How do you train? Tell me everything you think/know/do.

spiritdogs

Wilkes uses a clicker, so many people like to use him as a defense for mixing methods, but, by definition, what he does isn't clicker training in the sense that clicker trainers understand it.  I would much prefer that people call it Wilkes training, or mixed training, but make the distinction between those who practice "clicker training" and those who are "training with clickers."   It is NOT the same thing.

 

 He was a clicker trainer when Karen Pyror liked him ;)

 I've been involved with clicker training for long enough to have seen when it started to take this strange turn. This "clicker user vs clicker trainer" argument, the idea that if you really want to benefit from clicker training it is the only method you can use, that to truly be a clicker trainer you must shape the behaviors and what I think of as "clicker extremism".

corvus

Liesje

corvus

There can be quite a bit of pressure in clicker training. It's a great environment to build up a dog's confidence and teach them to work through a little frustration or anxiety because you can control it to such a fine degree. 

 

I guess it depends on how you define pressure.  Like I said in Karen's drive thread, to me there is a difference between creating stress and applying pressure. You can create stress by applying pressure but I don't think that because a dog is out of balance/stressed a bit during training means that that pressures have been applied.

 

I was responding to Jason's comment about a dog needing to learn how to deal with a correction so they can cope with some form of stress or pressure in their life. I agree that a dog needs to be able to cope with things not going their way, but I don't believe I need to correct my dogs to achieve that. I am very careful about monitoring how my dogs cope with strange and unpleasant things and deliberately set about allowing them to gain success with coping strategies I think are safe and appropriate and removing stressors if it looks like they will not have success with the coping strategies I encourage. Example, a few months ago Erik and Kivi met a strange dog at a dog park that kept running at them and nipping them. Kivi dealt with this worrying situation by coaxing the dog into a game with him. I would expose him to any number of flakey dogs as long as he is happy to employ a non-aggressive coping strategy and has success with it. On the other hand, Erik tried to avoid the dog and give it "don't hurt me, I'm no threat" signals, but the dog kept coming back at him. I left the park with him because he was not having success with his non-confrontational coping and I felt that any moment he was going to try something else and I wasn't sure that it would be something I would want him to have success with. I would have stayed if it looked to me like he was going to change to another non-confrontational strategy. Learning another of those would have been worth the risks with the flakey dog IMO. Although I will point out that I was confident it wasn't going to hurt anyone.

I don't want to protect my dogs from every stress they might encounter. I want them to learn to cope with stress on their own. I believe it will help them to be confident, well-balanced dogs that can handle themselves when things don't go the way they like. 

 

Fair enough, I was just curious because "quite a bit of pressure" to one trainer and dog can be absolutely nothing to another.  That is why I'm curious as to how you define pressure and how you apply it.  I see pressure as something more tangible than putting a dog in a stressful situation.  Most people do not have any reason to use the type and the amount of pressure that Jason and I are thinking of.  For one we are engaging in breed-specific training and testing with dogs supposedly bred to work through a significant amount of pressure and stress.  A dog not purposefully bred for such work probably has no business being subjected to it, nor dogs that don't possess the same nerve (like my Kenya and Coke are trained totally differently than Nikon).  This does not necessarily mean applying corrections from the handler to impose pressure, but in previous threads on here people have used extremely broad definitions of "correction".

I was not talking about stressful situation "in general". In the case of a flakey dog, I would get my dog out of there too. I don't care if my dog is 90lbs GSD, full of fight and drive, and the dog bugging him is a 5lbs chihuahua.

Rather I was thinking more about the stress that comes from the handler/dog relationship. Michael Ellis, who does a lot of training with markers and is really good at it, is a big proponent of introducing dogs to escape training early (early as in before they are a year old). Escape training is basically where you apply pressure on the dog until he submits and as soon as he submits, you turn off the pressure. In competitive obedience or field trial they will sometimes use it to train retrieve and in Schutzhund they use it for retrieve and tracking (again, just some trainers, not all trainers - obviously Liesje did not do forced retrieve with Nikon) and if you have seen it in practice, you know - when done on an adult dog - it is NOT for the faint of heart.

Obviously when Michael Ellis does it with a pup or a young dog, the pressure is very very insignificant, just enough to make the dog uncomfortable. But the principle is the same: stress the dog, show the dog how to turn off pressure, let the dog learn to work through discomfort. Why is this important? To me, it's important because it teaches the dog a valuable lesson about working through pressure/stress THAT COMES FROM ME and it's also a good place to teach the dog sometimes you will have to do something even if you don't want to. You have to do it because I say so. And if the dog has been through this kind of training, then when he does get a correction here and there from me, it is not a big deal to him. Submission is not a big deal. Pressure is not a big deal to him. Because he knows how to turn it off.

And this applies not only to sports but also to everyday life. For example, with my dogs, when I train positions, I train it first motivationally and then when I done with that, I go back and train it again with some kind of compulsion. It does not mean I drop the hammer on the dog. With my corgi mix who is a soft, nervy dog, it means I put my hand on her butt and very softly (but insistently) push her down in a sit. At first she didn't like it at all and would flip out because she didn't like to be "pressured" into doing anything. But I kept it up and overtime she comes to accept that when I put her in a sit, it's no big deal, there is no need to flip out, BUT she still needs to sit - even if she does not want to and does not like to be told what to do, she needs to accept pressure from me and when she feels the pressure, she needs to comply.

I think it's good you bring up Ellis in this context.  He has show that you can use pressure AND freeshaping at the same time.  One does not negate the other.  I like how he demonstrates "yielding to the leash" as a very basic form of working through pressure (and he is basically freeshaping it with mark and reward, he applies the pressure, lets the dog figure out how to turn it off rather than luring the dog, marks the dog's behavior, and rewards).  I think it's a good example of something that relates to *every* pet dog and owner (not just specific Schutzhund stuff).  I did a lot of that exercise with Coke for his obedience training, and of all my dogs he takes the least amount of pressure (but actually recovers the fastest).  I think there's a video on Ed Frawley's site...

AgileGSD

This "clicker user vs clicker trainer" argument, the idea that if you really want to benefit from clicker training it is the only method you can use, that to truly be a clicker trainer you must shape the behaviors and what I think of as "clicker extremism".

Well, to be fair, I noticed a significant change in my dogs and the way they reacted to clicker training when I went from "training with a clicker" to "clicker training". It's not the same IMO. One is not an extreme form of the other. It's more about what the dog expects and how aroused they get. When I was training with a clicker I could never have done some of the things I've done with a clicker since I started clicker training. Like using the clicker as an interruptor as per the LAT game, for example. I couldn't do much shaping at all with my older dog when I was training with a clicker. He found it too stressful to work through. I'm not a clicker purist. My little guy Erik can be a tricky dog. He is very opportunistic, pushy, and confident. He tries things many dogs don't try. I love that about him, but it takes a bit of decisiveness and confidence on my part to handle him. Clickers get him excited and I don't always want that. However, for all intents and purposes I guess you would call me a clicker extremist. If that is the case, I think it simply means I'm using a clicker correctly. It's not just a marker.

Jason L

Obviously when Michael Ellis does it with a pup or a young dog, the pressure is very very insignificant, just enough to make the dog uncomfortable. But the principle is the same: stress the dog, show the dog how to turn off pressure, let the dog learn to work through discomfort. Why is this important? To me, it's important because it teaches the dog a valuable lesson about working through pressure/stress THAT COMES FROM ME and it's also a good place to teach the dog sometimes you will have to do something even if you don't want to. You have to do it because I say so. And if the dog has been through this kind of training, then when he does get a correction here and there from me, it is not a big deal to him. Submission is not a big deal. Pressure is not a big deal to him. Because he knows how to turn it off.

It's nice in theory, but I don't like it in practice because unless you plan to put quite a lot of pressure on your dog on a regular basis I don't think it's necessary. I don't think it's necessary because I get the same result if I use nothing but food. Erik hates having his teeth looked at. The first time a vet tried to do it when he was 4 months old he threatened to have her hand off. So we taught him to hold still and let us lift his lips and inspect his teeth on command. We used food and did it in small steps. He still hates it, but a complete stranger can look at his teeth and he will go still and wait passively for it to be over. This came in handy the day he got something stuck across the roof of his mouth. His conditioning overcame his distress and he went still when I pulled his mouth open so I could see what the problem was and yank the offending object out without having him so much as try to pull away from me. I'm not going to pretend I never put pressure on my dogs or make them do something they don't want to do. But it's part of living with me and the more I counter-condition them through any troubles they have the fewer troubles they have in future. They don't need to learn to accept pressure from me. I don't intend to use much. All they need to know is that I'm a good bet and if I do something they don't like I will probably have something good for them afterwards. To me, that is a much better lesson for them to learn than that I do mean things to them but it's not so bad if they stop trying to make it stop. In the meantime, I do lots of massage and gentle desensitisation to being handled. I work up to rough handling like scruffs and pushing and pulling them. It doesn't bother them because I habituated them to it in a safe atmosphere.

I find it rare to have one of my dogs fail to do as asked if they know what I am asking. Generally they are so used to being rewarded for it they do it before they think about whether they want to or not. Kivi has his spitz moments, but that's what steak is for. Occasionally Erik especially will get this mulish "I don't want to." look on his face, in which case I usually say "fine, if you don't want to I guess you don't want any of the good stuff I have for you." I can't even think of a time when he hasn't changed his mind. If he's not complying with something the reward history isn't strong enough, or he's hurt/frightened, or he doesn't know it in this context.

I had a great chat with the guy that oversees the police dog training unit for my state last week. A lot of the things they need to do with their dogs are just not necessary for pet dogs. He was telling me about the tests they do for "nerve" and bounceback. He seemed embarrassed about it, possibly because my supervisor, who is an extremely empathetic person, saw it once and probably thought it was a bit rough. I don't for a second think they don't need to test for those things and get their dogs into defence drive at some point. People are going to try to hurt them when they are police dogs doing their job, and they have to be able to work through it. But pet dogs don't need to and many probably can't. To me, putting pressure on an animal just means upsetting their equilibrium. I think it's a wonderful means of very carefully teach them coping skills and strategies, and to give them confidence. Animals get a dopamine rush when they successfully alleviate pressure. It's a powerful tool, but one that should be used with caution IMO.

AgileGSD

spiritdogs

Wilkes uses a clicker, so many people like to use him as a defense for mixing methods, but, by definition, what he does isn't clicker training in the sense that clicker trainers understand it.  I would much prefer that people call it Wilkes training, or mixed training, but make the distinction between those who practice "clicker training" and those who are "training with clickers."   It is NOT the same thing.

 

 He was a clicker trainer when Karen Pyror liked him ;)

 I've been involved with clicker training for long enough to have seen when it started to take this strange turn. This "clicker user vs clicker trainer" argument, the idea that if you really want to benefit from clicker training it is the only method you can use, that to truly be a clicker trainer you must shape the behaviors and what I think of as "clicker extremism".

 

I don't think it was a "strange turn" at all.  I think the distinction came when people started to see the results, and learned that you can get excellent results without the use of corrections.  That is not to say that it doesn't come at the price of being willing to let the dog make mistakes.  The biggest deal seems to be that humans are so opposed to the dog making any mistakes.  Isn't that how most organisms learn?  Good trainers in all disciplines are able to control the dog's environment during the training process so that the mistakes are not life threatening, so how is it a problem to simply avoid reinforcing mistakes and reward good responses?  Even if you train with correction, you don't let a dog run off lead until the recall is installed, right?  I also don't think you should take my comments to mean that you cannot benefit from clicker training unless you use it to the exclusion of other methods.  That is not the same as saying that you might do better if you did.  Even I don't train my dogs' every behavior with a clicker.  Horrors, I even use lure/reward if I think it will transmit the info to the dog sooner.  Why shape a "sit" if you can elicit it quickly with a lure, then fade the lure?  (Unless, of course, you are training a greyhound, in which case, a clicker can come in really handy for the ones who think they cannot sit if they are attached to a human lol). 

I really think the definitions are helpful only in that we can identify which camp we want to be in and be able to communicate that to others. 

BTW, what Karen Pryor likes or doesn't like is not gospel.  Even those of us who deeply appreciate many of her contributions do not regard her, or anyone else, as a guru or god.  Funny, but one of my colleagues once said to me that she knew I was a good trainer when I started to question the masters of the trade, holding them accountable for their hypotheses, and thinking on my own how to apply what I had learned about behavior for the ultimate benefit of my dogs and those of my clients.  We don't all do things exactly the same, but it doesn't mean that we all shouldn't always question the legitimacy of what we are doing.  Frankly, I've been around long enough to have seen clicker training in its infancy, too, and if something else comes along, maybe I'll try that.  What I know is that there are only so many things that alter behavior...medication, operant conditioning, classical conditioning.  There will always be new names for those things: Constructive Aggression Therapy, BAT training, etc.  But the premise underlying it all is the same. 

corvus

AgileGSD

This "clicker user vs clicker trainer" argument, the idea that if you really want to benefit from clicker training it is the only method you can use, that to truly be a clicker trainer you must shape the behaviors and what I think of as "clicker extremism".

Well, to be fair, I noticed a significant change in my dogs and the way they reacted to clicker training when I went from "training with a clicker" to "clicker training". It's not the same IMO. One is not an extreme form of the other. It's more about what the dog expects and how aroused they get. When I was training with a clicker I could never have done some of the things I've done with a clicker since I started clicker training. Like using the clicker as an interruptor as per the LAT game, for example. I couldn't do much shaping at all with my older dog when I was training with a clicker. He found it too stressful to work through. I'm not a clicker purist. My little guy Erik can be a tricky dog. He is very opportunistic, pushy, and confident. He tries things many dogs don't try. I love that about him, but it takes a bit of decisiveness and confidence on my part to handle him. Clickers get him excited and I don't always want that. However, for all intents and purposes I guess you would call me a clicker extremist. If that is the case, I think it simply means I'm using a clicker correctly. It's not just a marker.

 

So, using a clicker to freeshape = clicker training and using a clicker incorrectly = training with a clicker?  IMO if the clicker is not being used correctly it's not really training at all, it's more like fumbling with a clicker.  Either it's being used correctly or not.

Again I'm with Lies on this. I don't see why clicker and freeshaping have to be linked together. For me, the power of clicker resides in it's ability to pinpoint, isolate, capture moments in time for the dog, to communicate to the dog in a very precise way "THIS is what I want". But it doesn't mean that it only works in conjunction with freeshaping or it only works in an entirely "positive" training environment.

I just got done teaching the "Hold" command to my 11 months old pup and I used to a clicker to teach it but I didn't do it in the traditional shaping way. That is, I didn't hold out the dowel and then click the dog first for looking at it, then smelling it, then mouthing it, and then finally holding it (you can, of course. It just doesn't work well with this particular pup ...). Instead I physically put the dowel in the dog's month, put my hand around his muzzle, pressed down, told him to hold it and click him as soon as I felt him tighten his jaws. And over time I lessen the pressure of my hands until he was holding the dowel on his own. The clicker is invaluable in this process because it allowed me to pinpoint various key moments in this training: the moment when he first accepts the dowel in his mouth, the moment he first tighten his jaws, the moment he commits to holding the dowel on his own. Other people do it differently but the way I did it, there was not a lot of "WANT TO" in the exercise. Rather, right from the start, I impressed on him that he "HAS TO" do it but nevertheless the clicker plays a very important part in the whole teaching process.

It seems the distinction is really freeshaping and not clicker training? You can train with a clicker and not freeshape, and you can freeshape without using a clicker.  Personally, I love behaviors that are 100% freeshaped and backchained, but, that's not always possible or even safe.  For example, you can't really freeshape the the back transport during protection work!  I can't just let my dog try biting the back of the helper's thigh and giving my NRM, I don't think that would go over well!  There are some behaviors where you MUST set boundaries right away, and others where it's probably a good idea to completely freeshape.  I think it keeps a dog's mind fresh and encourages the dog to be proactive. 

I also think it has a lot to do with temperament (I'm beginning to sound like a broken record!).  Nikon is extremely outgoing and confident and also unflappable when it comes to corrections, doing things wrong, and not getting a reward.  He is a joy to freeshape because his temperament really lends itself to that training style.  His mind is always ticking and his body is always in motion so even freeshaping a totally unnatural behavior comes quickly to him using marker words and NRMs to keep him going.  On the contrary, Kenya is very lacking in confidence, simply not receiving the food treat or not getting praised is sometimes enough for her to shut down (it doesn't have to be any sort of verbal or physical correction).  With her it is difficult to set her up for success using freeshaping because she thinks she's doing something wrong and will give me a pitiful look like, "Just please tell me/show me what you want and I'll do it!"  It's sort of a viscous cycle of having this soft temperament and needing more direction and luring in training.  She was raised and trained for years by someone who did not use physical corrections with her, very familiar with the clicker, and is against luring but for freeshaping, so it's not as simple as saying if she had more of that training from the beginning, she'd be more adept at it now.

 

Again I'm with Lies on this. I don't see why clicker and freeshaping have to be linked together. For me, the power of clicker resides in it's ability to pinpoint, isolate, capture moments in time for the dog, to communicate to the dog in a very precise way "THIS is what I want". But it doesn't mean that it only works in conjunction with freeshaping or it only works in an entirely "positive" training environment.

I don't recall anyone saying that freeshaping had to be linked to clicker training.  Many of us freeshaped behaviors long before clicker training became popular, but the fact remains that a dog who is frequently corrected is harder to freeshape, whether you want to admit it or not.  That's because a dog that is often corrected has learned NOT to offer behavior that he thinks might elicit a correction.  Simple as that.  (I often feel like those dogs are slow to learn, but it isn't that at all, it's that they have learned, quite quickly sometimes, to remain still or to do a known behavior, rather than trying a new one.)  The dog that is rarely, if ever, corrected, usually has no problem offering behavior (unless the trainer tries to up the criteria too quickly), and in my opinion, is far easier to shape, whether you are using a clicker or not.

Jason L

I just got done teaching the "Hold" command to my 11 months old pup and I used to a clicker to teach it but I didn't do it in the traditional shaping way. That is, I didn't hold out the dowel and then click the dog first for looking at it, then smelling it, then mouthing it, and then finally holding it (you can, of course. It just doesn't work well with this particular pup ...). Instead I physically put the dowel in the dog's month, put my hand around his muzzle, pressed down, told him to hold it and click him as soon as I felt him tighten his jaws. And over time I lessen the pressure of my hands until he was holding the dowel on his own. The clicker is invaluable in this process because it allowed me to pinpoint various key moments in this training: the moment when he first accepts the dowel in his mouth, the moment he first tighten his jaws, the moment he commits to holding the dowel on his own. Other people do it differently but the way I did it, there was not a lot of "WANT TO" in the exercise. Rather, right from the start, I impressed on him that he "HAS TO" do it but nevertheless the clicker plays a very important part in the whole teaching process. 

 

First of all, I don't know why you would deliberately choose to be so forceful with the dog if you didn't have to be.  I could understand it, even if I didn't agree with it, had you said that you tried it the other way and failed.  But, as an initial attempt, what would have been wrong with simply teaching it using the back-chaining?  Why are people so afraid to try to teach positively, and why is there such a lingering, yet false, belief that positively trained dogs don't understand the concept of having to perform a behavior accurately and on cue.  The only difference is between forcing the dog to do it to avoid unpleasantness, or creating a situation where the dog really wants to do it because he might earn something he wants (which does not imply having to use food, tug, toy, or even a verbal acknowledgement, for every single repetition!!!!).

I think it just boils down to personal choice.  Anyone can justify any training method.  Also if it is a new behavior, and something as critical as retrieves (make up 40/100 ob points), people like me who are novices are more inclined to do it the way their TD and club peers are doing it because there's more support.  I chose to backchain it (and did it differently than what my TD would have instructed me to do) and stand by my choice and will train all future dogs using the backchain/freeshape method.  While I trained it my way, I watched two people train it by putting the dumbbell in the dog's mouth.  Maybe it's a coincidence but now both of those dogs are really inconsistent with picking up the dumbbell.  Unless the handler shows some pressure right away, they are often slow/reluctant to pick it up, or just blow off the exercise altogether.  Both dogs are also more likely to drop the dumbbell prematurely unless being shown pressure.  They are not bad dogs but what's clear to me is that they lack the motivation to do the exercise and also are still somewhat confused about what the behaviors really are.  But I'm sure there's thousands of dogs trained this way that do it just fine.

spiritdogs

First of all, I don't know why you would deliberately choose to be so forceful with the dog if you didn't have to be.  I could understand it, even if I didn't agree with it, had you said that you tried it the other way and failed.  But, as an initial attempt, what would have been wrong with simply teaching it using the back-chaining?  Why are people so afraid to try to teach positively, and why is there such a lingering, yet false, belief that positively trained dogs don't understand the concept of having to perform a behavior accurately and on cue.  The only difference is between forcing the dog to do it to avoid unpleasantness, or creating a situation where the dog really wants to do it because he might earn something he wants (which does not imply having to use food, tug, toy, or even a verbal acknowledgement, for every single repetition!!!!).

It really just depends of the dog. I'm not saying that this is the way to teach hold/retrieve for every dog. Actually, for most dogs the clicker retrieve is the best way to go. But then again - this goes back to an earlier part of this discussion - what's a lot of pressure to one dog is nothing to the other dog; how one dog reacts to correction is different than how another reacts.

As for why I would deliberately choose to be forceful with a dog when I don't have to be ... well, that probably deserves a new thread for discussion. I'll just say that I think it's not the end of the world to make your dog do a few "HAVE TO"s from time to time. As long as the relationship is good and the bond is strong, an occasional conflict is not the end of the world. In fact, used appropriately, conflict can function as a very valuable teaching tool.

Liesje

So, using a clicker to freeshape = clicker training and using a clicker incorrectly = training with a clicker?  IMO if the clicker is not being used correctly it's not really training at all, it's more like fumbling with a clicker.  Either it's being used correctly or not.

Not at all. I said that it was more than a marker. I don't even use a clicker with Kivi, and I don't freeshape with him but target instead. I still call it clicker training because the things I consider important about clicker training are still there. I still have a short, sharp marker, I still use rapid fire and concentrate on high reward rates, I still create a sense of anticipation and excitement about training, and I don't correct or punish during training. I am about helping dogs make good choices. They don't get it wrong. I get it wrong. The responsibility for their behaviour is entirely mine.

When I was "training with a clicker" it was just a marker. My reward rate wasn't high and I tried to give my dog lots of help. I still used a lot of interruptors and conditioned punishers. It wasn't until I stopped doing that that I realised how many opportunities to reward good behaviour I had missed. It was quite liberating ditching the conditioned punishers.