Za Neyko
in seveda vse ostale, članek o targetingu od očeta te metode,
Garyja Wilksa. Po mnogih letih premora je od pred kratkim spet v prodaji originalni alley oops
, znamenita premična tarča, katere sodobnejšo različico smo naročali nekaj postov višje.
Wilkes je na targeting šolal pse za obedience, saj se da s tarčo naučiti zelo filigranske gibe pri denimo poziciji na poleg ipd. Pomembno je, da vodnik razume poanto te metode, potem je prenos dotikov z različnimi deli telesa v vaje, trikce ipd. resnično neomejen. Keller Beland je dejal: "
I can train ANY behavior that the animal is physically and mentally capable of doing." Jaz imam psa, ki se ne odziva na hrano, zato so ga v eni pasji šoli označili za nesposobnega, težko učljivega ipd., saj se na učenje z luringom, kakor smo delali, ni odzival. Učenje s tarčo pa mu je zelo všeč, na ta način sem ga naučila vedenja, za katera so mi dejali, da se jih moj pes ni sposoben naučiti. Kakor ni težko spregledati
, sem velik nasprotnik luringa, a velik pristaš free shapinga in targetinga. Seveda pa tako kot moj pes ni učljiv na hrano, obstajajo tudi psi, ki niso učljivi na targeting. Je treba poskusiti in videti, kaj kakšnemu psu bolj ustreza. Za targeting lahko s tarčami veliko improvizirate; jaz trenutno psičko učim, da jaz ležim na hrbtu, imam noge v zraku, ona pa mi skoči na stopala. Kot tarčo sem zaradi njene oblike, ki mi je tu prišla prav, uporabila kar - kuhalnico
. V bistvu je kuhalnica prav izvrsten target stick
.
Targeting - Your second most important toolWhen I was about seven, my parents made a decision that they would come to regret - they let me take violin lessons. One of the horrible side effects of learning the violin is the "Gawdawful" made during practice. My first efforts as a violinist sounded about like a Bull Terrier at a cat show. That's when I learned about "fundamentals."
Fundamental behaviors are best described as the essential or foundational skills necessary to learn and perform an activity. Often, fundamental behaviors are intended as stepping-stones to more complex behaviors. For instance, before you can read and write you must master the alphabet and a pencil. One of the important fundamental elements of clicker training is "targeting."
Targeting is the process of triggering instinctive behaviors by controlling a dog's focus. The most common use of targeting is "luring" -- where the trainer uses a small piece of food to focus the dog's attention and thereby trigger a desired behavior. This ancient method precedes the Pharaohs - and has survived for thousands of years for one simple reason - it works. While food luring has some limitations, it was far superior to earlier methods of brute force.
While food luring has its place in the trainer's kit-bag, it has several obvious drawbacks.
1) It doesn't work well at a distance.
2) The animal may never be able to perform the desired behavior without looking for the treats.
3) The dog's enthusiasm may drop off severely any time he discovers that the food lure is lacking.
4) When using food for distance work, it is tough to prevent the dog from getting the reinforcement of the food, even if he does the behavior incorrectly.
To overcome these limitations without losing the essence of targeting is pretty simple. Instead of using actual food as the lure, we need a target that can't be eaten. That's where a "target stick" comes in handy.
Target Stick Basics: About twelve years ago, I developed training methods that relied heavily on non-food targets. Back then, I provided my clients with my first attempt at a production model target stick. My "Model T" was a 30" piece of ½" PVC. I coated the end cap with liquid vinyl for contrast and safety. My next model was a 3/8" wooden dowel with a rubber furniture tips glued on the end. I eventually came up with the collapsible, anodized aluminum target sticks that are now popular, but don't fool yourself - the art is in the trainer's hands, not in the material of the tool. To quote Jon Lindberg, a branch from an apple tree can work just as well. So, if you don't already have a target stick, get thee to a home improvement center and make one. Don't bother about making it collapsible - if you need to fold it up often, you aren't using it enough.
Holding a Target Stick: With your dominant hand, circle your thumb and forefinger into a classic "OK" sign. Put the clicker between your thumb and forefinger, while leaving your other three fingers extended. Grasp the stick with your remaining three fingers. This allows you to hold the clicker and stick in the same hand. Now rotate your wrist so that the stick is pointing straight down, as if you were stirring a pot with a skinny spoon. OK, you're ready.
Introducing the stick: There are a number of dogs who are initially frightened by a target stick. This is not an indication that the dog has been beaten. Dogs that are leery of target sticks are usually just reacting to an unusual object that needs to be investigated cautiously. To avoid getting off to a poor start, here are three ways to introduce the stick.
1) Hold the stick so that the tip is barely peaking out of your fist, with the rest of the stick pointing away from the dog. Slowly extend your fist toward your dog's nose -- click and treat if the dog investigates the tip or your fist. Over a series of repetitions, gradually extend the stick so that the tip becomes more and more visible. Continue until the dog is comfortable with seeing the stick in your hand. Gently integrate movement into the process so that a moving target stick is a "good thing."
2) Place the stick on the ground and put a treat under the tip. As the dog investigates, click, just before the dog gobbles the treat. After a few repetitions, hold the other end of the stick and lift it slightly off the ground. Continue to place treats under the tip as you raise the other end. Once you are sure the dog is comfortable with the stick, raise the tip off the ground. The dog is quite likely to investigate the tip - click and treat.
3) Smear a bit of hot dog or peanut butter on the tip of the stick and let the dog lick it off. Click and treat periodically, and replenish the peanut butter as needed.
Once the dog is comfortable with the stick, we'll start with some targeting fundamentals. First, hold the stick so that the tip is up in the air. Gradually lower the tip so that it is slightly above the dog's nose and a little to one side. Now be still and don't move. Give it at least 30 seconds. If the dog moves toward the tip, click and treat, even if the dog doesn't go all the way. Repeat this until the behavior is on a short cycle, like 5 seconds.
Here's the sequence:
1) Hold the tip in the air, so that the dog can't touch it - say "Touch".
2) Bring the tip down near the dog's nose and hold it still.
4) The dog bumps the tip.
5) Click & Treat.
6) Repeat.
If the dog makes an error during these trials, say "wrong" or "sorry" or your choice of sounds to indicate the error and put the tip up, out of reach - this tells the dog "end of behavior", no treat. Make sure you don't say the "wrong" word with any inflection - it's supposed to be informative, not discouraging.
Now that you have the basic behavior happening, it's time to stretch your dog's knowledge. Take a small object, like a baseball hat and place it on the floor. Begin the session with the stick behind your back, out of sight. Say "Touch" and then touch the object with the tip of the target stick. As the dog touches the tip, and the hat, click and treat. Do this about 10 times, quickly. On repetition eleven, leave the stick behind your back and say "Touch." Wait for at least 30 seconds. If you have built enough momentum into the behavior, your dog will probably move toward the hat - click and treat for any movement in the right direction! Continue this process until the dog will consistently touch the object when the target stick is not visible.
Once the dog will consistently "touch" the object, it is time to make the behavior a little tougher. After one of the repetitions, take a step backward. This places you one step away from the target. On the next attempt, as the dog touches the hat, click and then put the treat on the ground, right between your feet. (If you don't want your dog to take food from the ground, offer the treat from your hand, instead.) As the dog finishes the treat, say "Touch" (or "Hat", if you choose) and wait, quietly. After a few seconds, the dog will go and bump the hat or at least start in that direction. Click and treat for any movement toward the hat. On the next try, take another step backward as the dog is moving away from you and toward the hat. Over a series of repetitions, the target will be imperceptibly farther and farther away. Now pat yourself and your dog on the back. This is the fundamental behavior that we will use to teach a "go-out."
Over several training sessions, practice this behavior until your dog can touch the hat across the room, or across the yard.
Targeting Continued - a practical applicationOne of the most important tasks of training is controlling your dog's movement from a distance. Whether your interest is competition obedience, agility, search and rescue, field trials, or finding a Frisbee in high grass, using your dog's new skill at targeting can make directed movements a snap.
If you have been keeping up with the assignments, your dog should be comfortable with this sequence -- moving away from you, bumping one of your arrow targets and returning. This basic pattern is the key to making your dog's movements predictable and thereby controllable. To start this month's project, gather your clicker, treats, target arrows and of course, your dog. Select a relatively quiet outdoor where you can work off lead safely.
First, place a target arrow in the ground and move about five feet away from it. Sit your dog facing the target and stand slightly to the side. Point at the target and say "Touch" After all your practice last month, your dog should blast over to the target and touch the ball - click and treat.
NOTE: If you don't get this reaction on the first try, drop your standards and allow your dog to succeed at a shorter distance a few times. If you have been working on other things, go back and review last month's exercises and gradually build up to a short go-out.
Once you have achieved your simple go-out, repeat the sequence about 10 times, quickly. To make it a bit more interesting, take a step or two to one side or the other as the dog is moving away from you. This will insure that he must actually look to see where you are on the return. Your goal, for now, is to create a pattern that the dog can repeat almost without thinking about it.
ADDING A WORKING CUE:
At this stage of the behavior, a command would help to initiate each repetition. Since the behavior is currently a pretty poor example of the end product, it's a good idea not to attach the cue you eventually want to use in competition -- we will go into detail about this, next month, so for the moment, trust me on this. For the moment, we are going to add a "working cue", that will allow you to trigger the current version of the behavior.
Attaching the working cue:
Attaching a cue to a regularly occurring behavior is really easy. All you have to do it change the pattern a little bit. This time, just after you click and treat for a successful repetition, say "go" , "out"± or whatever cue you prefer, just before you think the dog is ready to go out to the target. Here's what the new pattern looks like ---
1. Wait until a pause occurs between the click and treat of one repetition and the start of another.
2. During the pause, say the command word in a normal tone of voice - well before the next repetition starts. (It is important to realize that you are attaching an initially meaningless word to a new behavior. For 20-50 repetitions, the word is not "causing" the behavior to happen. It is the learning of the pattern that will allow the dog to connect the cue to the behavior. The more consistent you are with the 20-50 repetitions, the faster the dog will learn the cue. )
3. The dog does the behavior.
4. Click and Treat
5. Repeat the sequence, starting at Step 1.
For those of you who are just starting out with clicker training, this process of adding the cue AFTER establishing the behavior may feel a little odd. For the time being, try to ignore the uncomfortable feeling and simply work through the exercises. Concentrate on making the pattern so precise that the dog will learn it quickly.
"Big T, Little t"
Once your dog will consistently "go-out" to the target, it is time to add left and right to the process. The easiest way to do this is to sit your dog in the same position you used for your "go-out" and change your position. If you imagine the shape of a big "T", you can easily visualize the new setting. The dog will be at either end of the cross bar - the choice of which one to start with is up to you. The target is on the other end to the cross bar, and you are at the base of the "T".
From the base of the "T", point generally in the direction of the target - but don't point directly at the target. Give your cue for "go-out". There may be a bit of confusion for a second, but if you have worked up to this gradually, the dog should blast over and touch the target. If she fails, drop your standards and get the behavior happening on a shorter cycle - like every 5 seconds, or so. Then try the new scenario again.
Once you have consistent performance on one end of the "T", move your position so that the dog is seated at the other end of the cross bar. It isn't necessary to move the target, just move the dog and yourself. Run this exercise about 20 times, so that you are getting consistent performance from the dog. Now that we have the movement to either side, it's time to start making it a little more complicated.
Now get a second target, and place one on each end of the cross bar. Sit the dog in the center of the cross bar and stand at the base of the "Big T". Use your arm signals and verbal cues to get the dog to go to either target. Remember to use "wrong" at the instant the dog heads in the wrong direction. When you are getting consistent responses, add a third target, about 10 feet past the center position. This will make our setup look like a "Little t", instead of a "Big T". Now you can send the dog away from you with your cue for "go-out". Practice this "round robin" until the dog's performance is smooth. You can add more arrows in order to increase the complexity and require more directed movements in between reinforcements.
Once you feel very confident that the dog knows the basic behaviors, you will need to know how to remove the targets. For some of you, this will be within a couple of training sessions, while others may choose to wait for several weeks. My best advice is don't be in a hurry to pull off the training wheels until the kid really knows how to ride the bike.
Pulling the target:
Until you can remove the targets the dog really doesn't know how to do the final behavior. The easiest way to accomplish this is to get the dog to make a simple mistake. First, start a practice session using your "Little t" setup. Make sure that the pace is pretty quick. After about ten successful movements, give the dog signals that places her at the far end of any of the arms of the "T." This is the moment of truth -- give the signal that will take the dog in a direction where there is no target. If the dog balks, say "Wrong" and go back to practicing, with special care to build up the speed and the number of targets the dog can bump, successfully.
Once the rate of performing is back up to speed, try the "setup" again. Move the dog to the far end of one of the arms of the "t" and give the signal that would take her to an area where there is no target. If the dog takes the tiniest step in the correct direction, click and treat - stop the session for a second and really whoop it up. Go back to your training diagram and slip in some of the "go to the target that isn't there" situations. If you stick with "Wrong" for turn-backs and clicks and treats for continuing in line, you can build a 10 yard go-out (with no targets) in a couple of sessions. Now remove all the targets and practice your "Big T - Little t" training without the targets.
Over a series of sessions, these exercises will give your dog the skills necessary to move quickly at your direction. One of the easiest ways to start using these directional signals is to place jumps in line with the targets - this will accustom the dog to taking obstacles after you give a direction signal. Once the dog is flawless with the process, it will be time to add a performance cue to each of these movements.