Saturday, October 1, 2011

Motivation

 I spend a lot of time thinking about motivation for the creatures that surround me.  If I can tap into that motivation helps make a willing partner instead of a resentful puppet during the training process. 

There's different "layers" of motivation.  It can be very easy to identify a motivation such as "will work for treats, or really anything edible", but there can be other, less obvious motivators that once recognized can help you teach a skill in a way that is efficient and works for that animal.

Recognizing how my animal differs in motivation from the "demonstration animal" will allow me to modify training ideas and regimes to help me be successful.

So.....let's talk about Tess.  (of course we are talking about Tess - it's her blog!)

Here the motivation "profile" I have identified for Tess.  This will also give my blog followers insight into the kind of puppy I have and provide a starting point for any advice you might have!!!! :)

Motivators

  • Tess is food motivated.  Very food motivated. Will work for food - really anything.  Even kibble sufficient for most tasks.....
  • Second to food only is her love of playing "tug".  She ADORES it.  I'm sorta "eh" about it - but I recognize that she will go through elaborate rituals and literally jump through hoops to get me to play tug, so I humor her once in a while.  In the beginning I watched her carefully, since I had heard that dominance can be an issue with dogs that tug that have been allowed to win a lot - but I don't see that in her.  She plays tug with me a lot like she plays tug with her best friend, Reed the Golden Retriever.  Sometimes he wins, sometimes she does - but no matter who wins, the toy is always placed back into "play" really quickly. 
  • Motion motivated/fixated.  To get her attention quickly, move it. She's a pointer, and bred to hunt.  This is actually causing more problems than being useful right now.  Butterflies and birds make it almost impossible to keep her attention in large grassy fields. 
  • Likes to do a "good job" - You can actually see her how pleased with herself she is when she's taught herself a new trick (catching a ball), or when she's figured out an answer to a question or a new way to do something.  Letting her figure something out actually gives her greater pleasure than me telling her what to do. 
  • Speed - she loves to RUN.  Chase is actually another favorite game - something I'm fine with when I am the quarry - and I exploit that shamelessly during recall games.  But something I try to avoid playing in the other direction if you know what I mean.....

Knowing her motivation help me to identify her preference for activities.  Preferences are both clues to making the training efficient......AND identifies areas that are inherent weaknesses.  Below are some of Tess's preferences, and how I use them to my advantage.....and some of the training issues we have based on those preferences.

Preferences

  • Tess consistently chooses "fun" and challenge over straightforward.  I think this is related to that "good job" concept above.  For example - when running around the yard, she rarely just runs - she would rather weave in and out under tables and chairs and leap over hydrants.  During training, I have to keep stuff fast paced and "multi-level".  She's quickly bored and impatient.  It's easy to push her too fast when we don't really have the foundations of the exercise down.  She'll do the task as long as it's new and exciting - but loses that consistency once she feels the game has become "hohum" and we should be moving on. 
  • Choses activity over relaxation.  If something makes her run really fast and jump really high with a large output of energy.....that's what she chooses.  
  • Choses the bold and independent path and makes quick assessments.  This is good because I rarely deal with a nervous or unsure pup.  BUT, keeping her attention (motion fixated) and keeping her close (because she's so bold) and keeping her interested in what *I* want to do, instead of her little independent self exploring the world....has been quite *interesting*.  
  • She's *highly* opinionated and *highly* creative.  She's quick to develop opinions and if I'm not observant enough to see what's going on, and then I have a lot of work to "undo".  She's convinced that while I'm clicker training her, she's actually training me and is quick to offer shortcuts and novel solutions.  *sigh*  And yes, sometimes I'm even duped into rewarding it with a treat before I know what I'm doing.  My brain is clicker stupid. 

Another example - the other dogs in the household

Last night, just for entertainment, I did some introductory clicker training with the other two dogs in the household - a 1 yr old Golden Retriever, and a 3 yr old German Shepherd.   The games I played and my process for "teaching" the games varied based on their very individual personality.  Reed (Golden) is lazy, craves physical contact, and lives to eat.  We played "Yer choice", an impulse game without the clicker, and I required him to sit for the reward (he's so lazy he lays down for everything).  It was relaxed and slow paced.  Harley (GSD) is anxious, craves attention (but not physical contact), restless, not particularly treat oriented, SUPER motivated by a "job", and thinks and offers incredible fast.  For her, I used a clicker and taught her to target my hand.  I needed the clicker in order to capture her thought because she moves from thought to thought so quickly.  The clicker also gave her approval in a non-physical way - but still gave her attention, and let her know she had "done good".

Each game was chosen with the dogs natural perferences and motivations in mind.  In time, both dogs will learn both games.....but I tried to set each of them up for success.

These principles are not just for dogs - I can apply these same principles to my mare Farley - and even to my cats.  Jonah (a cat we believe is probably a Siberian Forest Cat) has motivators that are more similar to most dogs, than to cats, and as a result it was easy to train him to come and sit.

If you can figure out what motivates the animal, with some imagination you can train it!

Further thoughts

The definition of a motivator is as broad as your imagination.  I'm learning from my classmates that have substantial experience in rehabilitating animals, that when considering motivation it's important not to project your own needs and feelings on the animal and try to evaluate them as objectively as possible.  For example, consider a rescue dog that has been abused (I use that term very generally) and has learned that attention = negativity.  While the human may see a poor frightened animal that they want to scoop up and reassure and give affection, the animal may want nothing more than to be left alone.  He is motivated by the need/want to be alone.  The kindest thing may NOT be to sit down next to the dog and provide reassuring pats - turning your back and ignoring the animal may be the best beginning.  The process and balance of working within their motivation (and perhaps over time gradually introducing the concept that affection can be a positive motivator) is part of the training (and rehabilitation process).

Thursday, September 29, 2011

Crate game!

I find that starting a blog and writing a welcome post is a bit awkward.

I'm rereading and categorizing posts from my "other blog" in conjunction with a project (see link under the header of this blog if you are interested) and I wince at some of my first posts. It was so hard, knowing no one was reading (and secretly relieved), and feeling confused where I should start.  Should I do catch up?  Just start in the middle?  How much backstory is needed?

For this blog, I took a different approach - I started writing posts where I am.  Today.  Right now.  Even though Tess and I have been doing foundation training for less than a week, I didn't try and go "backwards",  and figured I'd fill in the blanks as needed as we went along. Consider today a "fill in the blank!" :)

I've talked about crate games several times and maybe you are thinking "what the heck is a crate game?"

What I thought a crate game was - 
I said "kennel kennel kennel kennel" over and over and over, while directing the dog into the kennel.  Finish with a treat.  Usually do this only when you need to put the dog into the crate and are going to shut the door.  This was based on forming a (eventual) conditioned response to the word, associated the word with the action, and training as part of the daily routine and not necessarily spending a lot of time training the kennel as a tool.

Those of you that have been in this game longer than me (all of you....BTW) are hysterically laughing.  Those of you that grew up like me are scratching their heads and saying "what's the problem".

After 4 months:
  • Tess still wouldn't go into the crate on her own
  • She rarely "hung out" in it.
  • It remained a place that she would dart into to get toys out of - nothing more.
  • The one and only positive (but a big one) - Tess was quiet and obedient in her crate

What the crate game ACTUALLY is (as explained to me by W):

 The crate game involves no speaking, no gesturing, no clicking.  The point is to teach the dog that the crate is a "high value" place and a place that the dog wants to CHOSE to go.  It's their idea and choice to go into the kennel, and when they make that choice, they are rewarded.

Mechanics of the game 

With the Kennel door closed, wait until the dog sits.  Say "Yes Tess", open door and reward (or reward through the top of a wire crate).  Dog should remain in position.  If the dog moves forward or moves from the sit, close the door and wait.  Don't say anything - just close the door and wait.  Repeat until dog will sit with the door open.  Reward with the high value treats!

Next, with the door open, say "OK" (or release word).  Feed a piece of kibble (low value treat).  In my case, I didn't praise Tess - her encouragement/desire to come OUT of the crate was high enough to be a reward unto itself when combined with the kibble.  Then wait.  Eventually the dog wanders back into the crate....and you praise the heck out of them verbally and give them a high value treat.  Then shut the door and restart the game. 

At first you stand really close to the kennel door when you give the "OK", and perhaps they have a leash on.  Gradually move further and further back, eventually dissapearing behind a corner, or walking around the room.

As the last "repetition" of the game for the session, say OK, but don't give kibble - just stand there.  The dog should figure out that the answer is to go back to the kennel.  When they do go back, PRAISE them and give a high value treat. 

As the value of the crate goes up, the pup should get faster and faster about returning to the crate.  Eventually, start saying the word that describes the behavior of going back to the crate (kennel, crate etc.). 

Foundations established

I LOVE this game.  It teaches so much all at the same time, and I don't have to do anything - I don't even have to talk. 

1.  The sit stay - dog sits with the kennel door  open as you run around and even go out of sight.
2.  The come and release - the dog is still getting rewarded when they come to you.
3.  Introduction of "asking a question" and letting the dog come up with an "answer"
4.  Teaches the handler self control and NOT to spoon-feed the dog answers!  Harder than it sounds.....as the puppy sits in front of you expectantly and you know there is NO WAY they are going to figure out to go back into the crate......LOL. 
5.  Teaches the dog obedience, focus, and restraint in the absence of voice or physical controls


Game Notes -
  • Kennel reward is "high value" - hot dog, etc.
  • "OK"/coming out of the crate reward is "low value" - kibble. 
  • It's OK to look at the kennel when waiting for the dog to go in.  That is the only signal I give. 
  • No talking except to praise them when they are in their crate. 
  • When you do start introducing the word, say it ONCE (harder than it sounds....).  You want the behavior established BEFORE introducing the word, because using the word too soon potentially end up having a negative connotation and being ignored while teaching the behavior....TEACH the behavior before you associate the word! 
  • Having them sit in the BACK of the kennel is the goal, and once you have a good sit, it's OK to wait with the door closed and see if they offer to back up - REWARD!!!!  if they do.  



What our game looks like at this point.....
  • The sit with the door open at the kennel - pretty darn solid.  Sit stays have always been a strong point for her, so I'm not suprised.  We still have an average of 1 break per session, so I'll scale back our practice (stay closer to the kennel) for a couple of sessions with the goal of ZERO breaks and then try it from "afar" again.  
  • The release - Perfect.  She comes running to me, wherever I am for that ittybitty piece of kibble
  • Going back to the kennel - could be faster and snappier, definitely still gets distracted.  She seems to have picked a "pace" to go back and that would be a trotting "saunter".  W assures me that as the value of the crate increases, so will the speed....so I'm trying to be patient and not worry about it.
  • Sitting when I go up to the door - good!  After starting clicker training in other areas, she's been offering other behaviors when I don't open the door - such as getting further "back" in the kennel.  She is offering some "frustration" behaviors (rubbing her nose into the side of the crate etc.) so will continue to work on this by waiting it out and insisting that she sit in the back. 
  • Have started combining with the impulse game by leaving treats and food she has to step over as she goes in and out of the kennel.  Does "OK"  and continues to improve. 

As I learn more I'm sure I'll be modifying the game - if you have suggestions, please offer.  I don't have a good youtube video for this game, but I would encourage you to check out Susan Garrett's resources for this game, her website can be found on the "Resources" Page.

Tuesday, September 27, 2011

Behavior shaping - it really works

After working in the afternoon on clicker training and rewarding for Tess offering behavior, Tess actually started to offer behavior in the evening session during one of our "regular" games (crate game).

Normally, when I wait to give the treat in the crate to see if she will offer an improvement, or "try harder" for the treat, she just stares at me.

Stares at me with such concentration with those green hazel eyes that I feel compelled to start giving her the answer to the question.

But I don't.

At least, most of the time I can practice enough self control not to.

So I just stare back.

Sometimes I raise an eyebrow.

Very slowly, she might try something from her obedience repertoire......like s-l-o-w-l-y laying down.  As if to say "you haven't given me permission to do anything differently than what I'm doing so I'm not sure whether I'm even allowed to try something new?"

But laying down isn't the answer either.

In fact, the answer is to sit further back in the crate.

So, up to this point, in games like the crate game, she has slow to offer new behavior or modifications to her behavior.

After working with W and a cardboard box, as well as some other behavior molding tasks, it's like a switch went on in Tess's brain.

"If i don't get a treat within a few seconds and I know I'm suppose to be doing something, try something DIFFERENT".

Now in the crate game, when I delay the treat because she's sitting too close to the door, she's quick to offer all sorts of things - a down, a bark, a paw..... - and more than likely a scoot backwards into a sit.

Bingo!

A common criticism of clicker training is that if you don't have a clicker and treats with you, the dog won't offer the same behaviors that are being trained for.   But, keep in mind the crate game doesn't include a clicker, or any voice commands at all, yet using the clicker for a complete different exercise improved it!

Depending on what behavior or purpose you are training for, you use more or less tools.  Your voice is a tool, your body language is a tool, your eyes are a tool.  Treats are a tool, a clicker is a tool.  A leash is a tool, a collar is a tool, a head collar is a tool.  Do you see my point?  Relying on any one tool is foolish - whether it's your voice, your hand gesture, the clicker, or the treat.  Each has their place.  From what I've observed over the last couple of days, the treat and clicker are especially helpful in motivating a dog to offer all sorts of behaviors (some good, some not so good) and capturing that behavior in attempt to eventually "train" that behavior.

What did treats and rewarding for offered behavior earlier in the day gain me?  A dog who isn't afraid to try new things.  For 4 months she's gotten in trouble for doing anything other than exactly what she was told, especially because she is SO bold and independent - Sit, down, stay, come, leave it.  However, when I get into "real" agility with obstacles, I can't suddenly expect her to offer behaviors in conjunction with them, nor can I necessarily train an obstacle the same way I taught her to sit - I need her active participation.  Part of the foundation training I am doing now is teaching her how to THINK like an agility dog.  I need her to offer behaviors, and I need a way of capturing and rewarding the behaviors - thus the clicker and the treats. 

Without treats, my 6 month old puppy would be distracted and completely unmotivated to switch her attention from butterflies to me.  So distracted in fact, that a minimum of training would be done and even less learning.  A learned behavior doesn't rely on treats, but training goes a lot faster with less frustration and it's more enjoyable if treats are used. 

I haven't always believed in using treats to train, but the older I get, the more I realize that life is too short for me to spend it being frustrated or stressed.  Tess's life is even shorter and for the ten or so years that we get to share, I'm going to try and make that as positive and fun as possible for both of us. 

Monday, September 26, 2011

Crate Games & Unintended Consequences

Tess and I have been working on crate games...and I managed to create a monster.

The crate has never been Tess's friend.  It didn't matter that I was careful not use it as punishment, or that I associated it with treats, Tess still regarded it with suspicion and mistrust. 

That isn't to say that she's disobedient in a crate.  She sleeps in a crate, settles, and is a model citizen crated in the homeroom.

But it still wasn't her choice to be in her crate and so never volunteered to go in on her own time.

One of the first foundation games I introduced was the "crate game".  Simply put, she gets a low-value treat for coming to me outside of the crate (a piece of kibble), and a very high value crate (piece of hot dog) for going into the crate.....all on her own.  I don't tell her to go in or direct her in.  She decides that she would like a piece of hot dog, and so she decides she would like to be in her crate.  There's no nagging, coercing, bribing or luring - just a puppy who figures out the answer to the question "how do I get that hotdog?" and CHOOSES a behavior.  I'm quickly figuring out that this is a dog that loves a choice.  Force the issue and she goes along grudgingly - make her think it was her idea and there's nothing better that she would like to do, than to march into her crate.

Tess started popping into the crate when we weren't playing the game.  "Great!", I thought, "the crate is her happy place".....and I would give her a treat.

Oh yeah.   You can see where this is going.

In a matter of days, Tess now begs for treats by getting her kennel and waiting.  Some of the time she gets the treat, and sometimes she doesn't, but she likes the odds well enough to try.  As soon as she gets a treat (or doesn't) she pops out of the crate, only to try a little while later.

While the behavior isn't inherently BAD, it misses the point of the crate game and what I want her to learn about the crate - that it is a safe and fun place to hang out, play with toys, and take a nap.  I want her to ENJOY being in the crate for its ownsake - not because she thinks she might get a tasty treat.  The treat was there in the beginning to help mold the behavior - but it's important to keep in mind what the end goal/behavior is and realize when MY behavior is sabatoging our efforts.

I'm very new to the concept of shaping behavior, training with treats, and the philosophy of rewarding for offered behavior.  It's quite a switch from the conditioned response/command based training I'm familiar with.  It's so easy to reinforce the wrong behavior with any method of training, and I'm more likely to do so now, because I'm new to the training method.  So, I'm being very diligent about spotting "unintended consequences" before they become a horrendous problem.

The solution to the "crate for treats" dilemma?  When not playing the crate game, she'll get treats for being in the crate ONLY if she's hanging out "for real" - ie chilling or napping.  Thanks W!

***W is a classmate and dog trainer that is working with me and Tess. 


Sunday, September 25, 2011

Dog 'n Jog Event - Oct 15, 2011

If you live in the Davis, CA area, two UC Davis School of Vet Med clubs are putting on a "Dog n' Jog" as a fundraiser and public information event. If you love organized races, but are sadden by the fact that most require to leave Fido at home, this is the event for you.

Click here for more information and entry.