Tuesday, January 24, 2012

More about luring....

Someone asked an excellent question about luring and reinforcement and rather than answer privately I thought I would post it to the blog.

The question was if my observation was that people were incorrectly teaching the lure without teaching the command and they were then unable to fade the lure because the dog does not know the command. Good question because the big problem IS failing to fade the lure (if you use it) soon enough.  Luring itself is not automatically bad and can be used successfully.  In my response to the question though I have to admit that I am not too big on teaching voice commands. Not on purpose but I always end up teaching commands WAY after I teach actions. Take Amber for example - she is currently doing 4 weave pole sequences from a variety of positions and I have yet to teach her to officially follow a command to "weave". Something to remember tomorrow when I am training her as it is certainly something she needs to know. I don't think I have, as of yet, said to Steel "track" and I know that he recognizes no official command to find drugs. He does however know exactly what I want when I show him his tennis ball and clip him to his wide buckle collar (find narcotics). He also knows exactly when it means when I put his harness on and show him his frisbee (track). Obviously for certain exercises such as obedience the commands are necessary but I tend to let my body language give the dog information about what I want. Even for heeling if I stand up straight and place my left hand on my side Jill gives me pretty snappy heeling!

Getting back to the question though - luring is not bad. And of course commands are necessary and useful. Sometimes luring it is an excellent way to teach a behavior. However it can easily be a crutch to create the behavior long term. Again, the basic question is whether the food/toy/etc is creating the behavior (luring) or if it is being used to reinforce the behavior after it happens.  It is that later bit of clarification that is often missed. I often see people use luring to create the behavior and then think that the dog has been "taught". When in reality he may have just been following the lure and some body language cues. They then get tough on the dog for failing to comply. Or they give up, determine that the dog just doesn't like to work or isn't good at it. However an important step has been missed or neglected. Following fading the lure we need to reinforce the dog for getting it on his own.  The reinforcement rate is high at first and decreases with the dog's gain in proficiency. If the task is something complicated we need to reinforce approximations of the final behavior.

I hope this clarifies things a bit - or at least does not make them more confusing!

Saturday, January 21, 2012

Amber Update

What a difference a few months makes!

Here she is in October:


Cute but a little bit shy and insecure....

And here she is now!


Wild Girl!!
 She is still technically looking for the perfect home but in the meantime she is learning some agility and obedience and tracking. More importantly she is becoming very well adjusted to the world, a few things still scare her (like the automatic doors at Petsmart) but her progress has been amazing. I am used to obnoxious pushy dogs who tend to get in trouble rather than shy cautious dogs so it has been a learning experience for me. I have not forced her to "deal with" any of the things that concern her - I have just let her investigate the things that she is curious but cautious about. If she wants to avoid or leave I allow her to do so but in a calm cool and collected manner. We can retreat to a safe distance but not scurry away in a panic. It has worked amazingly well - we recently had a construction crew working on the front of the building and Amber ran the "gauntlet" of construction equipment several times a day. It was a great test of her adjustment and she trotted right through in her usual alert perky manner.

I will try to post more on where she is in her performance training but here are a few action shots in the meantime.

Cute and perky on my frisbee target!


Who says I don't have ball drive now!!??

I am coming!!

I hate this pic of me but it gives you a better idea of just how tiny she is!

Friday, January 20, 2012

Remember the Rottie Puppy?

The one who was dumped on the side of the road with her eyes barely open?
The one who passed through 4 different sets of hands to get to her new home in Florida?

Here she is now! I am thinking the "toy deprivation" needs to stop!!!

Tuesday, January 17, 2012

Luring and Reinforcing in dog training....

I have been slacking lately blogging because I have been spending everyday teaching dog training to students attending our Professional Dog Trainer's School. They are dog savvy people with either knowledge or experience or both. However we are teaching them a lot and they say they are learning a lot!

One of the things I noticed while training, and I see this quite often when working with others, is how many people have difficulty transitioning from luring the dog for behavior vs reinforcing a correct behavior. The basic definition of luring is using some sort of enticement to prompt a behavior. For example - we hold a cookie over the dog's nose and guide the nose/body into sit position. Is luring bad or evil - some will tell you it is and call it bribery. I prefer to think of it as a way to shape or teach the behavior that I want. Getting the dog's body into a place that I want to be. That said, I don't always use luring to train, it depends on what behavior I am trying to get.

The problem with luring is that it is easy to continue to use it, creating a dog that only responds to the prompt in front of his nose. This makes sense, while some dogs will anticipate the action or think you have the lure when you don't, we can't always expect this to be the case.  The problem that I often see is that people lure for too long. They then stop it altogether and wonder why they don't continue to get a correct behavior. They then discount rewarding the dog and start correcting the incorrect behavior.  The goal from the beginning should be to phase the luring out as soon as necessary (this will vary depending on your goals). An important transition, which is often shortcutted, is to phase out the luring and start reinforcing the behavior when it occurs. If you start reinforcing correct behavior or position at a high rate then you can decrease the rate and continue to see the behavior that you want.  Correctly done this can often be done quickly and the dog will stay engaged for a long period of time. However people often, because they have the prompt (food for example) still in their hands, continue to use it as a lure even though there intention is to use it for reinforcement. Been there and done that....

A simple distinction to ask yourself (at least it works for me) is:

am I using the food/toy/treat to create a behavior I want (luring)
or am I using it to reward a behavior I am seeing (reinforcement)

Both have their uses, we just need to clear which one we are doing and why.

Tomorrow I will introduce "TruckStop" the wanna be sitcom dog!!

Sunday, January 8, 2012

Habits are hard to break

For the past 10+ years my address has been PO Box 85. Now it is PO Box 574. I have moved a lot. In the past 10 years I have moved 5 times. Never has it been such a challenge to write down my new address. Three months later more than half the time I am writing PO Box 85 Bondville VT 05340. I don't just get the box number wrong, I get the whole thing wrong. Often.

Why? Because the first part, the "PO Box" triggers an automatic behavior. One I write PO Box I stop thinking and muscle memory takes over. I bet if my new address was something like 123 Main Street getting the new address correct would be less of a challenge.

Those who have trained with me, especially people working on changing a behavior in their dogs, have heard the following reminder:

"It is much easier to create a new behavior than it is to get the old behavior to stop."

Whether you are fixing an exercise for competition or trying to change a household manners behavior, it is often more effective if the new behavior is totally incompatible with the old behavior. At least at first. If we can break into that automatic, rote behavior that has been rehearsed over and over again before it starts then we are more likely to change or modify the event. Try to make the exercise or event have a whole new context rather than make it look similar to the old way. Remember that every time the exercise is done wrong the old way is being rehearsed and solidified. This is true often regardless of whether we reinforce the behavior or not.If you are trying to fix something try to arrange things so that the opportunity to rehearse the old behavior is not there.

I however, lacking that option, think I will just get pre-printed address labels. Maybe after a year or so I will have forgotten my old behavior/address. Maybe.....




Thursday, January 5, 2012

The advantages of letting the dog teach himself

My being involved in a wide variety of dog training venues - obedience agility police dog training SAR dog training tracking etc etc gives me the opportunity to watch a wide variety of trainers demonstrate and explain techniques. Sometimes I learn something new however most of the time I just learn ways to tweak what I am already doing to make it more effective. A lot of the time however I see stuff that makes me roll my eyes. Most of the time it is because people seem to think that dogs are stupid they are grossly underestimate how little the dog needs their help while learning. This is true for people who utilize negative training methods but, and this is always a surprise to me, it seems to be even more common with positive training methods.  In order for the dog to learn the task needs to be broken down into progressive steps for the dog to learn. However we also have to have patience and shut up and give the dog a chance to figure things out on their own.

One reason for this is that things that the dog self teaches appear to be more "hard wired". I can give a cue and hand guide the dog through an exercise many many times and they will learn. However if I manipulate the environment in some manner so that the dog offers the behavior and I reward it, my training time is much much less. The light bulb just seems to go on faster. The dog also learns to perform independently of me right from the beginning without some body language cue or encouragement or prompting also being necessary in addition to the command. The dog ignores me and just offers the behavior because I have never been a source of help. Or at least not that he would notice - one of my most repeated mantras is "if you are going to help the dog don't get caught helping". In most of my training I don't want the dog to know I have the answer to the problem. If I have made the exercise too hard then I need to make it simpler in a way he does not notice. I do NOT step in a show him or help him, especially while he is still working and trying to solve the puzzle.

This brings up my most important reason for letting the dog self teach:

Because it is more fun for the dog!!

Dogs love puzzles. Not only are they capable of solving complex problems on their own, many of them like to and are pleased with themselves when they do. Of course, just as with people, this varies between individuals but a lot of dogs are never given the chance. With our egos, we think that all the information needs to come from us and don't realize how much problem solving and self teaching the dogs can do in the right environment and with the right motivation.

I took these clips of Steel just prior to moving. I regret that I didn't make more of an effort to get better quality and more detailed video of him doing this - rather than sitting there giggling at him. There was no training involved in this, just me at first leaving his beloved planet earth ball on top of his kennel. Then of course I started putting it there just to watch the show.  The interesting part was not the accomplishment, it was watching him experiment with behaviors and then watch what happened. He quickly learned to push it to the right rather than to the left! Then later on he determined just how far to the right it had to go before he could reach it from the outside.

During the learning process....



"I got this all figured out...."






Sunday, January 1, 2012

Why I run....

I run to stay healthy and to manage my weight. Every once in a while I enter a race. To call them "races" is a little bit inaccurate however since to most people a "race" implies trying to run faster than other people. That would never happen in my case. I am a slow and methodical runner. Entering a race just gives me a training focus and an added motivation to get myself out and running regularly. Jill also helps with that.

I saw this on facebook recently and I just about hurt myself laughing. So totally me! Even Jill speeds up a little bit when faster runners pass us. I know she is wishing for a faster partner.


Being the first day of the year I felt obligated to get in a long run today. It didn't hurt that it was beautiful sunny, low 60's with a breeze through the trees while I ran. Lots of people were exercising probably for the same reasons. When I run my mind usually wanders in many directions.  Rarely does my schedule or my to do list enter my head, usually just random thoughts. Being New Year's Day and the first day of 2012 my thoughts were more philosophical today.

This area is relatively flat. Especially compared to Vermont. I always found training program that called for "hill days" as humorous. For a Vermont runner every day is hill day! The bike path that I run has a few places with gradual rolling hills. On the out portion of my out-and-back route I found myself grinding my way up a long hill. On the way up the hill I was wishing I had eaten fewer cookies over Christmas. I was also telling myself how nice it would be when I came back and got to run down that rolling hill.

However on the way back I realized that I had not noticed the easier portion of the run. Then I got to thinking about dog training (what a surprise right?!) and life in general. We notice when life is a struggle and when things are not going right. But during the moments when things are going well we don't notice it and blow right through it, focused on our goal.  I think I am a fair and patient trainer, experience has taught me how much more effective and enjoyable it is to be so. However the scientist in me tends to focus probably too much on a goal and over analyze my dog's performance and especially my own performance. Every year a large group of us that belong to an online dog forum post our dog goals for the year. Usually sports related goals involving getting certain conformation and performance titles with their dogs. There is nothing wrong with that in itself and I also have goals for what I want to accomplish. However more importantly, here are my goals for the year:

To try to do a better job noticing the downhill, easy portions of running, dog training and life. I'll see how I do.....

HAPPY NEW YEAR!