British Labs / No Force????

[Jennifer Henion]

[QUOTE=mitty;1041756]Ok here's a video of a guy teaching a baby puppy using a clicker. It was uploaded 3 years ago. With all the advances in clicker training, how should we be doing it differently? Better yet, show us:

Renee, we are not communicating on the same wavelength. In clicker training a baby puppy, not much is new. Clicker training itself is not new. How you apply clicker training to teach skills in field dog work are a bit new. Using the platform is not new. Using it in clicker training (magnetizing dog to platform with clicker training) to teach precise positions and go-outs is kinda new (2-4 yrs) Using a platform to teach the concept of go out and sit facing me when I blow the whistle (whistle sit) is kinda new - at least to me, just came up with it in the last 2 months. Building on that and using the platform Pairing the whistle sit with casting is a little new, at least to me.

Pat Nolan has done an amazing job of melding both clicker/marker training with force training, from the few videos I've seen and word from others. His stuff is new, cause he invented the ways of applying the techniques for teaching certain skills and concepts.

Same for ecollars. You started a thread about history of the ecollar and how it evolved and how trainers have used it in new and improved ways. Same thing in my view.

[gdgnyc]

Ok here's a video of a guy teaching a baby puppy using a clicker. It was uploaded 3 years ago. With all the advances in clicker training, how should we be doing it differently? Better yet, show us:

First of all, I like this video. Those not familiar with the technique: the trainer is using hand targeting (useful for clicker trainers), and is conditioning the clicker to mean "reward" i.e you have earned a reward (not necesarily a treat but that is for another discussion), not exactly the same thing as "you did it right" in my opinion.

OK, mitty I think your comment is for me. I will be taking some heat from clicker trainers for my point of view but I still "power up" the clicker each time before I use it. I very clearly heard the trainer in the video say "probably after 30 or 50 of those". As you notice, he does not power up the clicker before use, he goes right into training and the secondary reinforcer (clicker) is conditioned on the go. Someone watching this never gets the idea that the clicker needs to be powered up (Pavlovian conditioning). Almost none of the clicker training videos on Youtube that I have looked at start with powering up the clicker. I like to start off with powering up. After that the click becomes useful as a marker immediately. Also, I power up my clicker in a slightly different way than I have seen most people do. Maybe on the video, my videos only go to certain people. I have to add that clicker training is new for me (about three years in use) but learning theory is not. In college we called operant conditioning "instrumental learning".

I might add that I do not exclude other methods of training.

[DarrinGreene]

There hasn't been anything "new" about clicker training since Pavlov proved the theory. The "new" things are techniques for inducing the dog to perform the target behavior. Advancement's in "clicker training" could just as easily be applied to -r training. They are advances in training technique, not advances in learning theory.

Sorry folks but Karen Pryor and company didn't invent it. Pavlov and Skinner did. Well actually, mother nature did. Pavlov and Skinner just figured out the puzzle.

[O.clarki]

no force??

I am to lazy to read through and see if this was brought up.. I say no such thing as no force. When you push your puppies butt down and say sit - that is force. When you tell him to stay, heel or here and are using a lead to get compliance you are applying force. Force is more then FF and cc. it is incorporated into the most basic OB at a young age.

[jeff evans]

What do you all suppose was the original intent of the "atta boy tone" on the tri-tronics collars?

[gdgnyc]

What do you all suppose was the original intent of the "atta boy tone" on the tri-tronics collars?

Unfamiliar to me. What is it?

[sixpacklabs]

Ok here's a video of a guy teaching a baby puppy using a clicker. It was uploaded 3 years ago. With all the advances in clicker training, how should we be doing it differently? Better yet, show us:

I think we've already established that marker training is neither new nor revolutionary, so I think you're kinda flogging a dead horse here.

That said, the fellow in your video could be doing a lot of things differently. Not because there has been a revolution in clicker training, but because he's doing so many things so poorly.

Just a few examples:
1. He's picked a really poor training environment to introduce the clicker. The dog is distracted and not really engaged with him.
2. If I'm training a dog with food, I would never teach it that seeing food in my hand means it will get it. Quite the opposite.
3. His timing is all wrong. When you condition the clicker, you click and then feed. It's classical conditioning. The click must precede the reinforcement. The point is that the click predicts reinforcement. He shows the food before he clicks. That is backwards conditioning. In at least some of the reps, the puppy appears to be eating the food as he is clicking. That is simultaneous conditioning. Neither backwards or simultaneous conditioning is effective for classical conditioning. You want the dog to learn that the click predicts reinforcement, and that it is being reinforced for the behavior it was doing at the instant it heard the click. But the click must come before the reinforcement!
4. He appears to be lumping behaviors, a recall with a hand touch. Good marker training involves splitting a behavior into simpler elements making it easier for the dog to learn. As Bob Bailey says, "Be a splitter, not a lumper."

My guess is what the puppy really learned in this session, if it learned anything, is about luring. It learned to come to a hand with food in it. I seriously doubt it learned anything about the clicker. I certainly see no evidence that the sound means anything to the pup. Typically, when a pup has learned that the click predicts reinforcement, you will see some sort of reaction when the pup hears the click...it might tip its head or turn toward you in anticipation of the reinforcement, for example.

I imagine that all of us who've ever trained with a marker have done things wrong. Some of us get better because we study and practice. Some give up. They fail not because it's a poor method, but because their application of the method is poor. You could say similar things about the e-collar. Some people create problems through the use of the collar. Not because the collar is a poor tool, but because their application of the tool is poor.

If folks want to learn how to use a marker properly, there are lots of great resources....Karen Pryor, Michael Ellis, Kathy Sdao, Ken Ramirez, Bob Bailey, Morten Egtvedt and Cecilie Koeste, just to name a few.

[RedDog]

Renee

I would also use the word "good" and "yes" with the clicker because the clicker is not heard that far out when you start to work in the field with the dog. And food may detract the dog from his up close teaching and work. Good while he is a pup but you don't want him dropping the bird for food. IMHO
I do target teach on my dogs as pups because in agility there are contact points the dog must touch (seesaw, walkit and A frame).

I am just not sure clicker's value for long blinds etc in Ft unless you do mega association work up close work with the clicker and the pager on your collar then???Just thinking out loud! Maybe I shouldn't!!!! Any body else have ideas how that would work.

Lots of ideas. Some things I've tried with my own dog, some things I've done with a number of dogs. (and to get it over with, yes I'm a beginner, no we haven't competed yet)

My dogs learn to retrieve with food. They learn to retrieve while I'm offering food and while others are. They learn to retrieve and hold while I tempt them with food or toss food. They learn to retrieve a dumbbell/bumper/bird that is surrounded by food. If I thought using food would make my retrieve weaker or cause problems.... all the more reason to practice in those scenarios.

Long blinds: Clickers/markers may not provide more help or information there, depending on how the behavior is taught. I don't use a marker for this, not because of the distances involved, but because it doesn't add anything. My dog is reinforced in training with a well placed toy or bumper or with a toy tossed by a hidden training buddy or a ball from a remote ball launcher. If I can make my reward appear ahead of my dog, a marker doesn't add much. When DO I use a marker on blinds? If we're working on the concept of "stay on the line despite X factor". When my dog makes the right move at that point, mark and toss ahead. Dog was making a choice to do the hard thing or go the easier/unconscious route....and that is a moment where I specifically want to relay it's the right choice.

Pager/noise/something from collar as a marker: I've considered it but probably won't go that route. I don't think it would add much to what I'm doing now. Over the last few years I've grown to love the idea of different markers for different types of reinforcers. Short version: At a distance I'd want a marker for "that's right, the reward is straight ahead of you" and one for "that's right, come back for your reward".

Am I crazy? Maybe. Am I telling anyone else to do this? No. Just sharing what's worked for us.

[yellow machine]

I wish this thread would go away soon.

[Evan]

I wish this thread would go away soon.

That is inevitably how everyone ends up feeling about these threads.

Evan