Teaching the Retrieve
In WTM issues May and June last year, I wrote about ways of encouraging the dogs’ natural retrieving instincts. The retrieve instinct is so important for a trials dog even when he is not retrieving – it is the desire to get the ball, article, or toy that motivates the dog to track. I know we can teach the exercises using food, but we can usually get more drive, more ‘oomph’ by using a ball or toy. I always say that good working dogs prefer a ball over food and often have to be taught to eat when out as they are too busy to stop to eat! I think that food is better for getting accuracy and games are better for speed and drive.
When teaching the dog a formal retrieve for competition I use food to train him. With food I can work on refining different aspects of the retrieve. If I were to use the retrieve game itself how can I point out the bits I am not happy with such as mouthing, dropping, or not releasing? Once the dog has possession of the ball (or dumbbell, or whatever) he is self-rewarding and I am not important to him. I guess there are many ways in which I could reprimand him for his misdemeanours, but that could either dampen his enthusiasm or reduce cooperation, and this could make his reluctance to return and to release the article even worse. The chances are I would damage our relationship and cause confusion over my requirements for the retrieve. In order to avoid these potential pitfalls my dog has a play retrieve in which he can mouth or drop the ball/article if he wants without criticism from me. I work on improving this exercise over a period of months simply by encouraging the dog to place the ball/article in my hand for me to throw it again, by pretending not to encourage him! If he wants to keep it that’s fine – but if he wants me to throw it he will have to make a point of giving it to me. If we are not ‘having words’ over this exercise the dog will relax and work out how to get the best out of us – and this is usually just what we want from them.
Using food to teach a retrieve
You may be familiar with the ‘paradox of reward’. The paradox of reward tells us that ‘if you reward a self-rewarding behaviour, you reduce the frequency.’ Say, for example, you enjoy ironing (I actually have a friend who does! – sad but as they say; ‘there’s none so weird as folk’) if you start getting paid for ironing, the natural pleasure will decrease as the payment becomes more important. You could end up only doing the paid ironing and walking around in creased clothes because you cannot be bothered to do your own ‘unpaid’ ironing.
Shock horror, am I saying that if I reward my dog with food for fetching something to me, that he will be less likely to want to fetch things in future? Well yes actually, but this would be an extreme case and with sensible training it is unlikely that you would ever get to that stage with a dog that had a reasonable retrieve instinct in the first place. With a dog that had no retrieve drive anyway, at least you would get an accurate, reliable retrieve if you train it properly. To maintain balance then, I would use food to train my formal retrieve but also continue my play retrieves with no similarity between the two. So the command for my play retrieve would be "Fetch" and the command for my formal retrieve would be "Hold". If he messes around in the play retrieve I just ignore it, after it is a ‘play retrieve’ so I just let him get on with it and walk off.
To train a formal retrieve I start with an article that is easy to pick up and a pot of ready prepared tasty titbits and a clicker. The article might be an old walking sock folded into itself and stitched to prevent it unravelling, or the centre of a roll of toilet paper stuffed with rolled newspaper to stop it collapsing, or a suitable soft toy without dangerous eyes, or limbs and a tail that might encourage the dog to pick it up with his teeth by an addendum rather than developing a nice opened mouth pick-up.
I am not going to labour the points of clicker training as I have already covered those (WTM March 2012), but essentially I would show the dog the pot of food and ask if he wants it. When I have his attention on the pot of food, I would then draw his focus to the article and click as soon as he picks it up. He should immediately drop the article to claim his reward from the pot. If he does not drop the article it could be that he is self-rewarding (playing with the article) and therefore you are losing control. To overcome this problem, make no attempt to grab him or the article, but instead tease him with the pot of food right under his nose, lower the pot towards the floor and as his head lowers and his mouth opens to get the food, the article will fall out. Allow him to have some of the food and then take a larger piece from the pot, teasing him with it, throw it across the floor. You are now replacing the chase of the article with a chase for the food. The food should become more fun than the article. Remember it is easier to control the food than the article (once he has swallowed it, it is gone and you still have the rest of the pot of food in your hand). The article should remain boringly lifeless on the floor near you throughout. After he has eaten the thrown treat, draw his attention back to the article and click as soon as he picks it up and repeat as above.
If your dog is a ‘foody’ type you may have difficulty in getting him to move away from the food to pick up the article in the first place. Ensure he is; a) charged on a clicker, and; b) knows to move away from the reward to a hand target in order to earn the click and reward. Now all you have to do is to make the reward less inviting or less obvious (maybe behind your back) and the article more exciting by animating it and playing cat and mouse. Remember to look at the article not your dog, or he will get locked into you and miss the point altogether. Click the instant your dog pounces to grab the article and proceed as above.
You are essentially playing a balancing game: if the dog wants the food – make the article more fun: if he wants the article – make the food more tempting.
By keeping the desire to earn a click and get the food greater than the desire to retrieve, you will be able to keep control of the exercise. If you click as soon as the dog picks the article up he will not have time to mouth it. If you repeat this dozens of times the dog will grow to like the game and you can begin to ask for a little more. Aim to get the dog to lift the article up towards the pot. Using the pot of food as a focus and feeding the dog from the pot, close to the pot will encourage the dog to turn towards the pot as soon as he has the article in his mouth. (He does this in anticipation of gaining the treat from the pot.) You are now shaping the return with the article. To move this on to the next stage, after clicking allow the dog to drop the article feed at the pot and then throw a piece of food away (as above) and, leaving the article on the floor where it landed follow your dog towards the treat you have just thrown (just a couple of paces to start) and turn back to face the article. The dog will now have to go past you to pick up the article and turn back to you with the article. This is now beginning to look a little bit like a retrieve except that you are throwing the food for the dog to chase, not the article: the article stays where the dog has dropped it every time. As you both gain experience with this exercise and grow to enjoy it you will soon be in a position that you will be able to ‘punish’ the dog for mouthing or dropping without causing stress. If your dog really enjoys the exercise (and if taught as described above he will do) the punishment for the misdemeanour would be to discontinue the exercise. So the instant you dog mouths mark the moment – "oops!" and lift the food pot up to signal ‘game over’ and pick up the article. Do not look at your dog, give him ‘time out’ for half a minute or so and then invite him to start again. If your timing is good, he will soon work out what the problem is and correct himself if he wants the fun to continue.
Hand delivery of small articles and the present of the dumbbell are taught separately so as not to cause confusion in the early learning. For example, if the dog did a lovely pick-up and return with the article but started to mouth when sitting, or sat crookedly would we criticise the dog for the last bit? Doing so would of course reduce the chances of him being keen to repeat the exercise. However if he knows both bits and enjoys doing them as individual exercises, it will be easier to pull the two together at a later stage when the dog is more familiar with his job.
Allowing the dog to work out for himself what it is that you want, simply by verbally marking the instant the problem occurs and taking a short break in the procedure is sufficient for most dogs to learn how to get it right. (Please note: a verbal marker is not a verbal correction!) If your dog really enjoys his work he will want to please you and he will want the training to continue. Remember, after every session to ask yourself, "What was it like to be trained by me today? Did we have fun?" The chances are if you thought about your training and you both had fun, then probably either you or your dog will have learnt something.