Teaching a dog to play fetch

Question

I have a six month old Rhodesian Ridgeback who I have not been able to cultivate an interest in playing fetch. I take him outside and he is more interested in walking around and chewing on sticks than he is on playing any type of game.

If I throw something for him he will often get excited and run after it but then just ignore it once he gets there. He rarely ever decides to pick up the toy let alone run back to me with it. As such I never get a consistent opportunity to shape the game through clicker training, and oddly enough he doesn't seem interested in earning clicks/treats when we are out playing. Even to the point where he did exactly what I hoped for, fetched the toy, brought it back, I clicked and he didn't even take the treat he just wandered off sniffing the ground and gnawing on tree stumps.

He has responded well to clicker training in every other area, so the clicker is well primed, and he has started to learn the click/treat "game", but I have no idea why it is seeming to fail in this situation. Can anyone provide suggestions for how to get my dog to play fetch?

Answer 1

Teaching a retrieve is a great example of when you can use backchaining to get your desired effect. Essentially, when backchaining a sequence (in this case, go get something and bring it back), you start with the last thing in the sequence and add a ton of value into that, then slowly add in the steps backwards. Because you've built so much value into the last steps, the dog is always looking forward to them and so the earlier steps are reinforced by the later steps.

Here's a plan that we've been going though lately to teach our dog, Nova, to bring things back. We had a lot of work to do because she'd always run away with toys and wasn't into tugging much, but she's improved immensely and is finally starting to retrieve automatically and with enthusiasm.

Prequisites

If you have a dog that loves to tug, you have a huge advantage. If you don't, you may want to spend some time building value into playing tug (here are some reasons to play tug), but you can use food as a reward too.

Sample Steps

The first step is to get your dog to bring the toy to you from almost no distance. Just drop the toy, back up a bit (start off with one step or less), and encourage your dog to come to you. If he does, reinforce that decision (click and treat, just treat, play tug, etc). If that's too much right now and your dog runs off, start in a small space (bathroom, closet, corner) where the dog can't run off and self-reinforce. You can even do this sitting and simply ask for a hand-touch.

As your dog gets more confident, slowly increase the distance after you drop the toy. Allow the dog more freedom to make bad choices, while reinforcing the good ones. Eventually, you can play games such as running all around the yard or house with your dog chasing you with a toy. If he drops the toy, end the game and try again. When he catches up to you, have a party and play tug / give treats. Try doing this with different toys and things.

Eventually you'll have a dog who loves bringing you things. Now it's time to continue backchaining. Start with dropping the toy on the ground and letting your dog pick it up to play the chase game. As you progress, increase the distance you're throwing the toy and pretty soon you'll have a dog who runs out with enthusiasm, and immediately runs back with the toy. Remember to only reward average or better responses and not to compare your dog's "average" to others.

Further Reading

Denise Fenzi Series series on behavior chains.

Answer 2

Although I'm 100% in favour of clicker training, in this case it might not be the best approach to start with.

What you want to do is to make the dog enjoy and initiate playing fetch. To me this is closer to classical conditioning than to operant conditioning.

Clicker training, and positive reinforcement training is about operant conditioning, you condition you dog, through positive reinforcement, but the dog has to think his way to get the reward. Subsequently he "figures out" what actions are "good".

Classical conditioning is about innate responses to stimuli. It is in a sense more "primitive" and bypasses any kind of reasoning the dog might have. When you take its food bowl he just gets hungry, there is no reasoning involved. This is opposite to the operant conditioning where you make the dog think, eg. for all the basic cues (sit, down, etc.).

In a way you want to play fetch with your dog so that he enjoys it. This is not to say that dogs don't enjoy what you train them to do, but here we want a very "primitive" reaction, you take the toy and the dog, without thinking starts playing.

In addition it seems that you tried to clicker-train him in an environment with a lot of distractions, that might be the reason why he didn't respond to your clicker. Whatever the real reason, if it doesn't work, that means you have to "go back a step", either in term of difficulty, distraction or attention (too long training session for example).

So how do I answer your question with this ?

You should classical condition your dog toy = fun = excitement. You mentioned that he wandered off sniffing. Take him on leash, try to see where he want to go sniffing, unleash him, and as you let him go, throw the toy. Doesn't matter if he goes to the toy at this point. Ask squirrels for help, once the dog sees them and is ready to dash to them, throw the toy. Do the same we he gets out of the car at the park. You can do the same association with food. First "throwing the toy = food", then "throwing the toy, the dog runs = food", etc. You got the idea.

If you do this, maybe 10 times, you'll have the dog conditioned and excited as you throw the toy. This will then be a perfect starting point to move on the operant conditioning, as you did with the clicker to a real "play fetch".

Try also to set up clicker training sessions (for whatever behaviour) in more distractive environments. To teach him "play fetch", or more generally for outdoor training sessions, try to use the distractions and the environment at your advantage. Use it as a reward. Fetching the dog, OK you can go sniff on your own, etc.

Always make sure that the dog can succeed. For the first part (classical conditioning) that means throwing the toy only if the dog is somehow excited. Not at the end of a long walk. For the clicker training part, don't do it if the distractions are way to high.

Answer 3

The answer on backward chaining was very good and I would like to suggest an augmentation to that approach. The OP stated that the dog is very interested in picking up sticks. Thus I would clear the yard of stray sticks and then take one stick and use that rather than the ball as the toy that the dog is to fetch (falling the gradual backward chaining approach). I would carry this on through eventually throwing the stick. At that point you could start again at the beginning of the process using a ball or some other too; don't worry the second time through will go MUCH faster than the first and the transfer to ball play will be almost immediate.

Answer 4

Try different items. Your dog may be bored with the same old ball. Try different toys or balls for fetch. Give better treats. My dog refused to do anything for cheap milk bones or any other wheat, corn or soy based treats. I then got a snack plastic bags to put in my training pouch I clip on my pants, and filled it with real tiny cut up meat, and my dog came to attention, and was willing to learn in order to earn a valuable great tasting treat. I cut up tiny bits of meat, put in tiny snack size ziploc bags and freeze. Then when training time, take out a frozen bag, and the treats are cold. My dog also loved apple slices, baby fresh carrots, cooked sweet potato slices, etc too. The cheap big commercial dog treats don't taste very good, so dogs don't really like them compared to real food. When you have great tasting treats your dog will be more attentive.