Quick overview of how and when to reward.
Dogs need to get paid. In daily life, payment might be a kind word or a skritch behind the ear. But agility is demanding and complicated, and the pay scale needs to be higher. We’ll use both food treats and play (retrieve and tugging) when teaching new agility skills.
In order to be effective, the reward has to come quickly: AS an action is occurring, not afterwards. A delay of even a few seconds can be too long. If you wait until the dog has finished performing the action before getting the treat to him, he probably has already gone on to think or do something else when the reward finally arrives. The dog must clearly understand what he is being rewarded for.
But it’s difficult to get the treat to the dog at exactly the right moment. We could throw it at the dog as he goes through or over an obstacle, but obviously that wouldn’t work!
Here’s where clickers come in handy. Clicker training is simply a way to communicate with dogs. Initially, the clicker means absolutely nothing to the dog. It’s just a mechanical sound. But by pairing the sound of the clicker with a treat, the dog quickly learns that when he hears the sound “click”, a treat always comes.
Technique. “Charging the clicker”. Just click/then treat, click/then treat twenty, thirty, forty times in a row. Don’t ask or expect the dog to do anything. All we’re doing is pairing the sound with the treat.
Once the dog looks expectantly for the treat the moment he hears the click, we’re ready to go on.
A Bridge
The sound bridges the time between the action and the reward. Timing is everything in dog training — the clicker/marker tells the dog that what he was doing the exact moment he heard the sound is the action being rewarded… and bridges the time between the mark and the reward. Rewards (and punishment) need to be given within seconds of the action for the dog to understand what he’s being rewarded for; the marker allows you to MARK the moment, then take your time giving the reward.
But we don’t always have the clicker with us. And, for me, I already have the dog, the leash, the treats, a toy… I find a clicker awkward to manage as well.
We can use a verbal marker in exactly the same way as a clicker. Choose a short, distinct, unique word, and pair that sound with a treat. I use YES, but you can choose anything that’s easy to say, but you don’t use regularly with your dog.
Again, “charge” the marker by saying YES then treat over and over again until the dog looks expectantly at you for the reward when he hears YES.
What the clicker does:
- The clicker freezes the action, allowing you to mark behaviour as it occurs.
- The clicker predicts or promises a reward. Thus, every time the dog hears the clicker sound, a reward must follow.
- A clicker ends an action. When you click something the dog has done, you have promised a reward, so anything the dog does between hearing the click and getting his reward is immaterial. Don’t click an action, then make him sit to get his reward.
Goal:
The dog’s actions CAUSE the reward, the reward doesn’t cause the action.
A dog who is clicker-literate knows that his actions are what causes the reward, and he will become an active, engaged participant in the learning process.
CLICKER BASICS
- Start with a hungry, interested dog.
- A reward ALWAYS follows the click or marker word.
- Mark during the desired behaviour, not after it has finished.
- Click ONCE. If you’re really excited about the dog’s progress, give more treats, not more clicks.
- Mark-and-treat every correct response (continuous reinforcement).
- Release between attempts. You can throw a treat to get the dog to move away, then bring him back to try again.
- Set up for success! Don’t make learning difficult. If your dog is not getting it, step back to something easier, where he will be successful. If he loses interest and wanders away, put him up for a bit before trying again. (This is where a crate is handy.) Wait until he is hungrier.
Agility — it’s not about the obstacles.
Really! Agility is not so much about the obstacles as it is about getting your dog around the course between the obstacles. If you can accurately communicate to your dog, who is running at top speed, where you want him to go, you can get him on the obstacles.
So the first skill needed is for the dog to understand about following your hand and going where you point him.
Space
With your treat held in front, move into your dog’s space. To keep from inadvertently pushing your dog away with one hand, hold the treat between both hands and plant your hands against your legs. Have your hands at the height of your dog’s point of chest. Lean into your dog. You don’t actually have to step into him, just rock forward. When your dog yields by moving back or away, even slightly, reward by rocking back, moving away from your dog and inviting him back into your space. Give the treat as you invite him back. Reward the slightest try your dog makes at this point. You just want him to get the idea that you can take space and give space.
Teach to follow your hand
Have him follow in a circle around you. Your arm is straight and pointing out; bring your hand down to your side when you reward. This is the beginning of the dog understanding to come in close when your arm is down, and get out to a distance when your arm is up.
Then, try going straight as your dog follows your hand. First thing to learn is for you and your dog to go in the same direction of travel. Hold your hand with the palm parallel to the floor, shoulders facing forward. This tells your dog that you’re going straight.
Turn and go the opposite direction. Switch hands as you switch sides.
Turn in to your dog, pick up with the opposite hand and change directions.
Have your dog sit on your right side… lure across your body and treat on the other side. Repeat the other direction. Don’t let your dog go behind you — step back into him if he does. Then wait… and treat when he chooses to be in front.
SHAPING BEHAVIOUR
Agility is a string of tricks — unlike herding for border collies, or flushing birds for spaniels, there’s nothing natural about agility. A dog might run over a log across a river, or run around trees, but executing the dog walk correctly or doing weave poles is nothing natural. So because we’re essentially training tricks, we need to use a high rate of rewards, especially in the teaching phase.
We will introduce many of the obstacles by “shaping” the behaviour. You’re probably used to helping your dog, showing him what you want. There’s nothing wrong with that, but if you always help your dog, he’ll get used to waiting for you to help.
Remember, we’re looking for an active dog, one who understands that it’s his actions that cause the reward and not the other way around.
Shaping allows the dog to learn by trial and error. If he tries something that is rewarded, he’ll try again. If he tries something that isn’t rewarded, he’ll keep looking until he gets the reward. Like the childhood game of “Hotter and Colder”, shaping teaches by rewarding closer and closer approximations of the finished behaviour. The exercise is broken down into steps, and each incremental step the dog takes towards the finished exercise is rewarded.
Technique: Shaping the wobble board. Start by first rewarding anything that might bring him into contact with the board. If he looks in the right direction, reward. Then, reward any movement towards the board. Then reward if he puts a foot on the board. Next wait for two feet. Then all four feet. As you progress, the actions you rewarded in the beginning are no longer rewarded: you wait for and reward closer and closer approximations of the finished behaviour. Breaking the behaviour down into TINY steps encourages the dog to keep trying and allows the learning to progress quickly.
Reward frequently. Sometimes you may need to reward anything that’s different; any new try, even if it’s not getting any closer to the board. In the beginning stages, we need the dog to be active and engaged. If he has to wait too long between rewards, he may get frustrated or bored and wander off to see what the other dogs are doing.
What to do if you dog visits the other dogs
If your dog is distracted and wants to go to another dog, let him. Stand still and just wait. The other dog and handler will ignore your dog. (As the class continues, this is an excellent chance to work on attention and focus.) Your dog won’t get rewarded by visiting, and at some point he’ll maybe turn back your direction. The moment he does, become animated. Tell him what a good dog he is, and if he continues in your direction, say YES or click and wait for him to come in for his reward.
If he continues ignoring you, follow after and scuff your feet. All we’re looking for here is for him to glance your direction. The moment he does, immediately yield by backing up and inviting the dog back into your space. Reward if he moves with you; scuff back in his direction if he turns away.
Homework. Click and treat your dog for getting on any obstacle you can think of: put his feet in a bucket, on a board, in a laundry basket. The more you work on shaping behaviours, the quicker your dog will understand that it’s his actions that cause the reward. You will also improve your timing.
One Comment
Great post!!