There are so many traps you can fall into when it comes to dog training. When it comes to clicker training, we sometimes jokingly refer to these traps as the “clicker disease “. One of those ” clicker diseases ” is this to reward a little too often. We want the dog to succeed so badly and it feels so nice to give the dog a reward – both the dog and we ourselves will feel good about it – that rewards can become commonplace. However, this is a disease that only appears after a couple of years as a clicker trainer – in the beginning, the vast majority rewards far too seldomly. Then, even after several years as a clicker trainer, you don’t get to be stingy with the rewards – but you must be aware of what you are rewarding.
Dare to withhold the reward when something goes wrong. Let your dog try again and get lots of reward when he gets it right!
In a reward-based training paradigm, a reward that doesn’t show up is almost as much information for the dog as the actual reward. The missing reward means “try again so you have a chance to get a reward”. When we don’t reward, we want to see that the dog at the next try uses the information and gets it right. If the dog makes several mistakes in a row, we have to make the training easier, but quite often we can actually be a bit “tough”. If the dog has been successful repeatedly at a certain level and has been rewarded for that and then makes a mistake, it’s not a given that the best strategy is to make the next attempt easier. Instead, the best thing could be to not reward and then make another attempt at the same level of difficulty. If the dog gets it right that time, we’ll naturally reward heavily!
We had a really clear example at a course we held in England a while back. A flat coated retriever bitch had a tendency to spit out the object 11-15 inches in front of her trainer’s hand or to spit it out as soon as the hand showed up. On the first day, the trainer worked on backing up quickly and then helping her dog by slowing down and presenting her hand – and voila, the object ended up in the hand. After a while the dog caught up on that that was what led to the reward. If the dog let go of the object on the ground, the trainer would steal it and try again. On the second day the trainer could put the object on the ground a few yards away, ask the dog to get it and then encourage the dog to come to her and deliver it. After a couple of repetitions, we wanted to take it further, so that the dog could do the same thing even when the trainer was silent. The trainer placed the object just as before but didn’t say anything when the dog picked up the object. Then, all of a sudden, the dog let go off the object 11 inches from the trainer’s hand again. Instead of making it easier (by going back to encouraging the dog when grabbing), we placed out the object in exactly the same way and made a new attempt with the trainer completely silent. This time, the dog grabbed the object and delivered it swiftly into her trainer’s hand. Bingo! The missing reward had really informed the dog what paid off and what didn’t.
Let the dog do the job
On the theme of daring not to reward, there is also the problem when the trainer works harder than the dog. Some dogs learn to “exploit” this more than others. It’s not always a problem, but sometimes it is and then the problem needs to be brought to attention and solved.
I’ve encountered a couple of examples of this during the year, all really quite easy to solve. In this particular case, it had to do with trainers who bent down to receive the dog when he came to deliver the object. This had the effect that the dog always delivered at the height of the trainer’s knees or ankles. Practical for the dog, not so much for the trainer… Both dogs in this example were quite large (a curly and a Landseer) and they really have a practical height for deliveries where the trainer shouldn’t have to bend over.
What we did was try what would happen when the trainer didn’t bend over to take the object. The first time, both dogs released at the trainer’s feet and looked very surprised! “But, you usually take it right here?” The second time, the Landseer had such a look on his face: “Ok, I’ll put it in your hand then” – and then his trainer didn’t have to bend over anymore. She just had to remember to stand up straight and not bend over.