What’s In Your Reward Toolbox?January 16, 2019 2020-05-13 9:21
What’s In Your Reward Toolbox?
What’s In Your Reward Toolbox?
I love talking about rewards. I love exploring rewards. Finding the right reinforcers for one’s dog is one of the aspects that makes dog training so highly individual and specific to each single dog.
Rewards make our dogs repeat behaviors. Great rewards make them repeat behaviors with a higher chance and better accuracy than rewards that they only like so-so. When first starting out, owners usually buy a bag of generic treats and hope that their dog will as hard as he can for it. Often they are quickly disappointed that the dog seems to enjoy sniffing, barking and socializing with other dogs much more than his cookies. This is not only due to just starting out in training – it is also due to mediocre rewards. The owners are in a difficult situation – they have an untrained dog, and nothing that the dog will make a true effort to work for. It is easy to mistake the dog’s lack of enthusiasm for the rewards as a lack of enthusiasm for training itself: Without a lot of dog training experience, it can be tricky to understand and appreciate just how much it is the rewards that rewards make or break the training.
For this reason, I like to do a reward buffet in my beginner obedience and trick classes in the first lesson. A bunch of plastic cups, each one containing pieces of treats that are traditionally high on a dog’s list of food rewards: hotdogs, cooked chicken, roast beef, freeze-dried liver etc. I encourage everyone to take samples of everything and try them out during the different exercises we do in the class. It illustrates quite effectively how different the effort might be that a dog puts into learning, depending on what’s in it for him.
By the end of the lesson most owners have a much better idea of their dog’s high value food rewards and can hit the ground running in training with a bunch of tools in their reward toolbox.
Reward toolbox, I like this term for many reasons. On the one hand it emphasizes that there is more than just one reinforcer for any dog, on the other hand it illustrates that we can pick and choose our rewards based on the job on hand- they literally are tools that are suited for specific functions, each one with their own advantages and disadvantages.
Toys for example are great for rewarding at a distance and for anything that needs dynamic motion. It can be a lot easier to lure easy motion-tricks like spins or leg weaves with a toy rather than a treat. They tend to engage dogs better around distractions as well. Toys on the other hand are not my first choice if I need precision. The inherent dynamic and speed that toys add to the behavior they are rewarding can take away from precision, especially early in the training process. Give it a try – take a trick of medium difficulty, such as hind leg lifts, and ask your dog to do it with a treat, then with a toy.
What about the differences between toys? Toys that you can throw very well (such as balls rather than tug toys) are great to reward you dog at a distance and close to a potential distraction. The rush of chasing can overwrite their interest in something else. Here is an example:
When I first got my Border Collie Fusion, he was very bad about chasing wildlife. Luckily we had ditches close to our house, with a pretty much 100% chance to see ducks there on any given day. The ducks were perfect dog training helpers in that they were very predictable not only in that they would be there, but that they would also stick to the same location +/- 50ft. It allowed me to set up my training situations perfectly and with no surprises. We would approach the ducks at a large distance and just when Fusion noticed them, I would start to play with him. The next day we could approach them a little closer. I made sure that whatever I offered him as an option for not chasing the ducks was really good – for him, this was mostly thrown tennis balls. Eventually we progressed to a point where I could throw the tennis balls into the ditch right next to the ducks, and he would swim to get it without batting an eye (the ducks also didn’t – I guess I desensitized the ducks in the process as well).
But even treats have different options for being rewarding. I like to use lick-able rewards (such as peanut butter or spray cheese) for dogs that are anxious or nervous. The licking itself calms them down, and they can experience a more sustained reward than if one simply fed them a treat. It takes rewarding from discrete events (one treat – another treat – another treat) to a continuous experience. This can also be helpful in counterconditioning/desensitization as it allows the dog to be fed truly the entire time that a trigger may be present.
Build your reward toolbox based on the dog you own. The reward toolboxes of my three dogs are all different, and what works well for one doesn’t get the other excited at all. Especially as they are puppies you can shape their reward preferences a bit though.
When my dog Kix was a puppy, she was very wild, very demanding and generally a bit of a challenge for poor Fusion. Fusion loves sticks – he will pick them up on our walks in the forest and carry them around with him for a long time, only dropping them to exchange for an even better stick. Baby Kix – of course – always wanted exactly the sticks that Fusion had picked and he would give them up to her, looking defeated.
In order to let him have his sticks and still satisfy Kix’s need to get her own nature toys, I introduced her to pine cones. Fusion has never shown any interest in pine cones, so this seemed a good option. I rolled them across the ground, threw them up in the air and generally made them very appealing to Kix. Within a week she would go and get her own pine cones when I told her to get a toy, and completely stopped taking Fusion’s sticks. They were now proudly carrying their specific toys on walks next to each other without any interest in the other one’s treasure.
(Here is another story about puppy Kixi and her unusual reward interests)
What is in your reward toolbox?