A Guide to Training with Treats

by Trish King

Having a happy relationship with your dog means learning how to work together. If there´s a topic you´d like to see Trish cover, email editors@fetchthepaper.com.

Why do people not want to use food in training? Some feel it’s somehow morally wrong, that the dog should want to work for us and do it for love. Some think that in the history of training dogs, food wasn’t used. Of course that’s not really true. Dogs have always worked for food, sometimes directly, sometimes indirectly. Our dogs get all the love they need for free. The real problem is not whether one uses food but how one uses it.

Consider that dogs are, at maturity, the mental equivalent of a 2 to 2 1/2 year-old child. Some are mentally a bit older. Some are younger. Their motivations are relatively simple. They want food, comfort, company, and they will seek that. They will avoid anything that is distasteful, frightening, or painful. They are also very selfish. Sure they love you, but they need you (and they should know it!), and they want stuff from you, all the stuff they require to live. They are also focused on what they want and what they don’t want, which can make training frustrating. Try training a dog who is focused on a squirrel. Very frustrating!

Food is what we call a primary reinforcer, that is, the dog needs food to survive. Therefore the dog is likely to put himself out to make sure he stays alive. Toys and balls are not primary, though they can also be potent motivators when used properly. Correction (punishment) is usually not the best way to get a dog to do something, though it has its uses for reducing certain behaviors. At any rate most dogs will work for food quite eagerly. Here is a guide:

Use food to:

• Train behaviors you want to see repeated
• Train alternates to undesired behavior (e.g., training your dog to sit instead of jump up)
• Train against the genetic grain (when you want your dog to do other than what his instinct says)

Food not necessary:

• To extinguish unwanted behaviors, like mouthing, chewing. jumping up
• When the environmental reinforcement is more attractive than food (e.g., just going through the door after a “wait” can be enough reinforcement for your dog)

Oh, is that what you want?

There are two stages of training: acquisition and maintenance. During acquisition, a dog is beginning to understand what you want. If you use inducements to help him he will learn faster. If you use punishment he might do the behavior but he will become cautious around you, and you may get “behavioral fallout.” Though your dog is doing what you want, his attitude may be depressed, or he may lash out at another dog or person.

During acquisition, you should reinforce your dog every time you get the desired behavior. If there are several components to the exercise, train just a bit at a time. The dog needs to be sure of what you want. We often think dogs know what we mean because they look so interested and alert as though they understand everything we say. They don’t.

Much of the time dogs just guess, and sometimes they guess right, sometimes wrong. They’re in a foreign country where they don’t speak the language. Sometimes they luck out but they don’t quite know how they did it.

A dog learns he has to do something when a certain “picture” is in front of him. Trish is standing, treat in hand, saying “sit.” If Trish doesn’t have the treat, the picture isn’t complete, and the dog won’t do the behavior. Just because you use a treat every time, however, doesn’t mean the dog has to see it before the behavior. And that’s where most people screw up. So, what to do? You should switch from the obvious treat to the hidden treat very quickly indeed, like within 15 to 20 repetitions if possible.

This seems really easy. Why don’t people use it? There are a few reasons. Mostly we’re impatient. Dog sits, gets treat. Dog sits, gets treat. Trish hides treat, dog doesn’t sit, Trish gets impatient, brings out treat, picture complete, Dog sits, gets treat. Success! Trish whines “dog will only do sit when he sees a treat.”

So, rule number one: hide the ever-loving treat! Quick! Rule number two: don’t give the treat if the dog didn’t earn it.

Now your dog has the behavior. Sit perfectly, treat hidden. Still this is frustrating. Sit, treat, sit, treat, sit, treat. Where do we go from here? Two places.

1) Twofers. Ask for a sit. Dog sits. Quickly ask for another sit. Reward. You just got two behaviors for one reward. This is called variable or random reinforcement. You are asking for more behaviors per reward as the dog gets better.

Imagine learning to type. In the beginning your teacher congratulates you when you type a rudimentary sentence. However, as you get better his congratulations come only when you accomplish more challenging tasks. That’s what we’re doing here. Now, people often want to stop giving treats altogether. Let’s go back to our example. Say you’re getting paid minimum wage while learning to type, and you’ve gotten really good. Your employer says, “Trish you’re darn good at that typing stuff. You’re so good that we’re going to stop paying you!” Don’t know about you, but I’d be gone! Never stop paying; just change the pay.

2) Adding other behaviors. Now your dog knows “sit” and “down” really well. If you say “sit” she sits; if you say “down” she downs. Combine them. “Sit, down” (treat), “sit, down” (treat). And every once in a while “sit” (treat) “down” (treat). You can go on with this ad nauseum. “Sit, down, come, wait, roll over” (treat).

Why don’t people get to this point? One reason is guilt. Guilt is not a good motivator for dogs! It is for people, however. When they look at us as if they are starving we feel obligated to dispense food. Honestly, it’s not necessary.

Perfection is not possible. 100% recall is not possible. 100% sits are not possible. Even those people who brag that their dogs are totally reliable are wrong. You want something that’s 100 % reliable, get a pet rock. Your dog will not be perfect until you are. And none of us ever will be. If in doubt about whether you are expecting too much from your dog ask yourself the following questions:

Is she capable of doing it physically?

Have I trained it until the dog is 90% reliable? If not, train some more.

Is the environment one in which I have trained before? If not, train in many places.