour to all these areas to become reliable
When Will You Stop Reinforcing With Food All Together?
The short answer is never. Similar to how we need reinforcement via a paycheck when we go to work, our dogs also like to see a reward for performing behaviours. It doesn’t mean that you have to reward every behaviour every single time, but you should reward occasionally as reinforcement keeps the behaviour strong and reliable.
Remember: Reinforcement Builds Behaviour. If you want to lose a behaviour over time then stop reinforcing it.
A good example of this is what I see often happen with recalls. Many times when someone calls their dog back to them it is good for them but less good from the dog’s point of view. The freedom at the park is over, playing with other dogs is over. Coming can quickly become something that your dog doesn’t want to do. But… by using periodic reinforcement when we call the dog and reward him and then release him back to whatever he was doing before then your dog will believe that coming to you is a good thing and rewarding to do so.
One way of fading reinforcement is to chain behaviours. You might ask the dog to sit and then to spin and only mark and reinforce the second behaviour. Gradually you can build up the number of behaviours chained together in this way. For example, competitors in obedience trials or agility may need to perform over 30 plus individual behaviours without reinforcing with food. For more detailed information on this topic check out Janice’s post on Fading Reinforcement by Chaining Behaviours
Keep Going Signal
A keep going signal is a way of letting your dog know that he is doing the right thing but it is different from marking because we don’t offer a food reward. “Good” is often used as a keep-going signal.
In the chaining behaviours example above, we want a way to let the dog know that he is doing right by sitting even though we aren’t rewarding it with a treat. We could use the keep-going signal after the sit to let the dog know he was still doing the right thing and then ask for the spin and mark and reward that.
The Reinforcement Timeline
When you are training a behaviour you will go through the same general pattern for each exercise:
- Follow the Lure (to create the behaviour)
- Fade the Lure
- Add Cue
- Reduce Reinforcement
- Fade to Random Reinforcement & Praise
Build Slowly on Exercises