User:Lindsay Ridgeway/Reward-based Field Training for Retrievers/Principles

Principles

This section contains important information for training your field retriever. Besides being prerequisite to the chapters which follow, this chapter contains information that you may want to reread and refer back to time and again during life with your dog.

Discovery Training

In many canine performance sports, the dog is trained to perform behaviors that are not natural to the dog. Trainers use extrinsic rewards such as treats, and aversives such as leash pops, to develop a reinforcement history for the desired behavior.

But when a retriever is being trained in field work, a new principle comes into play. . .

Motivation

The single greatest responsibility of a field trainer is to build your dog's motivation for the game.

While motivation is always enjoyable to see in a performance dog, and may affect scoring in other sports, it plays the highest role in field sports, especially Field Trials. The reason for this is that a successful Field Trial dog is required to overcome difficult challenges, such as rough terrain, heavy cover, and long swims in cold water. These are not merely trained behaviors. The dog must above all want to retrieve in order to find the will to meet such challenges.

In behavioral science, the term motivation is distinguished from learning on the basis of whether the behavior is more or less permanently installed. Thus, a dog who performs a behavior on one day but not another may be more motivated on one day than another, while a dog who learns a behavior on one day will then repeat that behavior on subsequent days.

In this book, the term motivation takes on a more specific meaning. Here, we use motivation to describe the amplitude of the behavior, the enthusiasm and vigor with which the behavior is performed.

Many factors affect a dog's motivation:

Establishing Operations (EOs). Examples are whether the dog is hungry and well rested.
Frustration. A dog with a tolerance for frustration will show higher motivation when appropriate levels of frustration occur, such as when a tug toy is kept out of reach until the dog makes sufficient effort to reach it. Of course, if the frustration level is too high for the particular dog on that particular occasion, the dog simply gives up.
Reinforcement History. If a dog is reinforced for low levels of enthusiasm, he will learn to perform at those levels. One frequently sees this with trainers who "cheerlead". Their intent is to rev the dog up, but because the dog finds cheerleading reinforcing, in effect the trainer is saying, "That's the way I want you to perform these behaviors."
More factors to be listed

Building the Relationship

Many factors can contribute to relationship building. For example, in Susan Garrett's Ruff Love, relationship building is based primarily on the dog learning that all good things come from the trainer.

How you reinforce a behavior can also have a major impact on the relationship. Reinforcing with tug is excellent for relationship building, because it engages the dog's prey drive in a cooperative enterprise (see Tug).

Another way to develop your relationship with the dog is through cuddling (see Cuddling).

On the negative side, a lack of consistency in your responses to the dog's behavior, the unskilled use of aversives, and other training errors can detract from the relationship. Even if you are committed to reward-based training, you can see this happen at times, for example if you are feeling tired or impatient. Always watch the dog to see whether he is engaged and enthusiastic in your training, and when you see his behavior going the opposite direction, stop what you are doing. If you are unable to immediately adjust your attitude, it is far better that your dog spend the next few minutes or hours resting in his crate than that you continue with a counter-productive training session.

Rewards and Aversives, Positive and Negative, Reinforcement and Punishment

This book uses the terms reward and aversive, positive and negative, and reinforcement and punishment with fairly narrow definitions:

A possible reward is a stimulus (such as food or play) that is given after the dog has offered a behavior. If giving the stimulus results in the behavior becoming more likely thereafter, the stimulus is considered to have acted as a reward (also called a reinforcer).

A possible aversive is a stimulus that occurs after the dog has offered a behavior. If giving the stimulus results in the behavior beoming less likely thereafter, the stimulus is considered to have acted as an aversive (also called a punisher).

A stimulus can also be recognized as a reward if a behavior decreases when that behavior results in the dog losing the opportunity to obtain the stimulus.

A stimulus can also be recognized as an aversive if a behavior increases when that behavior results in the dog escaping or avoiding the stimulus.

Positive use of a stimulus means that the stimulus occuring is contingent on a particular behavior. For a reward, the result is an increase the likelihood of the behavior. For an aversive, the result is a decrease in the likelihood of the behavior.

Negative use of a stimulus means that the stimulus not occuring is contingent on a particular behavior. For a reward, the result is a decrease in the likelihood of the behavior. For an aversive, the result is an increase in the likelihood of a behavior.

A stimulus is considered reinforcement if and only if it results in a behavior maintaining or increasing its frequency. In other words, the litmus test is not whether the dog seems to like or dislike the stimulus. The only way to recognize reinforcement is whether the dog will work for the opportunity to obtain a positive reinforcer (a reward), or will work to escape or avoid a negative reinforcer (an aversive).

A stimulus is considered punishment if and only if it results in a behavior decreasing. Again, the litmus test is not whether the dog seems to like or dislike the stimulus. The only way to recognize punishment is whether the dog will work to avoid losing the opportunity to obtain a negative punisher (a reward), or will work to escape or avoid a positive punisher (an aversive).

Note that a reward can be used for both reinforcement (by giving the reward after a behavior) and for punishment (by removing the opportunity to obtain the reward after a behavior). Similarly, an aversive can be used both for punishment (by administering the aversive after the behavior) and for reinforcement (by enabling the dog to escape or avoid the aversive with a particular behavior).

Note also that the terms positive and negative apply to both reinforcement and punishment, and can be used with both rewards and aversives. Since this book is based on methods that use rewards for both positive reinforcement and negative punishment, that is the reason this book uses the term "reward-based" rather than "positive" in the title.

The Specialized Retriever

Does a field champion need "roll over" or "down"?

The Training Schedule

We Americans typically live our 24 hour days with a continuous stream of awakened activity, followed by a full night's sleep.

Many dogs approximately shadow our schedule, though they may also snooze during the day and perhaps wake up to play at times during the night.

But performance dogs, and especially dogs trained for Field Trials, are often on a different kind of schedule: two or more short training periods each day, with periods of rest and sleep in between.

It is not always easy for amateur trainers to provide such a training schedule for their dogs while maintaining their own busier schedule, but there may be good reasons to try:

A well-rested dog is less likely to injure himself.

Quiet time gives the dog an opportunity to assimilate what he learned during the preceding session, to ponder it and to sleep on it.

Dogs on a training schedule bring all their energy to those short training periods, and tend to work with high motivation and focus. By contrast, dogs shadowing a living schedule learn to pace themselves, bringing less intensity and concentration to each individual task.

In Field Trials, your dog will be competing against dogs with professional trainers who run a string of dogs. Those dogs typically get one training session in the morning, and another in the afternoon. The rest of the time, they are in their kennels, resting and sleeping and assimilating the day's lessons. Unless you have reason to believe that more activity will give your dogs a competitive advantage over that time-honored approach to training, even though you may train only one or two dogs, you may want to provide your dogs with the same kind of schedule that the Pros give their dogs.

Biting and Barking During Training

Normally, only train one thing at a time, but biting and barking are always relevant.
Don't allow biting when taking treats. Hold the treat deep in your hand, and require the dog to use his tongue to reach it.
If you reinforce barking, it can become part of the trained behavior. Common examples: "Throw the ball," and barking when backing up.
Therefore, don't reinforce barking. Instead, turn your back, wait 10-30 seconds, whatever amount of time results in the behavior declining.
If turning your back is not enough to cause the barking to stop, see the section Barking in the chapter "Problem Solving".
Whatever method you us, be absolutely consistent, and respond at the very instant of the bark.
Don't allow the biting or barking to continue during initial training of a behavior with the idea of fixing the problem later. It can transfer from being a general behavior to becoming specifically associated with that behavior, and it's much more difficult to untrain later.