Conditioned “NO” Beats Reward-Only Every Time
Ivan BalabanovIvan Balabanov · 영상 · 14분
Conditioned “NO” Beats Reward-Only Every Time
0:00
All right guys, I'm gonna show you something interesting with one of my dogs today. Many dogs love chasing squirrels, rabbits, cars, bikes, you name it. But not all dogs are the same. Some dogs chase for fun. It's kind of the thrill of the run. It's only the chase part alone of the complete predatory sequence. So they may stop at the tree, bark excitedly, but ultimately they kind of bounce back to you. And this is what I call play predation because there is really no intent to kill. Now there is this other type of dogs like the dog that I'm going to show you. They chase with purpose to kill. They lock in hard. They stay frozen in a stealth mode for an hour waiting to strike or just choose option B and go in a wildest chases disregarding any obstacles along the way. Never really thinking about self-preservation or anything. This is the full predatory drive. The motivation is deeper and harder to override because succeeding, the catching and the killing is hugely self-reinforcing. The dog literally pays itself with adrenaline and completion of the instinct. Now, let me tell you a little bit about the dog that you're going to see shortly. Her name is Kalinka. She's a three-year-old Malinois female. And I got her back. I bred her and I got her back when she was two years old. So she had already some history. It was the reason she came back was because she was just too much dog, too high energy, just a young and active working Malinois that doesn't really fit well living in a condo in the city. However, she mastered the skill and the skill is hunting and catching and sometimes killing squirrels. So the reason I'm giving you that background is because we need to make distinction between the two types of dogs, the one that play chase and the one that have very different intention. The first three ways to control such behaviors are extremely limited and and they typically rely on few protocols. Descentization, counter conditioning, differential reinforcement and what you would call predation substitute protocols. However, none of this is effective in controlling dogs that chase with the purpose to kill and have history of doing so. Let's look at the desensitization for example. It's a protocol that gradually exposes the dog to the trigger in a bigger distance or with a low level of intensity. So the distance for example is big enough to where the dog doesn't react yet and gradually we start getting closer and closer to the trigger, a squirrel in our example. Now, if you believe in anecdotal stories and want to give this a try, go ahead. But keep in mind that the
3:50
sensitization is meant for very different purposes. It's typically for fear based reactions and phobias. In predation, the dog is not afraid. They are highly motivated by their instincts to kill. This is genetic programming that is really deep. Anybody can desensitize a dog to a squirrel that is 200 feet away. However, the moment that squirrel bolts under their nose, the biological threshold is crossed and game over. That dog is gone. The counter conditioning approach. What does it mean if you are not familiar? Basically, is pairing the exposure to the squirrel with high value treats. The idea, the intention is to change the dog's emotional response from hunt to look at the owner for food. Next, we have the differential reinforcement. Differential reinforcement alone. It's also not successful. The idea here is to reinforce different behavior other than chasing the squirrel like sit or watch me, whatever you can think of. Now, there is also the predation substitute training. And the idea here is to encourage the dog to chase and catch a toy instead of the live animal. The problem is that if the reinforcement present is presented after the chase starts, the dog is already locked in. The predation motor pattern system is activated. And at this point, no other reinforcement are acknowledged by the dog. Done. The dog is not the only one. Like you can go as far as even to offer another squirrel from your pocket. It doesn't matter. The dog is already locked on the target and it's mentally, if not physically as well, gone. So, let's be realistic. Wildlife, cars, bicycles, they don't show up on cue whenever you're ready to deal with them. That's not real life. Some of the approaches that I just mentioned, they will be successful in controlling a dog that is play chasing. But they will fail reliably with the dogs that have different intention and history. Let me tell you why it doesn't work. The first reason is what I call absence of cost. The force-free trainers basically refuse to attach a cost to the behavior. It's like dancing around the real problem, trying to convince the dog that they have a better, more interesting activity to offer than killing the squirrel. Certain dogs find the adrenaline of the kill far more rewarding than your suggestion. They simply outbid you and take the squirrel. End of story. So, what I'm going to show you now is how I can control behavior extremely reliably without the complicated protocols, without toys, food bribes, without any colors or long lines on the dog. Just a verbal cue that stops the dog every time.
7:27
In my case, I use the word no, but that's really a word that I choose. It doesn't matter which word you will pick. That cue is called conditioned or secondary punisher in the textbooks. Okay, enough of the talking. So, let's step outside so I can show you the setup I have created to demonstrate how conditioned punisher works. And then I will take a linker out on a squirrel hunt. So, let's do it. Natalia is going to help us. So, typically this is where the squirrels hang around. Do a little movement on the squirrel. Yeah, good. So, that's basically the idea. There is no chance that she will think this is a setup. Unless we let her explore. She knows that there are always squirrels here. So, they are either here or by the bird feeders on the other side. Or, of course, anywhere around. So, I don't know. Sometimes she may come and really make a very quick run. And sometimes she can come in that very stealthy mood and trying to stalk and catch. So, the plan is that she sees this squirrel, she goes for it, and I tell her no. And what I'm expecting to happen is that she will not continue. That I will actually be able to stop her. This is very different than, again, a playful dog that's, you know, having different intentions. Then grab and kill the squirrel. So, I will talk to you later why I can say no and I can control the dog. Okay. So, let's let her out and see how that goes. Kalina. No! Good. Come. Hey, Kalinka. So, now she knows where it is. She isn't moving. However, she does not go. The reason she does not go is because I have already told her that this is off limits. And it's very important. This is what we are talking when we talk about behavior control. Kalina, come. You see that avoidance, right? Kalinka. Okay. However, now I'm going to let her be free and she will hunt anything else. Again, I can let her go and have fun. I can let her go and catch something. Or, I can control the behavior verbally without anything on her. I can control the behavior. That comes with legit training techniques. Not rainbows and unicorns. No positive reinforcement, differential reinforcement programming can accomplish this. Kalina.
11:22
Hey, mama. Come this way. Okay. Okay. Okay. no good good mama come my squirrel hunter she's very skilled at this yeah here is how it works the work now is basically paired with an aversive consequence over time the signal alone stops the dog you can you can think of this as a transfer of function the no signal basically stops being just a sound and starts functioning as the consequence itself when used fairly with clear timing and the proper order of classical conditioning it will provide fast reliable control of behavior at the most critical times when you need it it gives even the most problematic dogs true freedom so they can enjoy off-leash walks without the constant stressful management now for the all force free advocates that believe that they can do that it's very much impossible to stop a dog that has history of actually hunting and killing small animals through differential reinforcement through counter conditioning through the synthesization to have that kind of control of their behavior this is absolutely not possible that's all i wanted to show you leave comments and i'll be happy to discuss further
0:10
Alright guys, I'm going to show you something interesting with one of my dogs today. Many dogs love chasing squirrels, rabbits, cars, bikes, you name it. But not all dogs are the same. Some dogs chase for fun. It's kind of the thrill of the run. It's only the chase part alone of the complete predatory sequence. So they may stop at the tree, bark excitedly, but ultimately they kind of bounce back to you. And this is what I call play predation. Because there is really no intent to kill. Now, there is this other type of dogs, like the dog that I'm going to show you. They chase with purpose to kill. They lock in hard. They stay frozen in a stealth mode for an hour waiting to strike. Or just choose option B and go in a wildest chase is disregarding any obstacles along the way. Never really thinking about self-preservation or anything. This is the full predatory drive. The motivation is deeper and harder to override because succeeding the catching and the killing is hugely self-reinforcing. The dog literally pays itself with adrenaline and completion of the instinct. Now, let me tell you a little bit about the dog that you're going to see shortly. Her name is Kalinka. She's a three year old Malinois female, and I got her back. I bred her and I got her back when she was two years old, so she had already some history. It was the reason she came back was because she was just too much dog, too high energy, just so young and active. Working Malinois. That doesn't really fit well living in a condo in a in the city. However, she mastered the skill and the skill is hunting and catching and sometimes killing squirrels. So the reason I'm giving you that background is because we need to make distinction between the two types of dogs. The one that play chase and the one that have very different intention. The force free ways to control such behaviors are extremely limited, and they typically rely on few protocols. Desensitization, counter conditioning, differential reinforcement and what you would call predation substitute protocols. However, none of this is effective in controlling dogs that chase with the purpose to kill and have a history of doing so. Let's look at the, desensitization for, for example. It's a protocol that gradually exposes the dog to the trigger in a bigger distance, or with a low level of intensity. So the distance, for example, is big enough to where the dog doesn't react yet. And gradually we start getting closer and closer to the trigger. A squirrel in our example. Now if you believe in anecdotal stories and want to give this a try, go ahead. But keep in mind that desensitization
3:52
is meant for very different purposes. It's typically for fear based reactions and phobias. In predation the dog is not afraid. They are highly motivated by their instincts to kill. This is genetic programing that is really deep. Anybody can desensitize a dog to a squirrel that is 200ft away. However, the moment that squirrel bolts under their nose, the biological threshold is crossed and game over, that dog is gone. The counter conditioning approach. What does it mean if you are not familiar? Basically is pairing the exposure to the squirrel with high value treats. The idea, the intention is to change the dog's emotional response from hunt to look at the owner for food. Next we have the differential reinforcement. Differential reinforcement alone is also not successful. The idea here is to reinforce different behavior other than chasing the squirrel. Like sit or watch me. Whatever. Whatever you can think of. Now, there is also the predation substitute training. And the idea here is to encourage the dog to chase and catch a toy instead of the live animal. The problem is that if the reinforcement is presented after the chase starts, the dog is already locked in, the predation motor pattern system is activated. And at this point, no other reinforcement are acknowledged by the dog. Done. Like you can go as far as even to have a, offer another squirrel from your pocket. It doesn't matter. That dog is already locked on the target, and it's mentally, if not physically as well, gone. So, let's be realistic. Wildlife, cars, bicycles. They don't show up on cue whenever you're ready to deal with them. That's not real life. Some of the approaches that I just mentioned, they will be successful in controlling a dog that is play chasing, but they will fail reliably with the dogs that have different intention and history. Let me tell you why it doesn't work. The first reason is what I call absence of cost. The force free trainers basically refuse to attach a cost to the behavior. It's like dancing around the real problem, trying to convince the dog that they have a better, more interesting activity to offer than killing the squirrel. Certain dogs find the adrenaline of the kill far more rewarding than your suggestion. They simply outbid you and take the squirrel. End of story. So what I'm going to show you now is how I can control behavior extremely reliably without the complicated protocols, without toys, food bribes, without any collars or long lines on the dog. Just a verbal cue that stops the dog every time. In my case, I use the word no. But, that's really word that I chose. It doesn't matter which word you will pick.
7:37
That cue is called conditioned or secondary punisher in the textbooks. Okay, enough of the talking. Let's step outside so I can show you the setup I have created to demonstrate how conditioned punisher works. And then I will take Kalinka out on a squirrel hunt. So let's do it. Natalia is going to help us. So typically, this is where the squirrels hang around. Do a little movement on the squirrel. Yeah. Good. So that's that's basically the idea. There is no chance that she will think this is a setup. Unless we let her explore. She knows that there are always squirrels here. So they're either here or by the birdfeeders on the other side, or of course, anywhere around. So I don't know. Sometimes she might come and really make a very quick run, and sometimes she can come in that very stealthy mood and trying to stalk and catch. So the plan is that she sees the squirrel. She goes for it and I tell her “No!” And what I'm expecting to happen is that she will not continue, that I will actually be able to stop her. This is very different than, again, a playful dog that's, you know, having different intentions. Dumb grab and kill the squirrel. So I will talk to you later. Why I can say no and I can control the dog. Okay. So let's let her out and see how that goes. Kalina. No. Good. Come. Hey! Kalin. So now she knows where it is. She sees it moving. However, she does not go. The reason she does not go is because I have already told her that this is off limits. And this is very important. This is what we are talking when we talk about behavior control. Kalina, come. You see that avoidance, right? However, now I'm going to let her be free and she will hunt anything else. Again, I can let her go and have fun. I can let her go and catch something. Or I can control the behavior verbally without anything on her. That comes with legit training techniques, not rainbows and unicorns. No positive reinforcement differential reinforcement programing can accomplish this. Kalina. Hey, mama. Come this way. Kalina. Hi. Come this way. Come. No. Good. Go ahead. Mama. Hahaha! Come. Oh! Oh, okay. Okay. My squirrel hunter and she very skilled at this. Yeah. Here is how it works. The word “no” is basically paired with an aversive consequence. Over time, the signal alone stops the dog. You can think of this as a transfer of function. The “no” signal basically stops being just a sound and starts functioning as the consequence itself. When used fairly with clear timing and the proper order of classical conditioning. It will provide fast, reliable control of behavior at the most critical times when you need it.
13:30
It gives even the most problematic dogs true freedom, so they can enjoy off leash walks without the constant stressful managing. Now for the all force free advocates that believe that they can do that. It's very much impossible to stop a dog that has a history of actually hunting and killing small animals through differential reinforcement, through counter conditioning, through desensitization to have that kind of control of their behavior. This is absolutely not possible. That's all I wanted to show you. Leave comments and I'll be happy to discuss further.