The Get Out is fundamentally a Bend away from the handler’s position. The performance means that the dog should increase lateral distance. We make this an important distinction between a Get Out and either the Tandem Turn or Back Cross. In the case of the latter two movements we expect the performance from the dog to be a continuing turn, hard aback, reversing direction. In the Get Out we want the dog to shift his path at some lateral distance and then continue in the relative direction from which the movement was started.

We introduce the Get Out to the dog using a single obstacle that most dogs will attack with great relish… the pipe tunnel. Note that in the illustration the change in the dog’s path is a modest bend of maybe 5′. There is no option aside from the tunnel aside from—potentially—the wrong entry to the tunnel; so it should be easy to condition.

After having practiced the introductory step three of four times, the escalation is a matter of modest progress. In this illustration the pipe tunnel has been moved only 2′ further to the dog’s left. Note that the handler’s step is less intrusive and less forceful. We’re counting on the previous conditioned response to allow the handler to be a bit more modest and subtle in his movement cue.

We will practice this step with the dog three or four times.

Again we’ve moved the tunnel another 2′ to the dog’s left making the Get Out a bit more dramatic. Note also that the handler’s step is even less forceful and intrusive. It’s worth discussing at this point the dynamics of the handler’s movement when taking the step and, frankly, which arm the handler should use. It is my own believe that the handler is taking a step to the side at about 90 ºwithout rotating his torso. The step might be in the oblique but the lesser the degree of the intrusive step then the weaker will be the signal to the dog. Given this discussion, the handler really should be using his inside arm (the arm nearer to the dog) as though fending or warding something coming up from behind (which surely the dog is). If the handler were to use his counter arm that is likely to introduce a rotation to the handler’s torso which should mean for the dog to turn hard aback, as in a Tandem Turn.

I’ve included in this illustration a dog’s path colored red. This is an error that is often uncovered at this point in training. The clever dog will understand that the ultimate destination is to the pipe tunnel and so may want to cut to the chase and will wind up avoiding the bothersome jump #2.

We’ve taken another step by moving the pipe tunnel yet another 2′ from the dog’s path. By now the dogs should be keen to bend away from the handler’s position. Note that the handler’s step is barely noticeable or intrusive to the dog.

Now we take a proofing step. We introduce an option for the first time to the dog. An option might be considered a course that is more logical than the one the judge actually numbered. We see these all the time in competition as the misdirection is one of the course designer’s favorite tools.

It might be necessary in this step for the handler to return to a more intrusive step. But likely you’ll find that the dog has been well conditioned to the Get Out and is happy to bend away into the pipe tunnel.

With that in mind …

This is what I call the smart aleck step. It is an interesting test of the efficacy of our training. Now we want the dog to go away straight over the third jump into a sequence that is entirely different and supposedly logical. However, you’ll find that notable percentage of dogs will want to repeat the performance that has been earning the praise and reward. The red line indicates the errant path that many dogs will take after the foregoing training of the Get Out.

Generalizing

Generalizing can only be described as the dog’s ability to understand the performance when elements external to the desired performance differ from the training condition. In this case the handler

In this case the handler is asking for a Get Out on a different set of jumps, in a different part of the building, heading a different direction, and to a different object obstacle, a jump rather than a tunnel. It might be worthwhile to condition this several times with a more intrusive step to help the dog understand the desired performance.

As a training initiative we might remove the dummy jump forward of jump #2 for awhile in order to re-condition the Get Out that we worked so hard for the dog to understand while using the pipe tunnel.

With all of the foregoing in mind… note that all of our Get Outs have had the dog bending away from the handler’s position to the left. Does this mean that the dog understands a Get Out if the bend is to the right? Probably not. All of these conditioning exercises need to be set up and practiced in the opposite direction. A Get Out means that the dog should bend away from the handler’s position, increasing lateral distance, whether the bend is to the left, or to the right.

Note that the Get Out exercises are accommodated on the floor: the introduction to the pipe tunnel in the upper right, and the generalizing in the lower left. It is useful to break up the tedium of drilling be running other sequences from time to time. Here are a couple of useful sequences.

White Numbers – This sequence has a couple of interesting moments. On the approach to the weave poles the handler might get the dog into the weave poles with a simple Front Cross. This would, however, have the dog on left to get past the dummy jump and to get the dog into the pipe tunnel. Rather than a Front Cross, the handler might consider using a Flip (a combination movement, Front Cross followed by a Blind Cross) so that the dog will enter the weave poles on the handler’s right side.

On the other hand, if the handler uses the Front Cross and has the dog on his right side he could draw the dog past the dummy jump on his left, and then use a Tandem Turn on the flat to flip the dog back into the pipe tunnel. This is a simple exercise in timing.

Dog Training Mode

When the handler is in Dog Training Mode he is armed with these distinct tools:

· A clear vision of the desired performance

· A marker

· Reward for the dog

A clear vision of the desired performance – It is important that the objective of the training is clearly defined in the mind of the trainer. For example, if we are teaching the dog to do a jump, that means “over the bar and between the uprights.” Should we worry about whether the dog keeps up the bar? In the beginning, we surely should not, because that is beyond the scope of our definition of performance.

A marker[1] – Whether the trainer works with a clicker or has a well developed verbal marker really makes no difference. The marker should be given the dog at the precise instant that he offers the desired performance, not before, and not after. If the trainer is using a verbal marker it should sound as though he is supremely happy about the dog’s performance. It doesn’t really pay to be lackluster and boring. Dogs like enthusiasm and will respond to it in kind.

A mistimed marker can muddy the water considerably for the dog. If the trainer marks too early the dog might turn back before finishing the desired performance and has every right to expect that he should be rewarded because that’s what the marker means. If the marker is offered so late that the dog has engaged in one or two or more behaviors after the one that is the object lesson then the dog might be considerably confused as to which was the behavior that earned the marker and the subsequent reward.

Reward for the dog – This might be a food reward or a game with a toy. Note that the marker was a bridge between the desired performance and the application of the reward. In general the timing of the reward isn’t terribly critical. However I used to train with a woman who told me this thing “To get the timing of the reward correct all you have to do is count… one-thousand one… one-thousand … too late!”

What do you do if the dog doesn’t complete? My own response is completely neutral. That means, I don’t give the marker and I don’t give the reward. I allow the dog to sort it all out. I want him thinking… what do I have to do to earn the reward?

The Light Bulb Moment

When we’re in dog-training mode we seek the Light Bulb Moment. That is the precise moment that the dog understands exactly what it is that earns the praise (marker) and the reward. And when we’ve got electricity going to that light bulb the dog will begin to offer. He’ll start playing you, and working you, and doing that thing over and over again to earn the reward. This is the moment of true success. The dog has been trained to the fundamental task that was in our vision of performance.

We’ve put such a high reward value on a specific performance that the dog will be keen to offer the performance to earn a reward. Indeed you’ll find that dogs will offer a performance with a high reward value at the most inopportune moments. For example if you’ve been spending a lot of time teaching a dog a two-on/two-off position on the bottom of a contact obstacle he may veer away from an obvious direction on course to go plant his butt at the bottom of a contact obstacle. This should not be discouraged really. The trick will be to counter the high reward value of the object lesson with reward for the more mundane.

On Negative Markers

I personally don’t use a negative marker. I work so hard to create a powerful positive association with agility that I find it counter-productive to introduce any kind of negative whatsoever. I’ve not always been this way. Let’s just say that as an empirical learner I’ve learned the error of my ways through the productive teaching of a dozen of my family dogs.

A negative marker might range anywhere from a neutral and bland “wrong!” verbal… to scooping the dog up and beating the crap out of him. Admittedly, that’s quite a range. There are certain trainers in our game who have won fame and reputation by getting the right kind of dog for negative training. The dog should a) have a tremendous work ethic, and b) be quite hard. A hard dog is one who rebounds from abusive training. The fact of the matter is that the type of dog I’ve described here is rare enough that I could safely say that your dog probably isn’t one like him. Negative and abusive training will ruin your dog for the game of agility.

The problem with negative or abusive training is that the dog is wrong so often that he is unlikely to offer performance. Offering just gets the dog in trouble so he will be tentative and slow and frankly very reluctant to offer performance. If, on the other hand, you are a positive trainer who trains with effusive reward it is likely that your dog will be keen to offer performance. Is this right? Okay, how about this? Trying again… what do you think of this?