Features September 1999 Issue

Teaching Fetch Using Positive Reinforcement Techniques

Not all dogs are naturals at fetching. Here’s how to teach a reluctant retriever to fetch using positive reinforcement techniques only.

Who has not watched in awe as a Border Collie at a local park sails through the air, snatches a FrisbeeTM in mid-flight and dashes back to her owner, dropping the Frisbee and waiting in eager anticipation for the next throw? Playing fetch with your dog is fun. It’s also a great way to strengthen the dog/human bond, satisfy your dog’s prey/chase instincts, and provide enough exercise to work off that excess energy that can make him a challenge to live with. A formal retrieve is also required for upper levels of obedience competition.

Some dogs are natural retrievers. Teaching them to fetch is a matter of directing the behavior into the right channels. Other dogs are not, and while teaching them to retrieve may look like an insurmountable challenge, it’s not as difficult as it seems. There are limitations, of course. Your 150-pound Newfoundland may never sail through the air like a Border Collie, but she can certainly learn how to fetch.

The old way

Fetching provides great exercise for dogs,
but the game gets tedious if the dog doesn’t
know how to "play" properly.

At one time in the not too distant past, the dog training world almost universally agreed that dogs had to be taught a “forced retrieve.” If you wanted a reliable retrieve, dogs had to know that they would be punished if they refused to pick up the designated object and bring it back. Years ago, my terrier mix, Josie, was the unfortunate victim of this training philosophy.

Against my better judgment, convinced that my trainer knew best, I taught Josie to retrieve using the traditional coercive “ear pinch.” We were preparing for the Open Class show ring exercises, Retrieve On The Flat and Retrieve Over High Jump. My trainer was a top ranked, nationally-known obedience trainer and competitor. I admired and respected her. I was just a lowly dog owner – what did I know?

Ignoring my uneasiness, I dutifully folded her ear flap over the choke chain, said “Take It!,” and pinched. When she open her mouth to yelp in protest, I popped the dumbbell into her mouth. Voila! She was learning to retrieve.

It worked. We flew through the Open Class with ease and earned our CDX (Companion Dog Excellent) title in three shows, with a high score of 197.5 out of a possible 200 points.

Meanwhile, we started training for Utility, where we would have to do the Scent Discrimination exercise. In Scent Discrimination, the dog doesn’t just retrieve a dumbbell, he must distinguish the one that has his owner’s scent on it from several lying together on the ground, and retrieve only that one. The exercise is done twice – once with leather articles, once with metal. The leather dumbbells were no problem for Josie, but she hated the metal ones.

Lots of dogs don’t like to hold metal in their mouths. Teeth scraping on metal must give them a “fingernails-on-the-blackboard” sensation. There are tricks trainers use, such as spraying the dumbbell with a clear plastic coating. We tried all the tricks. Josie still wasn’t buying.

“Pinch harder,” my trainer encouraged. “You have to make her do it.”

Josie and I practiced hard. The Directed Jumping and Directed Retrieve were easy for her. The Signal Exercise was a snap. But when I brought out the Scent Discrimination articles the light faded from her eyes and she gave me pleading looks, begging me not to make her do them. I persisted – until one day when I brought out the articles Josie hid under the deck and wouldn’t come out.

Finally, I realized how wrong the ear pinch was. I put away the articles and never brought them out again. If training meant destroying the relationship between me and my dog, I was no longer interested.

Negative reinforcement
Many trainers still subscribe to coercive methods for teaching the retrieve. The ear pinch is a widely-used, force-based method of training that utilizes a concept known as “negative reinforcement.”

Negative reinforcement means that the dog’s behavior makes a bad thing go away. We pinch the ear (bad thing), the dog takes the dumbbell, the ear pinch goes away. The dog learns that if he doesn’t retrieve, he will be hurt. He chooses to retrieve. As with many force-based training methods, it works with a lot of dogs, a lot of the time. It worked with Josie until we encountered the metal scent articles.

Unfortunately, there is a very real potential for negative side effects when we use physical force to train; side effects that can permanently damage the relationship. The dog learns to associate your hands with pain. He may lose his enthusiasm for training. Worse, he may lose his trust in you. Although traditional trainers like to believe that a forced retrieve teaches the dog that he has to fetch even if he doesn’t want to, in reality the dog can always choose not to retrieve and risk the consequences, like Josie did when she hid under the deck. The dog always has a choice.

The ear pinch is not the only coercive method used to train the retrieve. Blanche Saunders (now deceased), a highly respected obedience trainer in the 1950s and 60s, teaches a forced retrieve in her book, The Complete Book of Dog Obedience. Her method of teaching the dog to hold the dumbbell is inarguably harsh:

“Every time your dog drops the article, hold him tight while you cuff him across the nose. Say ‘Phooey!’ in a displeased tone of voice . . . Each time he drops it, the correction becomes more severe.”

I’m happy to say I never cuffed Josie across the nose.

Positive reinforcement: The new wave
Whether you just want your dog to bring back the tennis ball you throw for him in the back yard or you have your eye on advanced obedience titles and Frisbee championships, there is an effective way to train your dog to fetch, using the concept of positive reinforcement (in which the dog’s behavior makes a good thing happen).

In positive reinforcement training, you get the dog to offer the behavior you want without using force, and then reward him for it. All living beings repeat behaviors that they find rewarding.

When you want to train a simple behavior, like sit, it’s easy to see how it works. You hold a treat over the dog’s head, he sits, you use your reward marker (such as a Click!, or the word Yes!) to let him know that he just did a rewardable behavior, then you feed him a treat. By clicking and rewarding the sit repeatedly, you get the dog to offer sits more and more frequently and reliably. Once you are sure he will sit, you add the verbal cue, “Sit!” so he offers the behavior when you ask for it, not just when you move the treat over his head.

"Bring it back to me!" In parks full of distract-
ions, you may have to make it clear to your
returning but wavering dog that a treat is im-
minent upon his return.

Over time, you reduce the use of treats through a process known as random (or variable) reinforcement, clicking and rewarding some sits, but not all. At this point, if you consistently Click! and reward only fast sits, for instance, you can “shape” the dog to sit more quickly when asked.

The retrieve is a complex behavior. While some dogs are natural retrievers who pick things up easily and willingly, others are not, and have to be encouraged to take things in their mouths. Even natural retrievers may learn behaviors like “keep-away,” that interfere with a good game of fetch. How much effort you need to put into training the retrieve depends on your dog’s natural inclinations as well as your training goals. A formal show ring retrieve is considerably more complex than simply asking your Lab to drop his tennis ball at your feet so you can throw it again.

Breaking tasks into steps
Whenever you want to train a complex behavior, you need to visualize the final product (in this case, the retrieve), and break it down into small steps. For the back yard fetch you want to throw the ball, Frisbee or toy and have your dog run after it, pick it up, bring it to you and give it back. Broken into small steps it would look like this:

1. Wait politely until I throw the ball
2. Run after it when I throw it
3. Pick it up
4. Hold it in your mouth
5. Bring it back to me
6. Drop it when you get here

Let’s look at how you would train this simple “back yard fetch.”

Although we are going to examine the steps of the retrieve in order, you don’t have to train them in order. Once your dog knows each of the steps you can put them together in the right order to make “Fetch” happen.

• Wait For Me To Throw
Dogs who are excited about retrieving are often obsessed with their Frisbee or tennis ball, sometimes to the point of being dangerous. You can lose a finger if Skippy tries to grab the ball from your hand as you get ready to throw. We can use negative punishment to teach Skippy to stop jumping. Although we associate the word “punishment” with harsh corrections, negative punishment is not harsh or physical at all. It simply means “the dog’s behavior makes a good thing go away.” Positive trainers frequently use negative punishment because it doesn’t involve the use of physical force.

Ask the dog to sit, and hold the ball up to throw. If he leaps for it, whisk it behind your back and wait for him to sit again. Every time he sits, the ball appears. Every time he leaps at you, the ball vanishes. The first time he stays sitting when you bring out the ball, say “Yes!” and quickly throw it. (This will happen much sooner than you think – it often takes less than five minutes.) This part is positive reinforcement: Skippy’s behavior (sitting) makes a good thing happen (you throw the ball). In this case you don’t need a food treat. Skippy gets a “life reward” – he gets to chase the ball, which is even better than food! From this moment on, Skippy never gets to chase the ball if he jumps up; only if he sits. Once he figures this out, he’ll sit his little heart out to try to get you to throw!

• Run After It When I Throw
Lots of dogs will chase something that is moving but won’t pick it up. That’s OK – the pick-up is a separate behavior. Choose a toy that your dog really likes, play with it with the dog until he gets excited, then toss it a short distance. If he runs after it say “Go!” and when he gets to it Click! or say “Yes!” and feed him a treat. He may even pick it up. If he does, be sure to Click! and reward. At first he may only go part way toward it. That’s OK too. Just be sure you Click! while he is headed toward it, not after he turns around. Remember, the Click! marks the behavior you want him to repeat. If you Click! too late, you reinforce him for coming back to you, rather than for going toward the toy.

As he gets the idea, you can Click! only for increasingly closer runs to the object. If he does a short run, don’t do anything at all. Don’t say “No,” don’t Click!, and don’t say “Go!” again. Just wait. When he realizes he’s not going to get clicked he may head for the object again. (This is a very good time to Click! and reward.) If he doesn’t, calmly try again, tossing it a shorter distance this time. This may be a sign that you have raised the criteria too quickly and you need to take a step back. It is a common training mistake to try to move ahead too quickly. It seems logical that if Skippy will run after the object when you toss it five feet, he will do the same at 10 feet. But he might not. We might need to increase the distance by increments of one foot rather than five feet.

• The Pick-Up
This can be either the easiest or the hardest part of a retrieve. A natural retriever will do the pick-up in his sleep. In fact, most puppies naturally pick things up. If you constantly punish your baby dog for puppy pick-ups, you can squelch a budding natural retriever. Instead, if you put away all inappropriate items and consistently reward him with a Click! and a treat for picking up his toys, you will encourage his retrieving tendencies.

The goal of training should be to
keep the trusting relationship be-
tween dog and handler intact.

If your dog is not a natural retriever, don’t despair. Designate his most favorite toy as his fetch object. He only gets to play with it when you do the fetch game. Now set it on the ground. (Don’t throw it!) If he picks it up, Click! and reward. If he only sniffs it, Click! and reward. If he just glances in the object’s direction, Click! and reward.

In the beginning, reinforce the dog just for paying attention to the object. In any series of “attention” responses with the fetch toy, sometimes he will sniff or touch it, sometimes he’ll just look at it, and sometimes he will put his mouth on it – maybe even pick it up.

Once he understands the game, you can up the ante (this is called “raising the criteria”); you only Click! and treat if he touches it. Later, you Click! only if he actually puts his mouth on it, and finally, only if he picks it up. Once he is routinely picking up the toy, add your verbal cue of “Fetch!,” “Take It!,” “Get It,” or whatever you plan to use.

If at any time your dog “quits,” that is, he stops playing the game, you may have raised the criteria too quickly, or you may have trained for too long. Training sessions should generally be five to 15 minutes in length, several times a day. If you get two or three really good responses in a row, stop the session with lots of praise and a “Jackpot!” – a whole handful of treats. It’s always better to stop when you and your dog are having fun and winning, rather than when one or both of you are bored or frustrated.

• Hold It
The pick-up is only half the battle. Skippy has to hold it in his mouth if he’s going to bring it back to you. In any series of pick-ups, sometimes he will hold it longer than others. Once he is picking the toy up easily, gradually raise the criteria by clicking and rewarding for longer and longer holds.

“Gradually” is the key here. Your increments will be in fractions of seconds at first, and it is critically important that you Click! while the toy is still in your dog’s mouth! If you consistently Click! too late, after he has dropped the toy, you are rewarding him for dropping, not holding.

• Bring It Back To Me
Now it gets easier. As soon as Skippy is holding the toy for three to five seconds, back away from him when he is looking at you. (You can try calling him to you, but sometimes saying his name will make him drop the toy.) He should start moving toward you, hopefully with the toy still in his mouth. Click! and reward. He will probably drop the toy when you Click!, but that’s OK, as long as the Click! happens while the toy is still in his mouth.

Gradually raise the criteria so he comes closer to you before you Click!, and in short order he will be bringing it all the way.

• Drop It
You can practice this piece of the “Fetch!” any time Skippy has something in his mouth. Offer him a treat. When he opens his mouth to take the treat, say “Drop It!” or “Give!” in a happy tone of voice. (If you use an angry or intimidating tone he may hold tighter rather than drop.) Eventually he will “Drop!” on the verbal cue without the treat. Then you can Click! and treat after he drops, and by using random reinforcement, over time you can fade the use of the treat. (This is also a useful exercise for teaching him not to be protective or aggressive to you when he is playing with his toys.)

If he doesn’t want to trade the object for the treat, try dropping one or several treats on the floor, or use a tastier treat. Do this a lot with his own toys. You can then give the toy back (or toss it for him) as a reward also. He will learn that giving you the object keeps the game going. If you only do this with things he is not supposed to have, he will learn that when he drops an object he loses it forever, and he will become less and less willing to give things to you when you ask.

You can decide if having Skippy drop the object at your feet is acceptable, or if you want the object placed in your hand. Dropping at your feet is easier. Just let it fall when you offer him the treat. If you think he will try to grab it when you reach for it, keep him occupied nibbling the treat in your hand while you reach down and pick it up. Then let him have the treat. If you want him to place it in your hand, you will need to slip your hand under the object at first so it falls into your hand when he drops it. Later, you can insist he place it in your hand by ignoring it if it falls on the floor, until he picks it up and tries again.

Putting the pieces together
Now that Skippy knows all of the pieces, we can put them together. He sits and waits politely until you throw his ball. He runs after it when you throw, picks it up, holds it, brings it back to you and drops it when you ask. His tail is wagging, his eyes are bright, and he is eager for you to throw again. Yes, he has a choice to retrieve or not. He always has a choice. If you’ve trained well, he’s having fun and enjoying the game. What do you think his choice will be?

Today, Josie fetches a wide variety of objects with a wagging tail and a happy gleam in her eye. In recent years, when I reintroduced her to the retrieve using positive methods, I realized that she had never really been very happy about retrieving, despite her 197.5 scores. She used to dutifully retrieve the dumbbell under stress, in fear and anticipation of a correction. Now she joyfully chooses to fetch when asked, confident that she won’t be punished. We never went back to the metal scent articles; I’m not anxious to resume a show career. But I’m confident that we could, if we wanted to.

Also With This Article
Click here to view "Show Ring Retrieving."


-By Pat Miller

Pat Miller, a dog trainer from Salinas, California, is a regular contributor to WDJ. For information about her training classes, click here.

Comments (1)

Great post! I love the breakdown of tips by desired action. I also found this one useful, which also talks about using positive reinforcement to train your dog: http://www.ilovedogs.com/2013/04/using-positive-reinforcement-to-train-your-dog/

Posted by: marthailovedogs | April 24, 2013 1:38 PM    Report this comment

New to Whole Dog Journal? Register for Free!

Already Registered?
Log In