by Deborah Jones, Ph.D.


Kafka is a 2 1/2 year old unneutered male Italian Greyhound.

Presenting Problem:

During agility training Kafka will exhibit the 'stress zoomies' and race around the training area out of control. According to his owner "I think he's stressed because right before he starts zooming, his ears are pinned back tight to his head and his body is very taut and rigid. He always comes to a dead stop right before running off, and usually turns his head to one side. There's always that moment of hesitation when he's heard me give a command, thinks about it for a second, and then runs away".


Kafka came to his current owner at about 18 months old. Prior to that he had lived with his breeder. He was shown in conformation and did a bit of lure coursing (but preferred chasing the other dogs to chasing the lure).

Since coming to his current home Kafka has been training in obedience and agility. Kafka's owner has been training mostly on her own. She reports that she has used a "semi-clicker approach to training". Kafka has earned one leg towards his Companion Dog title.

More recently Kafka and his owner have been training at an agility club. She reports that she fits his training in between advanced classes with her other dogs.

Previous Episodes:

Kafka's owner reports that this behavior was first exhibited while training in her backyard. She was working on having him walk on a slightly raised board when he took off and raced around the back yard. This behavior now occurs two to three times during a training session.

His owner reports that the "zoomies" are most likely to occur in connection with the dog walk, broad jump, and (most recently) tunnels. At a recent trial Kafka's owner noticed that he kept refusing tunnels. Working later at home she discovered that when she would hesitate at a tunnel he would stop. If she insisted he complete the tunnel, he would start zooming. She also reports that this behavior is more likely to occur towards the end of a course or sequence.

Unfortunately, it seems likely that this behavior will increase in frequency and generalize to other obstacles. More on this below.

Previous Treatments:

Kafka's owner has tried a variety of techniques to deal with this problem.

She states that he comes back faster is she doesn't call him, but instead turns her back to him and ignores him. When he returns she has him finish the obstacle he ran away from, then rewards him and stops.

If she wants to continue working and Kafka is still not concentrating she puts the leash on him. He seems to relax and works better with the leash on than without it.

She works with Kafka in short training sessions (5 minutes) and obstacle sequences (4-5) in order to minimize stress.


Kafka is exhibiting behaviors that would be considered 'positive stress' reactions. In this case the term positive refers to adding something. Kafka is becoming more active (adding behaviors) when under pressure.

When feeling unsure, pressured, or overwhelmed Kafka uses unfocused activity (running laps wildly) as a way to decrease his stress level. This technique works as it releases the stress that has built up and allows him to relax and regroup. It is a self-reinforcing behavior that operates on the principle of negative reinforcement.

The behavioral definition of negative reinforcement is the removal of an unpleasant stimulus that increases the probability of the behavior it follows. Negative means taking something away, in this case the stress, and reinforcement refers to something designed to increase a behavior, in this case the "zoomies". For Kafka, performing the "stress zoomies" relieves an unpleasant emotional state and is very reinforcing. It is a coping strategy that he has discovered on his own.

Because engaging in this type of behavior is reinforcing and stress-reducing, it is very likely to increase in frequency and to generalize to other situations as well. If it works in one situation, Kafka is smart enough to try it out and see if it works in other tense situations as well. Pretty soon, even in only slightly tense situations, zooming will become a default (automatic) behavior.

We all engage in stress-relieving activities and, if they work, we continue to perform them. Humans have a number of options for releasing stress through actions such as eating, drinking, smoking, exercising, cleaning house, etc. While any of these activities, taken to excess, have the potential for harm, some are more inherently dangerous than others. Dogs don't have as many behavioral options. Typical positive stress reactions in dogs involve running, jumping, lunging, spinning, and chewing. They are all ways of releasing tension and energy.

*In another column I will present a case that deals with negative stress reactions. These are often seen as 'shutting down' and absence of behavior, rather than an increase in behavior.


1. Being aware of the warning signs is an important first step in managing this behavior. Kafka's owner has mentioned that there are certain signs that she has noticed prior to the zoomies. She has stated that he tenses his body, pins back his ears, hesitates, turns his head to one side, looks at her, and then takes off. These early warning signals are useful as they will give her a chance to interrupt him before he goes into full flight mode.

2. It's very important to try to manage situations so that the zooming does not have a chance to occur. Once it happens, two undesirable things have happened. First, Kafka has had another opportunity to practice an unwanted behavior and make it a stronger habit. Second, Kafka's owner has moved from training mode to crisis management mode. The best she can do at that point is damage control, but she has not taken proactive steps to solve the problem.

As soon as she notices the first warning sign Kafka's owner needs to change the flow of events. She can do this by training and practicing some more acceptable stress-relieving exercises (described below). Before Kafka has made the decision to run, she can divert his attention with these exercises and activities. Once Kafka's tension has been released through these exercises, he can be asked to work again. She can also avoid the build-up of stress by changing the nature of the exercises that Kafka is asked to perform and moving him from more stressful to less stressful activities as appropriate (more on this below).


It takes some experimentation to discover exercises and activities that will help any particular dog to release stress. While one dog may respond well to calming activities like stretching and massage; another may benefit from active games that involve jumping and tugging. Also, be aware that asking for compliance to obedience commands when the dog is in this overexcited state could be counterproductive. A dog who is overly stressed will not respond in the usual manner, which then frustrates the handler and may lead to more problems. Instead, engaging in mutually enjoyable activities can be helpful here. Below is a list of possible exercises to try out and observe their effect.

*Stretching. Stretching exercises can allow for some focused activity that will help to release stress. These are not obedience exercises and can be done with a food or toy lure if desired. Lead your dog through a series of sit/down push-ups, stand/down accordian style movements, and sit/stand on hind legs jumping jacks. You can also teach and practice spins in both directions (most dogs have a preferred direction). This can also be developed into a cute 'chase your tail' trick. A play bow type stretch and a stretch of the hindquarters can all be taught.

*Massage. You don't have to be a masseuse in order to provide relaxation for your dog through touch. However, there are a number of books and videos available for those interested in learning specific techniques. Just remember to keep a close eye on your dog's reactions to massage. Dogs will move away from uncomfortable sensations or give small signals such as pinning back their ears and quickly turning to look when you hit a sore spot. Keep the massage calm, slow, and quiet. The purpose is to relax, not to energize.

*Tricks. Tricks can serve the purpose of giving the dog an outlet for his physical energy while keeping him focused on the handler. Roll over, crawl, sit up, and high 5 are all fun possibilities. Once the trick has been taught, it can be used BEFORE Kafka has become overly stressed as a way to defuse his tension. A 'quick trick break' randomly in the middle of an agility run would be a great way to keep Kafka's interest and enthusiasm high. An example sequence combining tricks and agility might be jump, tunnel, spin, A-frame, roll over, high 5, jump, teeter, sit up, etc.

*Games. Kafka's owner might play some games with him to defuse his tension. Again, it is very important that the games are played BEFORE Kafka becomes overly stressed, not after. Different dogs enjoy different games. Some like tug and can play it nicely; others love to be chased. Some like a wrestling type game; others like to bounce and jump. Kafka's owner should work at finding out what his favorite games are and play them with him before, during, and after agility.


Kafka's owner has been observing his behavior and has noticed cues in his body language that suggest that "zooming" is about to begin. She has also noticed that Kafka is more likely to be stressed by certain obstacles, including the dogwalk and broad jump. These are very important and useful observations.

In addition, it is important for Kafka's owner to start monitoring his stress level (based on his body language and behavior) at various points during agility training and showing. Imagine a stress level of 10 being the most intense stress reactions that Kafka could exhibit. With a stress level that high Kafka would be in the midst of an incredible zooming session and would be unable to control his reactions. On the other extreme, a stress level of 1 would be a very, very relaxed dog (probably lying quietly at your feet). Either extreme would not be conducive to learning or performing in agility. What we would like is an 'optimal level of arousal'. An ideal level is one that has enough excitement to motivate activity, but not so much that it induces overactivity or overanxiety.

Kafka's owner should practice rating his stress level (from a low of 1 to a high of 10) at different times in training and trialing. For example, when they are warming up and just starting out, she might note that his stress level seems to be about a 3 (low average). However, as he approaches the end of a difficult sequence and comes upon the dogwalk, she might note that his stress level is around a 9 (very high). It takes keen observation skills to make accurate ratings, but this is very important to closely monitor stress levels with dogs who are easily overwhelmed. The average scores of 4, 5, and 6 would be ideal to observe while working. For the present time, a stress level score higher than 6 is hitting the danger level for Kafka. When his level seems higher than that, it's time to introduce a stress-reliever, an easier or preferred exercise, or a short break. Continued attempts to train or show will be counter-productive when the stress level increases.

In addition to monitoring and rating Kafka's stress levels, his owner should also observe and rate the obstacles and exercises on the amount of stress they seem to induce. The obstacles that cause the highest stress reaction earn a 10; while easy ones earn a 2 or 3. It is important not to overwhelm Kafka with lots of high stress obstacles in a single training session.

Kafka's owner should also think very carefully about how she plans her training sessions and how she structures the exercises within them. Taking the time to sit down and draw up a training plan will be immensely helpful. Rather than jumping in and running a couple of sequences here or there, Kafka's owner should seriously consider what she is training and how she will go about reaching her goals for him. A training plan that slowly and carefully introduces new techniques and challenges can help to relieve stress in both Kafka and his owner. The training plan doesn't have to be complex and comprehensive to be helpful. For example, Kafka's owner may decide to work on 3-4 obstacle sequences that include at least one contact obstacle. When he is able to run a variety of these sequences smoothly, she might then add an additional challenge such as a simple discrimination or a blind cross. In addition, she might decide that at least once in the sequence (randomly) she will take a 'stress break'. The key is to be proactive in planning her training sessions rather than going out to 'see what happens' and 'hoping for the best'.

By introducing mild to moderate training challenges slowly and carefully, Kafka's owner will slowly raise his tolerance for stress. She can then add bigger challenges or more challenges within a training session.


One way to approach this problem is to use zooming as a reinforcer. Since Kafka has shown that he finds it to be a good stress reliever, his owner could use it to reinforce good performances and to keep his stress levels low. To take control of zooming and put it on a cue, Kafka's owner can start by playing with him until he gets very revved up and excited. When she sees him about ready to burst, give him a "go zoom!" cue and run as fast as possible with him for a short distance, verbally encouraging him to run. Then call him back and reinforce him (with a very high value reward) for returning. Use this release cue regularly during play and training. Instead of trying to avoid zooming at all costs, this technique allows the owner to take control by deciding when it occurs and by giving permission for it.

Becoming more creative in her application of reinforcers may also help Kafka's owner keep him focused on her and on the work. One possibility is the use of jackpots during training. A jackpot can be either a large amount of a reinforcer or a very special reinforcer. For example, rather than giving a dog one treat for his efforts, you might give him a whole handful. For some dogs, having a chance to eat the treats right out of the container is special and makes a big impression. Others enjoy being fed a large number of treats one right after the other. Another way to jackpot a dog is to give a very special and rare reinforcer. I use salmon and have found that a few small bits of this delicacy makes a huge impression on my dogs. The jackpot should be used BEFORE the dog gives signs of being overly stressed. The owner should just stop suddenly and unexpectedly in the middle of a sequence and give the jackpot.

Using a number of different types of reinforcers during training sessions will also help keep Kafka's interest high and his stress level low. In addition to a variety of food treats, Kafka's owner should also use toys and play as rewards. She needs to keep him guessing about what wonderful thing she may have in store for good performances. It's very easy to get into a rut and offer the same old reinforcer time after time. When this happens our dogs get bored and we are no longer very interesting to them. It takes a bit of creativity and planning prior to each training session, but it is well worth the effort.


This is a situation in which many owners, trainers, and instructors would say that "the dog is blowing you off". This statement suggests intentional misbehavior on the part of the dog. It suggests that the dog completely understands your request, that he is physically and mentally capable of performing it, and that he consciously chooses not to. Underlying this statement is the suggestion that the dog is deliberately defying the owner's authority. This type of statement is damaging to the dog/owner relationship and shows a clear lack of understanding of learning theory and canine behavior.

The "blowing you off" explanation for training/performance errors often occurs when the trainer cannot identify an obvious explanation for the mistake. Instead of objectively analyzing the situation for causes, the trainer resorts to making an unsupportable assumption about the dog's motivation. Good trainers know that errors occur due to insufficient experience and training and due to environmental changes and causes. The dog's behavior is a reflection of the training history and the current situation.

An explanation of canine behavior based on assumptions about the internal emotional state of the dog, particularly when that emotional state goes beyond basic emotions such as fear and excitement, is ultimately not useful. Since dogs cannot tell us how they feel or what they 'think' there is no way to objectively test our theories about their internal states. We are only guessing, and our guesses may be completely wrong. The entire science of behaviorism is based on avoiding such errors by limiting ourselves to observations of actual behavior rather than resorting to guesswork.

Back to Articles Index

© 2005
Built and Maintained by Pet Web Designs

Home | About | Books & Videos | Reviews | Workshops | Calendar | Seminar Contact Form | Photos | Articles