Teaching
the “Handler Proof” Find/Refind
By Deborah
Palman
II. To Train a “Handler Proof” alert behavior (applies to all types of searches or behaviors):
A. Remove the handler from the teaching process as much as possible.
B. Be aware of any cues that are created by the teaching process and remove them through further training. Eventually the only cue that should trigger the alert behavior should be the finding of the victim or substance the dog is trained to search for.
III. Behaviors are taught by:
A. Ignoring unwanted behavior.
B. Rewarding the behavior you want.
1. Be aware of inadvertent “rewards” or cues:
a. Eye contact, looking at dog, paying attention to dog.
b. Moving hands toward reward (s).
c. Moving toward dog.
d. Talking to the dog - commands, praise or otherwise.
2. Simple but hard for most dog owners to apply unless they put duct tape over their mouths and a paper bag over their heads.
C. Use “approximations” to create a finished behavior. Rarely will the dog do the entire behavior the trainer is looking for, so “approximations” of this behavior or attempts the dog to respond must be rewarded so the dog will keep trying. The behavioral criteria are then gradually raised to create a finished behavior.
IV. Steps for Teaching Find/Refind alert behavior for Air Scent SAR dogs:
A. Goal is to teach the dog to:
1. Find the victim.
2. Return to the handler, who is walking at any angle away from the victim and not paying attention to the dog (I like to have handlers pretend they are looking at their compass when they are walking, as this is what happens in searches).
3. Stop the handler’s forward progress and get the handler’s attention, communicating that they have found the victim through some obvious behavior (bark, jump, sit in front, body slam, whatever).
4. Lead the handler back to the victim.
B. Pick a dog for training that has the proper temperament and socialize him/her sufficiently.
C. Develop a good relationship with your dog.
1. Have a reinforcement history.
2. Avoid teaching with and using compulsion.
3. Dog should want to stay around you and pay attention to your location. When you take a walk in the woods, the dog should stay in your vicinity and not run away. If dog does not, you need to train this (rewards for “check in,” recall, etc.).
4. The dog should look to you for leadership and when he/she needs help.
5. Don’t over control or nag the dog about his/her behavior.
D. Develop a reward that can be transferred to a total stranger. Food, tug, toy, wrestling, heavy petting, whatever the dog likes that allows him/her to happily interact with a stranger. Do not rely on social attraction alone. Social attraction is quickly satisfied when the dog finds the person and does not last. It can also not be withheld or “timed” so that only a specific behavior is rewarded. A special reward must be developed.
E. Teach the dog a “soliciting” or demanding behavior to get the reward. If your dog is not over controlled and you have a good relationship, the dog probably already has ways to get your attention that you understand. Barking, jumping on you, nudging you - the behavior depends on the temperament of the dog and your relationship.
F. Do Runaways to start:
1. Have the training “victim” reward the dog until the dog is eager and happy.
2. Victim runs away and hides only a short distance away, just out of sight (less than 75 yards or so).
3. When victim is hidden, handler lets the dog go with a search command.
4. Handler quickly but quietly follows the dog into the victim.
5. Victim rewards the dog for at least 60 seconds or more. Handler praises during this.
6. Wind direction is not crucial - you are only teaching the alert behavior at this point, not the search. Always teach the alert before you start the dog searching, or the dog won’t know what to do when it finds the victim after searching. If the dog knows the alert behavior will bring the reward, then it will happily perform the behavior after finding the victim. Teaching the dog to “search” for the victim is not necessary if the dog is motivated to search by the anticipated reward.
7. Victim rewards the dog immediately when the dog arrives, starting with praise and attention and escalating to play, feeding when the handler arrives. Do not withhold the reward for the find/refind. Doing a bark alert here would be done by with holding and shaping the behavior.
8. Victim always rewards the dog so the dog expects the reward from the victim, not the handler. This focuses the dog’s attentions on the victim and helps to prevent the dog from short stopping the exercise when he/she returns to the handler, and helps to prevent “false” alerts where the dog may try to solicit a reward from the handler when he/she has not found the victim. Only in rare cases where the victim is unable to reward the dog should the handler reward the dog.
G. Do more runaways until you can see the dog knows the game and is eager to play. More may not be better here ... only do enough runaways so you can see the dog is eager and expects a reward from the victim. Doing too many may make it harder to teach the dog to return to the handler.
H. Do not do more than two or three runaways in a session. Too many bore the dog.
I. Next step is to set up a runaway around the edge of a building, with some cover near the start, or something not far from the start that blocks the view from the start to the victim. A building is ideal. Trees can also be used. Victim runs away and hides around the corner, only 50 -75 yards away in a place the dog can easily find them.
J. Handler lets the dog go and remains at the start, hidden from the sight of the dog when the dog finds the victim.
K. Dog finds the victim, and the victim should ignore the dog. Any rewards must be hidden. If the dog bothers the victim too much by jumping, biting at clothes, etc., the victim should discourage the dog mildly, saying no.
L. Handler and victim wait, ignoring the dog and not saying anything. If you wait long enough and the dog has a good relationship with the handler, the dog will return to the handler.
M. Handler doesn’t pay attention to the dog when it first comes in sight but tries to get the dog to come all the way back by crouching down, offering a food treat or looking at the ground.
1. The handler should not call the dog or talk to it, only use non-verbal signals without really appearing to pay attention to the dog.
2. In the ideal situation with a driven, mature and pushy dog, the handler can ignore the dog and the dog will come all the way back and demand the handler’s attention in some way.
3. Most dogs in training are not this strong because they are young or have relationship problems, so they may need some help, but this help must be minimal.
4. Acknowledge the dog’s first attempts to get attention by giving the dog your attention.
5. Keep praise to a minimum or don’t give any. Often praise at this point confuses or distracts the dog so they don’t remember to return to the victim.
6. Helps to have a second person there to watch and tell the handler what the dog is doing if the handler can’t help looking at the dog.
N. After the return to the handler, the handler waits for the dog to remember the victim and start back.
1. Handler follows dog, quickly but quietly.
2. The handler always follows the dog. The dog must lead. If the handler helps here, the dog will let the handler do the work and become dependent on it. Wait for the dog to proceed first, no matter how long it takes.
O. Big reward after the dog returns to the victim.
P. Repeat this “hidden handler” runaway until the dog consistently returns to the handler. Q. When the dog knows to return, the handler can escalate the ignoring phase when the dog returns and demand a more pronounced alert behavior.
1. For example, at first the handler may reward with his/her attention a whine if a bark at the handler alert is desired. Later the handler may not give attention until the dog barks.
2. Don’t be surprised if the dog
makes up his own attention getting behavior and modifies the exercise. Go
with the flow - the natural behaviors the dog offers are more likely to be used
in times of stress.
R. Repeat short runaways until the alert behavior is well set with the handler staying in one place. Runaways can be moved away from the visual barriers once the dog gets the idea. Different victims should be used if this has not already been done.
S. Do runaways and add motion to the handler. Standing still is a very important cue for the dog. When the handler starts walking away, the motion confuses the dog, so motion must be added very gradually to allow the dog to perform the alert behavior successfully.
1. The handler should move very slowly at first at right angles to the line of the dog’s return.
2. May have to just walk in place a first - motion causes most dogs to think they should follow the handler’s direction of travel.
3. Make sure the dog completes the alert behavior before returning to the victim, even if the handler has to stop moving when the dog arrives.
4. Handler adds more movement, faster and in different directions away from the victim, until the dog still performs the alert behavior even if the handler is moving away at a normal walking pace.
5. I consider the exercise finished if the handler can walk away from the victim looking at his or her compass, totally ignoring the dog. The dog should come back from the victim, stop the handler’s motion with the alert behavior and lead the handler back to the victim.
T. Don’t try to do “searches” (more than a runaway) before you have the alert behavior taught.
1. The dog won’t know what to do when it finds the victim and the handler won’t be able to teach it without a lot of commands and cues when the dog’s motivation level is lower.
2. We want the dog to “discover” how to get the hander’s attention and is/her reward without help from the handler. High motivation is needed for this, and high motivation will not exist after the dog runs for a time searching, unless the dog is already a trained and experienced dog.
U. Require the dog to perform the full alert behavior in training after progressing to longer searches. In actual searches, you will not do this, but if you don’t require it training, the dog’s alert behavior will fade.
V. Ignoring the dog is an excellent way to “proof” or test experienced dogs who know what they are doing. If you have not already done this with your experienced dog, be prepared for some amusing results.
V. Problem Solving:
A. Different dog personality types react different ways.
B. Handler/dog relationships affect training. Often this causes the dog to be reluctant to return to the handler.
1. Handler takes up an inviting position: crouches, pretends to have something good (toy, food), then, when the dog returns, the handler stops, stands and waits for the dog to remember the victim.
2. Handler holds out a piece of food for the dog to take (only one!!), and then follows the dog back to the victim. This can be changed to a hand holding food on the chest to get a jump, a sit, a bark, whatever.
C. Dog fails to return because the dog is trying to get the reward from the victim. Usually the dog is also bothering the victim, often for a toy. Wait this one out if you can.
1. Handler can keep the toy reward, present it to the dog when the dog returns. The dog will run to the victim to play. This causes the dog to be much quicker next time.
2. Handler stays in sight of the dog but ignores the dog and moves slowly away at an angle to the dog (so the motion is easy to see). This helps to bring the dog back. When the dog returns, the handler stops moving.
D. Dog stops part way back and stands, trying to make eye contact with the handler.
1. German Shepherd ploy to try to keep contact with both parties and make the handler do more work.
2. Ignore the dog until he/she comes all the way back and add a reward for the complete return, like a single piece of food.
E. Dog sniffs around and wanders near the victim or on the way back.
1. Try to ignore this unless it is apparent that the dog has lost all interest in the game. If the dog has lost all interest, go back to motivational work. If the dog still does not respond, you probably have a dog with poor working temperament or are using the wrong motivation.
2. The sniffing may be a ploy to get a reaction out of the handler or victim. If you react, the dog will be rewarded for the sniffing.
3. Ignoring the sniffing should cause the dog to speed up with repeated work.
4. Some dogs, especially males, just have to sniff some and you won’t get rid of it all.
5. Be aware of breed specific behaviors: for example, I was told by a German Shorthair Pointer breeder that GSP’s are bred to quarter. They rarely travel in straight lines. They run around and hunt. This makes them harder for us to read and understand in SAR training. Don’t expect them to travel straight lines.
6. Dog which aren’t let off lead much and need to explore and run need time before and after runaways to run around. If you take a dog who needs to run out of the kennel or car and off lead only when you do SAR exercises, you will not have a good training session. Let the dog out to run before, and make it clear to the dog that he/she is free to run after the reward is over.
F. Dog seems to forget all about the victim after he/she returns to the handler.
1. May be caused by the handler praising or rewarding too much on the return. Often seen in Golden Retrievers who, when you say “Good Dog” and pet them, go into an ecstasy of joy and forget everything that happened in the last 10 minutes. Tone down praise and running to the victim.
2. If cues were previously used in training (“show me,” “where is he,” handler moves toward the victim before the dog does), the dog may be waiting for these cues to respond. Wait it out. If the dog has done it before with the cues, the dog will do it again without them. Put duct tape on the handler’s mouth if needed.
G. Dog “overdoes” soliciting behavior on return to handler and won’t stop barking, jumping, whatever at the handler and return to the victim.
1. Have the victim carry all the rewards so the handler does not have anything to give the dog. The victim should always reward the dog.
2. The dog is doing this because he/she is being rewarded for it. This reward may be attention, talking or laughing by the handler, running by the handler, etc. Analyze what the handler is doing and remove this reward. Dog may anticipate a reward from the handler - just ignore the dog until he/she starts back towards the victim.
H. During searches, the dog gives the alert behavior when it has found fresh scent, but has not made contact with the victim and cannot not lead you back to the victim immediately.
1. Congratulations! Your dog understands his/her job and is trying to communicate with you!
2. Acknowledge the dog’s alert, but don’t reward until you reach the victim. Experienced dogs will do this in situations where they are prevented from reaching the victim. Not rewarding them here will not affect their work.
3. If the dog persists in alerting and insists on a reward, ignore the dog.
4. Be sure you know where the victim is and that the dog is not indicating on a victim you are not aware of.
5. An exception would be if you have taught the dog article search and the dog leads you to an article. Then you would reward for the article find.
6. Set up a different problem so you can reward the dog at the victim, if needed.