Our goal as trainers is a well-trained dog. Most of us want to move toward that goal as quickly as possible. In our quest for rapid success, we tend to raise the bars too soon, push our dogs too hard, and cause our dogs to fail.
Don’t give into this temptation! Nothing will sabotage good training as quickly as training at levels that are too difficult for the dog. It is important to keep your dog successful. Behavior cannot be reinforced unless it is correct. Good training cannot be rushed. Practice makes perfect. Success builds confidence and motivation. Therefore, for the best, the most solid, and yes, even the fastest results, train under what we call the dog’s threshold of success.
Here is our definition of this very important concept:
The threshold of success is the level of training at which a dog will be successful if trained at a level below it and will fail if the dog is trained at a level above it.
To state it another way, the height of the bar that delineates between when a dog will be successful and when he will not is a dog’s threshold of success. When we train under the this bar, we will gradually bump up the threshold of success. When we train over it, we will bump it down.
The goal in training should not be to see how high of a level we can train at any given moment, but rather to bump that bar or threshold of success up giving our dog the ability to work at higher and higher levels in the future. We do that by training under the threshold.
Why train Under Threshold
There are many reasons that the success rate should be as high as possible. Here are a few.
1. Success creates good habits. Failure creates bad habits.
2. Success builds confidence. Failure reduces confidence.
3. Success builds motivation. Failure reduces motivation.
4. Success builds clarity and understanding. Failure can cause confusion.
4. Training under a dog’s Threshold of Success builds a stronger foundation for longer lasting results.
5. Practice makes perfect. You won’t hurt your dog’s training by making the training too easy.
When Your Dog Fails
Small amounts of failure followed by success builds resilience. Therefore, don’t worry about it when you do push your dog over his threshold and he fails. It will happen simply because you can never know for sure exactly where that threshold is in any given moment.
However, there will be plenty of accidental failures to build resilience. Don’t set your dog up for failure just to build resilience. It is more important to focus on all you are building through success.
Try to keep your dog successful, but don’t put too much pressure on yourself to be a perfect trainer! Simplify the task when your dog fails so that it doesn’t happen again and move on.
How To Increase the Threshold of Success
There are three variables that need to be changed for a dog to progress in training. These same variable must be carefully monitored because changing them too fast will cause the dog to fail. As we talk about these variables, we’ll discuss how to change them in such a way that the dog progresses in training, while still remaining successful and motivated to work. These variables are:
1. Difficulty of the exercise.
2. Criteria or expectations.
3. Rate of reinforcement.
How to Increase the Difficulty While Keeping Your Dog Successful
For a dog to progress in training, he must be able to master increasingly difficult exercises. Otherwise, training will not move you towards your goal of a well-trained dog. However, you must be careful as you increase difficulty. Here are some guidelines to help you balance increasing difficulty while keeping your dog successful.
1. Don’t increase difficulty so fast that you cause your dog to fail. You can’t always identify the exact level that your dog is capable is performing. You must guess. As you guess, err on the side of caution. You won’t mess your dog up by training at too low of a level. You can set your training back if you train at a level too difficult.
2. Only increase one aspect of difficulty at a time. For example, don’t increase both the time and the distance at the same time when training a down/stay.
3. Increase criteria gradually. A big increase in criteria will hurt confidence (even if the dog appears successful!).
4. Try not to go over threshold, but when you do, make it easier the next time so that the dog doesn’t repeat the mistake.
5. Remember that every time your dog is successfully working under threshold, you will be bumping the threshold up higher.
6. Remember that every time your dog is pushed too far and asked to work over threshold, you will be knocking your threshold down.
7. Remember that a dog’s threshold of success will vary depending on the environment, the dog’s energy level, the dog’s hunger level, the trainer’s skill level, and even on the dog’s momentary mood. Dogs, like people, have good days and bad days. They even have good moments and bad moments. We as trainers must always be evaluating where our dog’s threshold of success is and adjusting our training to stay under it.
8. Do not push your dog to work longer or with more difficulty than he’s able. Don’t worry if progress seems to be slow. Progress should be exponential, not linear. The basics need a very long time to train. Once the basics are trained, progress is surprisingly fast.
Increasing Criteria and Expectations to Earn Reward
During the initial training of a behavior, criteria for reward is nothing more than effort. Then, with reward placement, we move the dog towards more complete or perfect performances. We gradually increase our expectations of offered behavior before we mark and continue trying to create even more perfect performances with reward placement. Our dogs give us feedback by either his successes or his failures. If the dog continues to be successful, we can assume that we are increasing criteria at a good pace. If the dog fails, reduce criteria to keep the dog from failing twice in a row.
Eventually, the dog should be able to perform black and white behaviors perfectly without reward placement. What we mean by black and white behaviors are those behaviors or skills that are clearly either right or wrong. For example, performing a “sit” is black and white. The dog’s butt is either on the ground or it is not.
Once a dog has been trained to perform a black and white behavior consistently without reward placement, criteria for that behavior is always perfection. Don’t reward a dog for a down where elbows are raised off the ground or a down/stay where the dog is inching forward. Though criteria remains constant, the difficulty can be increased or reduced as well as the rate of reinforcement which we will discuss in the next chapter.
Expectations for Exercises With Degrees of Perfection
Other exercises are not black and white with regard to being performed to a right or wrong standard. These exercises have varying degrees of correctness. Examples of non-black-and-white skills would include sitting straight in the front or finish position and heeling in good position. The correctness of these behaviors is subjective. Whether or not these behaviors are correctly performed depends on the expected level of perfection. Staying in precisely the correct position while heeling, sitting straight when called to a front or finish position are examples of behaviors where perfection is the goal, but there is no clear right and wrong.
For the most part, those behaviors that are in our more subjective group when it comes to criteria are more complex behaviors where there are elements of the behavior that can have clear right or wrong criteria and other elements that are subjective as to correctness.
When we train complex behaviors, we break the behaviors down into the smallest trainable parts. Initially, we train every piece of complex behaviors separately. Therefore each piece of a behavior must be evaluated separately as to which group it falls. Heeling is an example of a complex behavior where some pieces of the behavior belong in one group and some fall in another. When eye contact is required, success is black or white. Eye contact is either given or it’s not. Accuracy of position, on the other hand, is an element of heeling that is subjective. There is no clear right or wrong. There are only higher or lower levels of perfection. We train the black-and-white portions of complex behaviors first. Then we work on the more subjective skills.
When increasing expectations and criteria for reward with regard to non-black-and-white skills, our expectations increase gradually. We do not expect perfection. Instead, we gradually increase our expectations as the dog’s skills grow getting ever closer to perfection. How we handle these less than perfect performances vary depending on the skill being learned and the level of training the dog has reached. We’ll discuss more as we get into the individual behaviors and skills.
Decreasing the Rate of Reinforcements
Rate of reinforcement can and should be reduced, but there is a process that must be followed in order to keep the dog motivated and interested in the work. Here are some principles that should help with shifting your dog from working for food to working for the love of the work.
1. Fade luring as soon as possible.
2. Fade reward placement as soon as possible.
3. Get the food off your body as soon as possible.
4. Never reduce reinforcements all together.
5. Instead of gradually reducing reinforcements in a linear fashion, move your reinforcement schedule from consistent and predictable to random and unpredictable.
6. Reduce reinforcements in such a way that the dog continues to think that a reinforcement could come at any time.
The Steps for Reducing Reinforcements
1. Use continuous reinforcement when a dog is first learning a behavior. Reinforce every correct response until the behavior is solid with required criteria and in the environment you are training.
2. Move to reinforcing the dog’s best efforts when the dog is proficient with the behavior in the current environment and when the behavior can be chained together with other behaviors. Chaining behaviors is asking the dog to do more than one behavior between reward events. For instance, if I ask my dog to down and then to sit, I would have chained two behaviors together. Effort must be reinforced while at the same time mistakes are not. Therefore, the dog must be set up in such a way that when a dog is putting forth his best efforts, there is a very high probability that he will be correct. We’ll talk more about this later.
4. When chaining together multiple behaviors, have the most difficult behaviors at the end of the chain so that they are rewarded more often than the easier behaviors. When you chain behaviors together, only the last behavior is actually being reinforced.
5. Begin using jack-pot rewards when you begin chaining multiple behaviors together between reinforcement. These are larger rewards that are given at least several yards away from where you are training. With a jack-pot reward, the dog hears the marker word and then runs to a designated place where the trainer gives him a small bowl of food in a crate or on the floor or ground. Because the dog is randomly being asked to work longer sequences and because the food is no longer close by, he needs a larger reward than a single treat or a few pieces of food to keep him motivated.
6. Always reinforce correct behaviors when the dog has struggled with particular behaviors. If a dog makes a mistake, make sure that during his next try at that behavior he is correct and gets reinforced for it. Make the exercise easier if necessary.
7. When your dog is very fluent, move to random rewards. Random rewards are more highly reinforcing than consistent rewards. Use random rewards only when the dog is very proficient in the environment and under the circumstances he is currently being trained. When mistakes are made, go back to continuous rewards for the behavior where the mistake was made until proficiency is again attained. Then chain together that particular behavior at the end of the chain so that it gets more reinforcement. Basically, when mistakes are made, go back to step #1 and work through the steps. If a dog is unsure, he needs a consistent reinforcement schedule so that he knows and understands and gains confidence as to when he is right.
8. Practice full run through’s rarely and only once every piece is well-learned in a large number of environments. When a full run through is done, follow these longer performances with drive-building short sequences sandwiched in between high value reward events.
9. Most importantly, be careful not to train over threshold. If your dog is failing, it is usually because he is not being reinforced enough. When mistakes are made, go back to continuous reinforcements (step #1) and gradually ,move through the steps again.
Random Reinforcements and a Full Obedience or Rally Run
When training a dog to run an entire obedience run, people tend to think that the key is to gradually decrease reinforcement until the dog is working for the length of time required to work in the ring. This type reduction of reinforcement will rarely work because as the reinforcement becomes less and less, the dog’s motivation to work follows suit and also becomes less and less.
In order to acclimate your dog to work for longer periods of time in a ring setting, there is no need to perform at that level over and over. Performing longer runs once or twice just before a trial is sufficient. These runs don’t need to be full runs. I recommend running a run that is about 3/4 the length of time of your upcoming trial run twice over the few days preceding the trial.
The key to preparing your dog for longer sequences between reward events is to continue building value in the work with varying levels of reinforcement and to keep the dog guessing as to when the mark could come. Here are some tips for preparing your dog for a trial where longer periods of time without reinforcement are required.
1. Sometimes, mark and reward the dog for simply setting up correctly and with enthusiasm.
2. Sometimes, chain together multiple behaviors.
3. Occasionally, throw in almost a full run.
4. Never do long sequences back to back.
5. Sandwich longer sequences between shorter ones.
6. Never let the dog stop thinking that the reward could come at any time.
When you go into an obedience ring with a dog that is not a seasoned competitor, you are most likely asking your dog to work over the threshold. Hopefully, your dog is prepared enough to where you won’t do much damage to his training. However, if done over and over without heavy reinforcement in between, your dog will lose significant motivation and confidence. What will enable your dog to perform at this level at a trial is training at home using random reinforcement and classical conditioning. Random reinforcement is important, but classical conditioning is the secret sauce for changing a dog who doesn’t want to work into one who loves it enough to do it even in stressful environments.
When you go into the ring at a trial, you cannot take food or toy. Neither can you take the threat of punishment (at least not for the long haul). The dog will eventually learn that neither reinforcement or punishment is there. The only thing left to motivate a ring wise dog is his own love of the work. That can only be built with classical conditioning. Classical conditioning can only happen with positive reinforcement and in order to compensate for the difficult environment of a trial, there must be LOTS of positive reinforcement.
You can’t drain your dog’s battery before a trial by consistently reducing reinforcement. You must heavily charge up the dog so that he’s full of love of the work. Then in order to not shock the dog’s system with a long series of behaviors without reinforcement in the ring, you must occasionally add in longer random work.
