Teaching Dogs to Leave Odor

2 Apr

Japanese Chin -NoseworkI volunteered the second day for the NW1 trial. My job was videotaping Interiors. Interior was again held in a classroom, but in a more traditional classroom with tight rows of desks. The hide was under the seat of the desk on the outside front corner of the room, opposite the entrance.  I was SO very lucky that the judge (very experienced K9 Handlers/trainer) shared his observations with the volunteers in the room between most searches. Luckily Ren doesn’t have a problem leaving odor, but I thought some of these takeaways could be helpful in certain circumstances.

-Teach your dog a search pattern from the beginning so the dog learns to search the entire room.

– The dog’s indication should be clear enough so that during training a friend could recognize behavior and call the alert with handlers back turned

– Don’t ever LURE the dog to then source then reward. Dog should put their nose on source first, then they get the reward

– Have other people in the room reward the dog so they aren’t just always looks at handler

– Don’t pre-cue your dog that they are correct (by hand in pocket) if you think they are getting close. Wait for the final response and the “YES” before reaching for the treats.

Here was the big one though. Dog’s should never be rewarded for an indication on odor after they have left and come back.  Many dogs in the NW1 went to the odor, never indicated, quickly left, and may or may have not gone back. The judge said this was a “trained” behavior. That during training, probably during non-blind hides, that the handler was doing something to cue the dog when they dog close to the odor that wasn’t happening in the trial (because the handler didn’t know the dog was close) and because that dog wasn’t getting the subtle queue or assistance, the dog was moving on.

Think of this scenario – you are doing a non-blind hide in class, your dog is getting close, maybe your breathing changes or you start slightly bending over, or your fingers re-grip the treat. Your dog has learned this pattern as a “hint” that he is close and he should alert. But in a trial he isn’t getting these pre-cues and simply walks away from the hide.  Another habit many of us are in is rewarding the very moment the dog gets their nose odor. I have even heard about instruction to try and BEAT the dog to the hide with food (which doesn’t make sense to me at all). Again, in the trial, the dogs aren’t getting the reward at that very moment. They have to first alert, you have to call it, wait for the judges YES, and then reward, which is again a very different picture then training. When the dog doesn’t get the reward the moment his/her nose hits odor (what the are expecting),  they move on.Whippet -Nosework

The judge suggested varying the delivery time for the reinforcement – moments, seconds, then minutes. Regardless the dog should stay at the odor. After the dog is confident it its job, the reward comes more for staying at the odor vs finding the odor.



3 Responses to “Teaching Dogs to Leave Odor”

  1. Elizabeth April 5, 2013 at 9:03 pm #

    I like Andrew Ramsey’s alert training and teaching the dog to persist at source. He has a training video, and also lots of seminars on. YouTube. He also teaches a search pattern right away. It helped me a lot, I like his style and his results

  2. Leah Gangelhoff April 10, 2013 at 7:32 pm #

    Hi Tracy,

    Many NW judges may give feedback based on their own method of training selection-tested high drive dogs to detect explosives (in this case), narcotics, agriculture, etc. Many of these high drive dogs respond similarly to those methods because of the traits for which they are selected (ability to work independently, extreme hunt drive, extreme endurance, zero environmental sensitivity or fear, drive for a toy over ALL else, etc.). Professional, law enforcement, and military K9 trainers need to do this selection-testing as they need a dog that will do well with their methods and complete the training in the most efficient and reliable manner possible. They may be driven by contracts, time frames for “finishing” a dog, money, the need to testify in court (so training the dog to perform search patterns right away is critical since they might need to testify in court that the dog covered every inch of the area or they may need to show the court that they have a standardized training method across dogs), and/or the fact that people could die if a bomb isn’t detected the first time around.

    There are many reasons why these methods may NOT be the best for many of the pet dogs involved in the sport of Nose Work. The beauty of this sport is that it is open to all types of dogs with all types of temperaments, drive levels, past training experience (from none to a lot), and behavior issues (environmental sensitivities, fear, anxiety, dog reactivity, etc.). Dogs with lower hunt drive, no toy drive, or behavior issues such as those above would never be selected to train for professional, law enforcement, or military detection — because the methods employed would not be effective with those dogs, it would take too much time to see similar outcomes, and the dogs would not be reliable enough due to behavior issues that could affect them during any particular search. The purpose here with our pet dogs in my opinion is not about the “result” as much as it is about the dog having fun and experiencing success with the least amount of stress or pressure.

    Many pet dogs entering the activity do not come in with those traits that are selection-tested in working K9s – some of the traits can be developed and some may never be there. That drive for a toy over ALL else in selection-tested dogs will prevail in situations where a pet dog may be impeded by behavior struggles or revert to “obedience mode” and look to the owner/handler for assistance and/or offer previously trained behaviors like targeting or body positions such as sit, down, stare when not at odor. With pet dogs that do not have these selection-tested traits, and that have not yet developed enough “odor obedience” (the desire to find and locate the source of odor overrides everything else in the environment including the handler), it can be critical for the handler to step back and not bring attention to themselves early in training by “presenting” a pattern or areas to search or by introducing a trained response behavior (among other things), that could bring attention to the handler and cause the dog to try to give the handler what the dog thinks the handler wants. There is always fallout regardless of methodology, which could be what the judge is commenting on, and this fallout is addressed throughout the course of training. The judge may see more of what he is thinking is the “correct” picture at the K9 Nose Work National Invitationals this June. ;0)

    In addition, unlike professional, law enforcement, and military training where the dog is trained “to completion” before pairing dog and handler together, most times in Nose Work, the green handler and green dog are being trained simultaneously. An inexperienced handler cannot know without more experience how odor may be traveling through an environment and thus may think their dog isn’t searching in the “productive” area and may feel the need to direct the dog to a more “productive” area or “present” those perceived areas to the dog. This can serve to lower the importance of odor to the dog (if inadvertently pulled or directed off of following that odor), foster dependence on the handler (which you don’t want), and rob the dog of developing the skills they need to track that odor to source on their own. We want the dog to understand that the odor is the most important thing of all and not a particular desired behavior(s).

    There are ways a knowledgeable instructor can engineer the dog’s experience without directly interfering with the dog. A knowledgeable instructor who understands how to work with a wide variety of temperaments, behavior issues, and drive levels will also be able to quickly ascertain what the dog in front of them in THAT moment may need in order to organically develop odor obedience and the desired skills needed down the line rather than to apply a method to each dog in front of them. The founders of K9 Nose Work®, who have years of experience selecting and training high-drive dogs for professional and law enforcement detection, have spent the last seven years modifying those methods for the widest variety of pet dogs. It would seem that they have learned a thing or two during that time :o).

    That being said, there are many very high drive pet dogs that could potentially do well with the methods that particular trial judge advised. This is why many competitive dog sport enthusiasts begin to selection test dogs specifically for the sport in which they wish to compete – so that they will do well in competition.

    I love this activity for the reason that there isn’t undue pressure on the dogs to “perform” in a particular manner. We have the luxury (no one’s life is at risk) of allowing them to freely explore their amazing scenting abilities and hopefully feel safe doing so. Some may do well in competition, whereas others won’t, but their owners may still wish to participate regardless – for the fun of it!

    • wyattjrt April 10, 2013 at 9:10 pm #

      Makes sense – was just very interesting to hear another point of view and way train from someone with experience. During my 20 years of agility training I have attended all kinds of seminars from many different people who all handled things a bit differently. Was all very interesting learning different ways to do things. You have to note the title of this blog – My2cents 😉

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: