Lightbulb moments

What you describe with River is pretty common for dogs new to clicker training. Conrad was like that originally, and it took quite a while to get him more animated and willing to experiment. Even for dogs trained primarily with lure/reward and few corrections, shaping is still a new paradigm and it take a while for a dog trained another way to understand what you're asking of him. I use a crossword puzzle to help me remember to not just stare the dog down when he comes and sits in front of me expectently. Once I start on the puzzle, that communicates to the dog, "Oh, that's not actually what she wanted me to do because she's not even looking at me, guess I'll try something else instead." I don't really need the puzzle any more, the dogs both now get what I want from them, but I employed it for the first few months. I just have to stay understanding that for 4 years all I wanted Conrad to do was sit still and not move and listen to what I told him. Now he's got to do the thinking for himself and it was a lot to get used to at first.

I would let River figure out for himself that you want him to grab the tug. It's really tempting to try to force these things but patience is really a virtue with clicker training. If he's a normal dog (mine aren't--they never pick anything up in their mouths and we're still working on this one) then he'll grab it once he realizes for really real that you want him to engage with it in as many ways as he can think up. You might try waving it around, making it fun and interesting, but I wouldn't just "open mouth, insert tug" so to speak. That doesn't really teach anything related to what you want him to do, it just teaches him to be passive while you stick stuff in his mouth.

Always treat when you click. If it's a bum click, don't sweat it, just treat and move on. If you c/t something that you didn't mean to, that's not a huge deal. Just don't do it again and the dog will figure out right quick that that's not what you wanted, and will move on to something else.

ORIGINAL: houndlove

What you describe with River is pretty common for dogs new to clicker training. Conrad was like that originally, and it took quite a while to get him more animated and willing to experiment. Even for dogs trained primarily with lure/reward and few corrections, shaping is still a new paradigm and it take a while for a dog trained another way to understand what you're asking of him. I use a crossword puzzle to help me remember to not just stare the dog down when he comes and sits in front of me expectently. Once I start on the puzzle, that communicates to the dog, "Oh, that's not actually what she wanted me to do because she's not even looking at me, guess I'll try something else instead." I don't really need the puzzle any more, the dogs both now get what I want from them, but I employed it for the first few months. I just have to stay understanding that for 4 years all I wanted Conrad to do was sit still and not move and listen to what I told him. Now he's got to do the thinking for himself and it was a lot to get used to at first.

I would let River figure out for himself that you want him to grab the tug. It's really tempting to try to force these things but patience is really a virtue with clicker training. If he's a normal dog (mine aren't--they never pick anything up in their mouths and we're still working on this one) then he'll grab it once he realizes for really real that you want him to engage with it in as many ways as he can think up. You might try waving it around, making it fun and interesting, but I wouldn't just "open mouth, insert tug" so to speak. That doesn't really teach anything related to what you want him to do, it just teaches him to be passive while you stick stuff in his mouth.

Always treat when you click. If it's a bum click, don't sweat it, just treat and move on. If you c/t something that you didn't mean to, that's not a huge deal. Just don't do it again and the dog will figure out right quick that that's not what you wanted, and will move on to something else.

Thank you houndlove, just the encouragement and reinforcement I needed.[]  One more question, I wouldn't need to pry open his mouth, if I were to hold the rope bone in my hand he would grab it with his mouth right away, so do you think I should do that once or twice and CT or still just let him figure it out on his on?

Thanks for the support.

Oh yeah, if he just grabs stuff with his mouth when you hold it, by all means c/t for that, then start moving it to the ground and let him make the leap from "I take things from her hand" to "I take things off the floor". With my guys I would have to like pry open their mouths because they are really not mouth-oriented dogs (they're more pawsy), but if River will readily take stuff, so much the better!

Enjoy!

And, anyone who owns herding dogs knows that they can solve problems, too.

I wonder if any of the researchers thought to ask what type of previous training those dogs in the experiment had.

Absolutely. We have to be careful to leave the minds of our dogs free if we want dogs that will "problem solve" around the farm instead of going into cruise control or reverting to some lowest common denominator behavior when they are left with no input.

I once watched my late trainer take a packet of sheep through a crowd (no fences). He was on a golf cart, his dog was behind the sheep just making sure they continued to follow him. He needed to get them across his farm during a busy event, but he wasn't in any hurry. He stopped several times and chatted with people in the crowd, once about 20 minutes. He never looked at the dog once or gave him a single command, though the sheep kept making casual efforts to drift off, and once were startled away outright by a passing ATV. The dog simply took care of the sheep, leaving his master free to have a good time with his friends.

I think sometimes we sell dogs really short. I know I've got big giant trust issues in reality though intellectually I know I'll never progress until I get over it.[] Clicker training has helped ME progress there, as has the mentorship of some very wise people.

ORIGINAL: houndlove

Oh yeah, if he just grabs stuff with his mouth when you hold it, by all means c/t for that, then start moving it to the ground and let him make the leap from "I take things from her hand" to "I take things off the floor". With my guys I would have to like pry open their mouths because they are really not mouth-oriented dogs (they're more pawsy), but if River will readily take stuff, so much the better!

You might try a toy that you can stuff treats inside of.  I did this with Tasha to get her to take something from my hand.  Then I'd opent he toy and give her the treat.  I then progressed to stuff on a couch or up against another object to make her have to pick it up to do anything with it.
http://forum.dog.com/asp/tm.asp?m=266491
Go here and watch the video.  Tasha would not pick up anything unless she wanted to shred it.  I've been working on this with her for months.  We got stuck at the pick up off the ground part.  She would take a toy from my hand and get a c/t.  She would pick a toy up from the ground if it was on something or up against the couch where she couldn't just lay down with it.  She would get c/t for that.  Every now and then, I would toss the toy on the floor and see what she did.  If she layed down, I'd pick the toy up and toss it away from her.  She knew that interacting with the toy was the idea, so she'd walk over to it.  If the toy was near me, I'd toss the treat away to keep her on her feet.  If the toy was away from me, I'd toss the treat on the ground near my feet so she'd have to come over to get the treat, then go back to the toy.  One day, she just picked it up!!  From the ground that is.  So she got a huge c/t for that.  It's progressed pretty well since then!

ORIGINAL: luvmyswissy


I think River is getting close to a light bulb moment too.   I have some questions too.  He is rarely geta excited, like the dogs in the videos I have watched, and his breed is know for getting bored quickly so I think this is part of it.  I started with one of his rope bones and held it in the air, when he touched it with his nose I would CT, I did this a couple of times as a warm up.  Then I threw it on the floor and every time he looked  towards it or walked near it I would CT, this exercise when on for a while because it took him a while to realize I would reward him for looking in that direction – then he would come and sit in front of me and wait.  Then I started tossing the treats on the floor to try and keep him from coming right back to me.  So he finally started to touch it with his foot, and I would CT.  He keeps coming back in front of me and sitting and staring.  While I ignore him he will finally go to the rope toy and touch it with his foot.  Now after two nights he is touching with his foot more often, but not quickly or excitingly but he is getting that that's what him will win him his reward.  But still, no excitement just bewilderment kind of.  Then I waited tonight for his nose to touch it and he picked up on that quicker that the foot touching – so I know he is getting it.  My goal is for him to pick it up eventually.  Will he just realize that or should I show him by putting it in his mouth a couple of times and CT to show him what I want or should I just throw it on the floor from the start and go through the touching figures it out on his own?
Also, sometimes I click to prematurely or to late, do I still reward?  Or pretend I didn't happen?

If you mess up, still reward the dog.  One poor response will not keep him from learning, but you need to insure that the dog knows that the click always predicts a treat.
If you are able to click 8-10 correct responses within 60 seconds, he is "getting it".  If not, you should lower your criteria a bit.  For example, lets say you have a CD with a bullseye painted on it in your hand and you are asking the dog to touch it with his nose.  He does that, and gets good at it, so you raise your criteria - now he needs to touch the inside circle.  He does that.  Next, you ask him to touch the little hole in the middle.  For some reason, he doesn't get enough reps there.  So, you reduce your criteria to touching the inside circle, and let him be successful at that.  You might then wait until he touches the hole once, even by accident, and jackpot for that. 

If you mess up, still reward the dog. One poor response will not keep him from learning, but you need to insure that the dog knows that the click always predicts a treat.
If you are able to click 8-10 correct responses within 60 seconds, he is "getting it". If not, you should lower your criteria a bit. For example, lets say you have a CD with a bullseye painted on it in your hand and you are asking the dog to touch it with his nose. He does that, and gets good at it, so you raise your criteria - now he needs to touch the inside circle. He does that. Next, you ask him to touch the little hole in the middle. For some reason, he doesn't get enough reps there. So, you reduce your criteria to touching the inside circle, and let him be successful at that. You might then wait until he touches the hole once, even by accident, and jackpot for that.

 
Thanks Anne!  I maybe moving to fast (I get bored easy..[8|]), thanks for the visual.

I wonder if any of the researchers thought to ask what type of previous training those dogs in the experiment had.

 
Excellent question. I wish I had thought of that. I would like to see the experiment retried with clicker trained dogs, i.e., dogs trained to problem solve. If I remember correctly, the dog was not a herding breed such as an Aussie or one of the american cattledogs.
 

Even for dogs trained primarily with lure/reward and few corrections, shaping is still a new paradigm and it take a while for a dog trained another way to understand what you're asking of him.

 
That's why I have patience.

I live in the NY area theres nothing slow paced about me![]  Patience is a virtue..

I need to get more reps before I move onto something else and I am also going to use more yummy treats and see if I can't some excitement out of the big guy.[]

I'm trying the shaping thing this morning, where I'm not commanding a certain obedience. But he has made the connection with the clicker. He knows I have treats in my hand but he also looks at the clicker, expecting a click then a treat. I also decided to shape something easy that he could come up with. While sitting on his haunches, he placed his paws on me. C & T. It's not important whether that becomes a neat trick or not. But it is a trick he will have taught himself and I can see if I can link it to a word and later, use the word to get the move.

So, I finally understand the reasoning behind the beginning of C & T. First, connection of the click sound and getting a reward. Click equals treat but it also equals the exact move that you wanted. But the point of the beginning where you click for any ole thing is not hap-hazard as it may appear. The point is to help him realize that he can train himself. That he can problem solve. I also see the viewpoint of wanting a dog that obeys because they want to, because it is something they discovered that is rewardable. It is, IMHO, quite close to wolves in a pack that submit readily to the alpha without a struggle. They "submit" because they want to. They motivate themselves to fit in. Some structure is resolved through conflict, but the rest of the pack will fall in line on their own, though there may be fluid ranking depending on context.

With Shadow, it happens to be treats that he likes. Shadow also likes to play tug and I let him win sometimes, though I can get him to drop the ball for a treat.

Something else I am beginning to understand. Clicker training is not anti-corrective, from what I can see. It is pro-decision. As I have mentioned, I have only used corrections to avert something, then lead with a command that can be rewarded. 98 percent of his training is treats and play. I think the clicker can reinforce that, especially with the method of allowing the dog to learn and define behavior how he understands it and all I have to do is shape that learning cycle to the end desired behavior. So easy, even an electrician can do it.

Plus, believe it or not, I value the experience of everyone here, from the softest dog to Willow. And people, including Spiritdogs have had tremendous success using the clicker. And I don't see using a clicker as not being alpha, I think of it as being alpha with another tool.

Here's a brain-teaser. If a dog can train himself to problem solve and find his motivation to get along with the group does that mean that the dog is not dumber than the wolf but that humans, such as myself, get better understanding at what motivates dogs?
 
ETA: While looking for Shadow's lightbulb moments, I get some of my own. And I must say that one must try it to understand it and in that, I am following the basic rule of scientific discovery. Observe, find the theory that fits the facts. I also think that self-trained behaviors may have the possibility of being in the dog's tool bag when encountering stressful situations. And results count. Who cares if c & t results in a dog following your lead better than if they were corrected as they might be in a wolf pack? You achieved it without the use of force in a way that the dog led himself to do. So, It could be a way of re-aligning the dog's desires with yours.
 
If nothing else, it's a lot of fun, but I expect more from it than just that.

But the point of the beginning where you click for any ole thing is not hap-hazard as it may appear. The point is to help him realize that he can train himself. That he can problem solve. I also see the viewpoint of wanting a dog that obeys because they want to, because it is something they discovered that is rewardable.

Yes! I get very creeped out when I hear people say that their goal in training is to have a dog who "has no choice but to obey." Sorry, as far as I can tell, a dog always has a choice. Now, you can (at the extremes of this attitude) make that choice be "obey or I hurt you really bad" and even convince a dog through some head games that the choice is "obey or I kill you" but that's still a choice. I feel that clicker training works with a dog's ability to chose, not against it. It in fact counts on the dog being able to choose, to problem-solve, to think independently, and to make a decision. A dog always has a choice, you can't even beat that out of them, so why fight it? Why not use that ability to choose and make it work for you? I think this is a big reason why a lot of people report that they feel like trainig their dog is " a battle of wills." Because it is. A dog has free will. If you're goal in training is to eliminate that free will, you are going to have a battle on your hands.

That's why for me, hand-in-hand with the actual methodology of clicker training and NILIF and all that good stuff, I just in general to try to be the Life of the Party for my dogs. Because I know they always can choose to either work with me or not, I want to give them a good reason to want to be with me and work with me and enjoy being around me. And so far, it's been pretty effective. My husband is a miserable failure in this regard and it shows in how the dogs attend to him (or don't, as the case may be). He doesn't have anything to do with taking them to the park or the ballfield or any of the fun places we go, he doesn't play that fun clicking and treating game with them, he rarely does anything structured with them. He just comes home, gets them all riled up for 5 minutes, then wants them to get lost and then he pets them as they rest in the evening. So who do they want to follow, me or him? Who would you want to follow? He doesn't really punish them or yell at them (much) but he also has given them no reason to give a crap about what he says. The other day as he was attempting to give them some obedience commands (and failing), I said, "I hope you know who's fault it is that they don't really listen to you." I hate it when he blames the dogs because they aren't hanging on his every word, when he has done zero work to show them why they should care about what he says.

So anyone got any suggestions for clicker-training one's spouse? []

So anyone got any suggestions for clicker-training one's spouse?

Well first you need a really high-value reward... []

I think I like clicker training so much because I'm coming from a background of horseback riding. As my trainer repeatedly reminded me when I was a 70-pound, 8-year-old girl learning to ride a 1,300 pound horse - "You can't force the horse to do anything it doesn't want to do!" While my trainer was definitely "traditional" - there were certainly no clickers, and we were taught the usefulness of a crop early on - all of her lesson horses had also been saved from the slaughterhouse and often had their own special "issues" to show for it. As a result, she was very sensitive to the horses' individual needs and personalities. We had to learn to be equally receptive.

I was definitely taught to be an "alpha" and, essentially, not take any crap. Still, it's a very powerful lesson when working with such an enormous animal (especially when you're tiny) - the idea is for the horse to willingly choose what you want him to do. There are certainly no illusions about "absolute control" in that field!

A dog always has a choice, you can't even beat that out of them, so why fight it? Why not use that ability to choose and make it work for you?

 
My reasoning, exactly, as well. The dog always exercises a choice. It may be a choice to ignore your command, it may be a choice to obey out of fear of pain, it may be a choice to bite a chunk out of your hide. Or it may be a choice to recall and down every time you command because he/she is motivated to listen to your commands, because they result in something enjoyable, more enjoyable than whatever self-rewarding behavior they were going to engage in. Now, some behavior may be instinctual or even a release of boredom or energy, such as digging in the flower bed. Some of this can be met with seeing to the needs, such as exercise, providing a place to dig, etc. So, one doesn't abandon the needs and environment for pure learning theory. But I can't see a reason why it can't fit in with everything else. As Glenda put it, I want to be the benevolent leader that my dog follows just cause I'm so peachy swell. Also, since I do treat training anyway, I think this can supplement that, having a clear, definable marker. In any case, the look of happiness mixed with intense concentration on me is kind of neat, too.

I copied my response from Shippo's touch thread. Does this count as a lightbulb moment?


[blockquote]quote:

Yay! I just started teaching it today, and he began picking it up in ONE session! Man he's a fast learne[/blockquote]



"Oh yeah? Well, I just had to try it and Shadow had it on the third click. I waited for him to offer that behavior. At #6, I was walking around the house with him following me and touching my hand. So, from a distance, I said touch and he approached and touched, c & t. I wonder if this counts as a lightbulb moment for him. "
 
ETA:
So danged easy. Sitting on my keister at the computer. Shadow came in from outside and wanted to interact. So, I thought, fine, we'll interact and you're going to learn something and like it. In 3 clicks, he had it. So, I had to "push the envelope" as it was too easy.
 
And I think, as Benedict said, touch would make a great alternate recall. It's something he thought of and it has a word that means something and it means endless treats. So, I wind up with a well-fed dog that heeds my call and then rests and stays out of trouble. Oh the tragedy ...