Overwhelmed by clicker training for the field

[EddieF]

By way of introduction (I've only posted in the photos forum before now), Winnie is a Wirehaired Vizsla just over a year old. She's done basic obedience, puppy kindergarten, and I've started doing some work in the field with her, hoping to one day have a companion hunter. I'm not interested in field trials.

But having exposed her to birds (she is excited by birds, seeks them out, finds them and points them - then chases them if they flush), but to really train her on the next steps I needed some help. The bad news is, I have a goal to do all this without the use of a shock collar, and hardly anyone training gun dogs thinks it can be done without one. The good news is, the guy who literally wrote the book on positive gun dog training recently moved to my area, and he came out to teach us some things!

I am new to clicker training, but he explained how to break something like 'retrieve' down to much smaller pieces and shape each step of that behavior using the clicker and rewards. Lots I haven't thought of before, like the difference between fetching a ball because it's flying through the air and she's excited about it, to a trained retrieve on command of a bird.

Anyway, I am SO overwhelmed with how much there is to do and how small each step is along the way! But the other day we decided to start so I just picked something. I want her to 'whoa' (stand/stay) on command, and then eventually move that to a whistle. So we just walked around the yard and I'd put my hand in front of her and say 'stay', if she stays standing I click/treat. If she sits, we just do it again. We'll work on it, increase the distance and duration, then assign 'whoa' instead of 'stay' (I use 'stay' now because she knows it, but I may quickly phase that out so as not to confuse her). And then eventually add the single whistle blast before the verbal command, then drop the verbal command altogether.

When I think about those steps, I can see how it can work. And how this all really can be achieved using these positive training methods. But all that work is just ONE tiny aspect of having a reliable hunting companion! It's fun to work with and train her, but do you ever feel like there is just so much to do it seems impossible to reach your goals? I used to think my goals were modest and simple, but the nature of clicker training, breaking things down and shaping behavior, adds so many steps to every complex behavior that it's really overwhelming!

Another question I have is, isn't it confusing to a dog if you're working on a few different things like touching a target or lying down or stand/stay, and let's say you haven't assigned a verbal command to some of those. If she gets clicked/rewarded in one session for looking at a dog bed, and the next session for giving a paw or whatever, and you haven't given those things verbal commands yet, isn't that confusing?

Sorry for the long post.

 

[Doberluv]

Will you be able to keep on working with that guy? That is wonderful that he lives near you. That would be the greatest if you could enlist his help for a while until you feel less over whelmed. The thing about clicker training is that once a dog has been trained this way...even for a short time and hasn't been trained using a lot of corrections, he learns that offering possible behaviors may earn him a click and a treat. So, they try things more instead of a dog that is corrected a lot...that dog will tend to not offer behaviors for fear of punishment. So, the odds of the clicker trained dog hitting on a clickable behavior are high. Once they get onto the game of clicker training, they become active participants, in other words and learning accelerates. And your learning accelerates also because you are being reinforced too all along the way, as your efforts are rewarded!

Once the dog gets onto the game and is c/t for small steps, the remaining small steps tend to go very fast...like 2 or 3 reps and he's already going to be succeeding at a more sophisticated example of the behavior you're aiming for. Once people start mixing positive reinforcement and corrections, such as collar pops, scoldings, etc...(those so called, "balanced" trainers) the momentum gets interrupted. The rules of the game become confusing and learning is slowed.

Another question I have is, isn't it confusing to a dog if you're working on a few different things like touching a target or lying down or stand/stay, and let's say you haven't assigned a verbal command to some of those. If she gets clicked/rewarded in one session for looking at a dog bed, and the next session for giving a paw or whatever, and you haven't given those things verbal commands yet, isn't that confusing?

No, it shouldn't be confusing. These clicked/treated behaviors, she learns are worth repeating. I usually don't introduce too many new behaviors all in one session. Maybe one or two. I don't even use sessions so much as little snitches of time here and there. (when I'm even training my own dogs at all these days) It takes time for a green dog to learn how to learn but once they do, they become real pros.

If you're comparing pos. methods to compulsive methods, I still think the dog learns HOW to learn better. Compulsive methods such as using a shock collar simply teach a dog how to avoid pain or discomfort and not to do something..... not so much how to offer new behaviors. It's too scary to offer new behaviors. Which way is more conducive to learning? Shutting a dog down or allowing him to try things and then receive information as to which things work to get a reward? Which way is he going to increase the volume of potential feed back from you? The more feedback...info, the quicker the incidental, unrewardable behaviors are weeded out of his repertoire and the quicker the rewarded, wanted behaviors will be repeated.

Now, proponents of compulsive training methods will say that it's the same difference...that the dog recieves info just the same as he is corrected for unwanted behaviors. Don't do this, don't do that, don't do this other thing. The thing is....I, personally don't want to be telling my dog, "no" all the time, even if he is rewarded some of the time. It has too much risk of shutting them down and they can't learn. It's much easier to tell a dog what TO do than what not to do. The list of what not to do is much longer than the list of what TO do.

The verbal cues do come later, once the dog is getting onto a new behavior. There's no sense in using them to attempt to elicit the behavior unil a strong association can be made. They aren't going to know what the word means at all for a while. In fact, using a verbal cue too early can cause your dog to miss the association. It can also cause the dog to associate the word with the wrong behavior....something you didn't intend. Get the behavior first, get it pretty regular, then add your cue word just as the behavior about to finish, while he's moving into the position, say, not when he's already finished. Do that for a while until you think he's made the connection. Then try using it as an elicitor when you're almost positive he's about to perform the behavior, just as his muscles are contracting in preparation. Gradually move back with your cue as you become more confident that he will do the behavior you're asking. If it doesn't work to elicit the behavior, don't keep using it... go back a step and work there a little more and then work back up. You don't want to ruin your cue word and have it wind up meaning something else to the dog.

I hope that helps a little. I would recommend trying to get that trainer to work with you and your beautiful dog.

btw: You're sorry your post is long? I out typed you. :rofl1:

 

[Doberluv]

The bad news is, I have a goal to do all this without the use of a shock collar, and hardly anyone training gun dogs thinks it can be done without one.

I don't know anything about gun dog training. But my Dad use to hunt with a couple of friends who had trained their own dogs. One guy had Brittany Spaniels and one guy had a choc. Lab. They were stupendously trained, phenomenal. I watched them often. And they did not use shock collars (don't know if they even were invented then) lol....and they did not manhandle or be cruel one little bit to their dogs. They were all gentle and treated their dogs like one of the family. They lived in their house with them, which at that time was still thought to be a mistake with hunting dogs in a lot of circles. Anyhow, I do not believe in shocking a dog to get what I want no matter what. If the dog can't be trained without, then I won't do that thing that I want with my dog. That's just my own philosophy.

Those people who think it can't be done may be too lazy to take the effort, time and skill it takes to train without a quick fix of electric shock. I know there are ways to condition a dog to not chase prey when un-asked. I would be careful in the meantime to not let your dog get the (self) reward (fun) of chasing the birds (or whatever it is you don't want her to do) until you have some situational control. I taught my Dobe not to chase deer or rather to come when called in spite of wanting to chase deer by using doggie zen concepts. I set up practice situations with faux prey and used a helper. Long story short....he learned that by not chasing the prey when HE chose, he would get to chase the prey in short order. The recall was further strengthened and he would come even if he were in mid chase of deer. He had a pretty intense prey drive, but he also was a Doberman. LOL. (come to Mama) I started when he was young and hadn't had the opportunity to be reinforced too much by chasing animals. (I live in a wilderness area and do a lot of hiking with my dogs) He wasn't perfect....did chase a bear off the property once with my male Chihuahua. That time it didn't work the first time. I did have to call a couple of times. But most of the time, he was very obedient.

 

[EddieF]

Thanks Doberluv, I've read some of your other posts on the subject and I was hoping you would chime in here.

I can work with him but this was a private session that was not cheap. He's still getting established, and in fact we're talking about hosting a weekend seminar for gun dog clicker training at our property because we have a lot of room. He'd let us do that for free in exchange for hosting 6 or 7 other dogs and owners. But he has been generous with his time via email, I just don't want to wear out my welcome and I knew there were some knowledgeable people here.

I have another question, even though I am overwhelmed, I have a trick in mind I want to teach her too. Just here and there along the way. And I may have even asked this here before, but clicker training seems to be the way to get this trick down. But you could use clicker training to teach a dog to raise their front paws in the air when you yell "touchdown," right?

 

[Doberluv]

LOL. That sounds like a cute trick. Does she ever put her paws up like that on her own accord? Can you get her excited and playful so she'll do that? If so, click at the very moment she's putting her feet up, even if it's not all the way up or long enough. Just reward for an approximation. When she's doing that more readily because she's learning that behavior gets her a scrumptious treat, c/t for that a few more times till it's pretty regular. Don't stay stuck though, at that level for too long. Some dogs have trouble getting themselves un-stuck so they can move along. It's a judgement call to be sure. Now, once it's pretty reliable, withold the c/t for that. Now, a regularly clicker trained, non punished dog isn't going to be put off by that....maybe only mildly disappointed momentarily. But these dogs seem to think...Okay, hmmm...I know the rules. When it stops working, all I have to do is try a little harder, try something else. Let's see...I'll raise my paws up higher and see if that works. Bingo! C/T. Then the dog starts repeating that every time. They learn the rules and play the game. Then you withold for that and the dog tries a little variation...maybe tries to jump clear up off his hind feet. If you don't want that, don't click or treat. (I'm a poet) Just wait. He'll try something else. You can c/t the examples which are of the longest duration, even if one is only a second or two longer. c/t. Wow! That works. All I have to do is reach up and keep my paws there for a little longer and I get the treat. Of course, I doubt they're going through all that detailed logic in their minds. I think it's more just this works. Do it again. This doesn't work. Don't bother doing that again.

One word of caution: Dogs trained like this will most likely try the behavior at all kinds of random times when you aren't asking for it and don't want it. Be sure to ignore. Turn your back, don't say a word. When she's settled down, you can ask for the behavior. Remember, don't add your cue until you can get the behavior going.

Sometimes capturing is a great way to go and sometimes shaping is more appropo. (I can't stop rhyming) Or use both for one trick. When you happen to catch her doing it, by all means c/t. If she doesn't get around to doing it on her own, see if you can rile her up a little so she'll do it, even a small aproximation of it. Then remember about c/t, the behavior getting pretty reliable and then witholding, waiting for better, then c/t etc.

There is also something called successive aproximation. That's when you are changing the environment to get the behavior to change. If you're using a pole or something to have her put her feet on so you can get her to keep her feet up longer, you might put the pole at a very low level at first. Then gradually raise it higher, then once that height is achieved, you might fade out the use of the pole or bench....whatever. (I wouldn't do this trick that way, but..)That use and regulation of a prop would be utilizing successive aproximation vs. shaping....where the nuances of the behavior itself are changing and being regulated by reinforcement. Like jumping. You start out with a low jump, then raise it higher little by little. Successive aproximation.

I really am not that good at clicker training. There are people here much better at it. I think I understand behavior and what dogs are capable of understanding, how they perceive things pretty well. But as far as all the ins and outs, no...I am not your best bet.

What about some books? Have you read Karen Pryor's books? She's sort of the queen of clicker training. Or websites? Some books might help too. And I bet someone will come along here that can really add something. Adojrts is one I think of, Dekka is another. I'm sure there are more, but I can't think right now. You might go hunting for them if they don't see this.

 

[EddieF]

Thanks again. I thought that trick would be fun while we're watching a game on TV. But I am a Redskins fan, so it's a trick that isn't likely to come up very often.

As for books, I do have the gun dog book and am getting through that, but as many training books as I buy, I simply can't get through them. Another thing I don't think I'll be able to do is to keep training logs, though this guy I worked with and others say it's important. I don't think I can do it.

 

[Doberluv]

I thought that trick would be fun while we're watching a game on TV. But I am a Redskins fan, so it's a trick that isn't likely to come up very often.

DVR or DVD? Does it have to be only when the Redskins are playing? How about the Mariners?

Oh yeah, the training log. I think it's a great idea. I never did that either with my dogs. :dunno: But then, they're not trained to the hilt either and they're not gun dogs. My dog with the most training was my now deceased Doberman.

 

[corgipower]

Great topic!! I'm enjoying reading it. I'm not a clicker trainer, so I can't give you any advice, except...

Another thing I don't think I'll be able to do is to keep training logs, though this guy I worked with and others say it's important. I don't think I can do it.

I have tried a few times to keep logs. It's never been successful. I do a lot of training just in the course of daily activities that have an overall effect but don't get documented. There are so many variables involved in training that it's close to impossible to document accurately. The dog may have had trouble with heeling, but was it because the wind was blowing or because of a scent, or because it was hot or cold or maybe my walk was different or maybe he didn't understand what I asked for...what goes in the log?

I do know a number of people who keep logs when they train scent work ~ tracking, odor detection. I would ask the trainer to show you his logs and to help you set one up if you're going to keep one.

 

[EddieF]

I feel like the burden, however slight, of logging things accurately will actually prevent me from fitting in little training opportunities here and there. I don't know, it just seems like something that will make the training less fun for me, which will result in a quick drop in frequency or enthusiasm toward it.

I have snakes, and used to log when they ate and how much. When they shed, how much they weighed. Then I finally thought, "why the hell bother?" They eat occasionally, and snakes are cool with that. Sometimes they don't eat but they'll almost never skip two meals in a row. They shed when they shed and they look to be at a good weight, so why make things harder? I know it's not the same, but it points to me knowing what I won't bother with.

 

[release the hounds]

I can think of a few times when having a "log" could have been useful, but overall I think the act of keeping it accurate and current would have sucked the life and fun out of all training and thus hurt the training rather than help.

Now that may be different for some people that thrive on the very minute details and documenting them. Not to say I don't pay attention to details, i just don't find the need to "obsess" over them. I use that word cause that's how it appears to me.

To me, training happens too fast to document it and it always changes depending on a thousand different variables. If I were to keep a log, it would be simple. Like any one thing that stood out. and my plans would be very loose and easily changed. IDK, for me, I know what i'm missing or what needs work by the end result and I don't need it on paper to remind me what to work on or what i've done. My training and the dog show me that.

 

[Doberluv]

I think the point of a log is to have a game plan, a general group of things you want to work on next time based on what was lacking the last time. I have taught various things in my life time. As a swim coach and intructor, I would make a lesson plan for the next day. I'd list off things that needed reviewing, aprox. how much time I'd spend on that. Then list some new things I wanted to introduce...how much time would be spent on that. I might expect one or two kids to have particular trouble in certain areas and give myself a little cushion to help them. There was only so much time in a day for a lesson or practice and it was most helpful and efficient to have a plan. Not everything was followed exactly, of course, because of variables. But there was a general outline. I did the same thing when I taught riding lessons and piano lessons. Now, with the owners I work with, I sometimes write something out for myself as a plan...things to not forget...like making a note to yourself not to forget a hair appointment. Don't forget to mention this or to work on that.

I think that is the same idea with a log.....you can list things you'd like to pay particular attention to, things you think need extra practice, new behaviors to show, what order makes the most sense in a particular session, if you're doing sessions persay, what kind of reinforcers are you going to use for which things. Which will come first, which will be used later in the practice. When you finish for the day, what transpired will be very fresh in your mind and you'll be able to make a more sequential and efficient game plan for the next time. Maybe you can keep it all in your mind without writing it out. Maybe you just are more comfortable playing by the seat of your pants.

I know some people keep much more exacting logs....like how high did the dog jump, how squarely did he sit, how much of the corner do we need to cut when circling around for the next jump, how can we do it....Measuring....comparing. I can see how that much detail would make me crazy. But I bet it helps a lot of people.

At any cost, I think it's most important to never ever let anything interfer with the fun you have with your dog or your relationship and bond with each other. If it becomes a drugery, there is no point at all. And in fact, the learning will be compromised. You must have a happy, positive attitude. If you're not in a good mood, forget it that day.

Just my .02

 

[lizzybeth727]

When I think about those steps, I can see how it can work. And how this all really can be achieved using these positive training methods. But all that work is just ONE tiny aspect of having a reliable hunting companion! It's fun to work with and train her, but do you ever feel like there is just so much to do it seems impossible to reach your goals? I used to think my goals were modest and simple, but the nature of clicker training, breaking things down and shaping behavior, adds so many steps to every complex behavior that it's really overwhelming!

Yes, I definately know what you're talking about here. I train service dogs, teaching chains of behaviors like retrieve, pushing doors closed, tugging people's jackets and socks off, turning lights off and on, etc. The dogs have to learn these complex chains, PLUS have to have enough obedience training to work politely - and do all of the above behaviors - in public with an infinate number of variables. AND we have to do it all in under a year. And of course, we use clicker training.

One thing to keep in mind when it feels like you're not getting anywhere is that the more foundation work you do with your dog, the better the end result will be. For example, I had a collie that I recently worked with, and I had to teach her retrieve. This collie, like many collies, HATES to hold things in her mouth. So you can imagine how frustratingly slow it must have been to teach her to hold things in her mouth. I worked with her on retrieve nearly every day, and to teach her to pick the dumbell up off the floor and put it in my hand, took SIX MONTHS. I literally cried the day it happened, it was one of my proudest professional moments. But, to get from that point - pick up dumbell, put it in my hand - to the place she is now - pick up any object, of any texture, carry it as long as I ask, and drop it in my lap - only took about six weeks. All that foundation work I did to convince her that the dumbell in her mouth is a good thing, really paid off in the end, as it made her final training go much faster.

I feel like the burden, however slight, of logging things accurately will actually prevent me from fitting in little training opportunities here and there. I don't know, it just seems like something that will make the training less fun for me, which will result in a quick drop in frequency or enthusiasm toward it.

I understand what you mean. I am supposed to log every training session I do at work; I'm honestly not very consistent in doing it, but it definately makes my training sessions more productive. It helps me remember where I was last time I worked on that behavior, and gives me a starting point for where I need to go in this session. But more importantly, I think, it reminds me how far I've come in training a behavior. Like I can look back and see that the first time I practiced stay, my dog had a hard time staying for even 1 second.... then 10 sessions later, he's doing great at staying for 30 seconds. Yeah, it seems like it slows you down, and if it sucks all the fun out of training then by all means don't do it, but I'd suggest trying it for about two weeks - long enough to really get comfortable with it - just to see how you like it.