Training Cycles: Teaching, Refining, Motivating, Proofing and Practicing
Remember that dog training is full of progressions, but it isn't consistently linear. Certain increments of advanced exercises can be trained independently and then sequenced; but even when your dog has learned all of the formal sequences, your direction of training sessions has to be determined not only by the skills, but your intent. Are you teaching something, refining it, proofing it, motivating it, or are you just repeating a common sequence to set a habit? (If the last option, be critical about your evaluation of whether the habit you are ingraining is a correct one or not!)
Here are some examples of cycles for some different exercises.
Heeling: This is a very complicated exercise, so you need to think about single elements, such as take offs, straight line endurance of attention, right turns, about turns, halts, etc.. Let's look at take-offs for this example.
For motivating energetic take-offs, you might do the racing game cues, and take off with a couple of running steps and release. When you have the energy you want, then you refine it by taking a normal step and doing a hand target to make sure the dog is heeling rather than rocketing. If necessary, you might say, “heel” and wait a second or two before moving. Or you could take a forward step and then a backward step, targeting on both steps. To proof it, you might have somebody say the forward command, and make sure the dog doesn't cheat. Or you could cause a distraction and tell the dog to heel, and release and praise if he does it right. Finally, practice the pretty take offs with a series of forward, halt, forward repetitions. Then perhaps end your session with a couple of motivational takes offs so you end with the dog wanting more.
Retrieving exercises: All retrieving exercises have the same basic elements and some common problems that need tweaking and maintenance work. They all involve the dog moving toward the object briskly and directly, picking it up cleanly, turning back and hurrying back on the recall portion and delivering with a correct front and finish.
Behaviors that we don't like to see in retrieves can be trained out, but some are easier to work with when you consider that they are related to attitude and habits associated with action patterns in other contexts. Look at what happens during play retrieves with toys, and make sure you aren't accidentally encouraging practice of naughty behaviors. Pouncing or shaking things, for example, are normal play behaviors, and it's darned cute, so we laugh at it when a puppy pounces on a toy and then grabs it and shakes it. “Killing” with the feet becomes part of the retrieve motion pattern. For the average first-time obedience trainer, the technicalities of the retrieves aren't even known until the dog has been killing toys for a couple of years, and then the owner tries to change the rules to the frustration of all concerned.
In order to change this behavior in a way that maintains enthusiasm, while running fewer risks than some of the directly physical means of correction come with, my favorite approach is to load the dumbbell with squeeze cheese. This motivates the dog to think about getting his nose and mouth to the dumbbell instead of his feet. So when you begin, you are teaching the skill of approaching in a different way physically. It takes a large number of repetitions of placing or tossing a loaded dumbbell and sending the dog, telling him to get it as he approaches to let him know you've put food there and he should look for it, and allowing the dog to eat the cheese before picking the dumbbell up, to create a new habit.This is a practice phase. (Remember that as you progress on this, silence as he approaches means no treat, and he should pick it up and hasten back.) At a minimum, I'd recommend 10 repetitions per day for at least three weeks with the dumbbell loaded every time before you start randomizing the presence of cheese (very SMALL amounts of cheese!). But the more deeply ingrained the habit is, the more maintenance you have to do on the new thought and behavior processes on a continued but random basis.
An additional way to motivate the new way the dog thinks about approaching the dumbbell is to use it to present rewards to your dog for other exercises, like recalls or stays or signals. Point at the dumbbell, let the dog watch you put his reward on it, and then present it to the dog. The dog starts to think about getting his mouth to the dumbbell in contexts when it is not thrown. You could also place the dumbbell off to the side of your working area with a treat on it, and release the dog to go to it when he has done something wonderful. It's your choice whether you just let the dog have the treat, or you ask the dog to then pick the dumbbell up. For long time pouncers, just allowing them to get the treat is good for awhile, but as soon as you are seeing non-pouncing approaches regularly, move on to having the dog pick up and deliver the dumbbell after eating the treat.
Another way to practice a controlled state of mind is to put in other commands (down, sit, stand, etc.) before or instead of a retrieve command when you throw a dumbbell.
Yet another way to modify the pouncing impulse is to cue the dog for a recall or a racing game while he is on his way to the dumbbell. This also is good for motivating tighter turns and faster recalls back to you. You'll have to experiment with the timing a bit, as cuing too early may get him to abandon the dumbbell, but that's okay; just tell him to fetch again, and likely he will do it without pouncing because you interrupted that pattern. Then run like the wind as he returns your direction to motivate the dog to drive hard on the recall part.
BTW, the glove retrieve can be approached the same way (although with solid treats rather than squeeze cheese on the glove).
You can also set up gates or other obstacles for your dog to work through and around on his way out to the dumbbell or a single glove to work on his commitment to task and distraction resistance.
Once in a while you actually do a formal retrieve; but whatever your problem area is, you will always be cycling through motivating or sharpening a particular response, and maintaining a consistent pattern of response as you cycle through the elements.
Now for some dogs, one of those approaches will be just what you need to fix the pouncing problem. But playing with all of them as you train is a form of randomizing the appearance and requirements of the task, and keep your dog sharp and alert in his work.