Sanji 4 months Recall

[sanjibasenji]

@yahtzee92 said in Sanji 4 months Recall:

Dang.. killer job. Can you provide a link to the training you're doing? Thanks!

Well, there's not a single link. She has a lot of free content on Youtube.
https://www.youtube.com/channel/UCfaslnxiHL2JHPEyV8goR5w

She also has an online course with two levels. One level is just the course (50 or so videos, checklists, schedules, expectations, etc., step by step), $167 if I remember correctly. The second Pro Level Plan is that course, PLUS personal unlimited written questions via their facebook page, and 3 live office hour zoom discussions with a moderator and trainer to answer any questions, about $400. HACK: I got the basic course and found one can upgrade to the Pro level for $75 per month. So I did that for one month, which I found useful to answer written questions, which were answered within hours.

https://www.howtotrainadreamdog.com/puppy-perfection-course/

I recommend scrolling through the entire page, though it seems a bit long. But it's all genuine.

The video on two-person recall training I think was the most helpful. But the work on "stay" training at the same time, and then doing a sit, stay, walk away, and recall, and you're making progress. It's SLOW though. Don't have expectations beyond what is realistic. Always try to make it FUN. That is a KEY element.

PERSONAL VIEW
It's a great foundation course. I learned 9 years ago from a local professional trainer. This was more in-depth and MUCH more convenient. The foundation skills, like the "bump game" don't seem important, but then later one discovers how helpful they are when one needs to get the pup accustomed to nail trimming, ear and teeth cleaning, etc. So, there's a lot there.

Limits: I have found that my basenji, at six months, reached a point where he's not as food motivated as he was, and consequently with "heel" training (covered in her course) we reached a point of marginal returns on the training I was doing. My expectations are probably part of it, and if I just kept at it over the next year, he'd improve and get there.

Nevertheless, about two weeks ago (weeks after I made that clip) I started with some electronic collar training, which is NOT shock training. One uses vibrate, beep, and lowest level electrical stimulation just to get the dog's attention, never to elicit a yelp or pain (in contrast to an electric fence, which, by the way, is working GREAT), and always in combination with positive reinforcement (treats or praise), as demonstrated in the video (link below). I also have found that my verbal praise works sometimes better now during heel training than a treat does, in part because he's understanding better with age what I want, and in part because he's more responsive to praise as he ages. He wasn't as emotionally in-tune or mature as a young puppy.

So the ecollar work has helped with heel work right away. I vibrate when he starts creating tension, and praise when he looks me in the eye and returns to my side. Sometimes, with greater distractions, I have to move up to sound. (He already knows sound, vibration, and painful shock because that's associated with getting close to the electric fence. But that's not how it's really supposed to work. Per the video, one is training with the ecollar using positive reinforcement, not for negative reinforcement -- that he might get a painful shock after the sound if he continues the behavior.

I found that the ecollar does does help with recall too, for sure in the short run. I don't know if this ecollar training for recall will "stick" and will work without the collar down the road, which is the goal. I'll find out over the next 6 months. But, as shown in the video, when he's intensely distracted and doesn't listen or hear me call or whistle, the vibration seems to do the trick, if one has done all the pre-training that we did with the online course. When I use the vibrate in those special situations, he'll turn to me and then will come. Or he may stall, and I'll vibrate again and he comes. About 95% of the time when he's with other dogs he comes by whistle or name. It's just when he's intensely distracted that the ecollar seems to help.

Qualifications: I never even try to recall when he darts off the path to chase something. I don't need to. He always comes back looking for me within a minute or so. I mean always. I'm very very comfortable going off leash in parks (would never do so in town). He gets along with all people and dogs. (The trainer we worked a few months ago for swimming lessons said he's the nicest Basenji she ever met.)

Ecollar training: https://youtu.be/SmBhKFdtMAY

[Zande]

Strange - I have never used anyone's training methods, weird collars or gimmicks and all my dogs, going back generations, have had super solid recall. Obviously only in woods and fields, never near highways or even country lanes. Main thing is to start them young enough and it stays with them for life.

I am very happy to say several of my boys' siblings are also able to run free. One of these days we will get them together for a Free Basenji Jamboree.

[eeeefarm]

@sanjibasenji said in Sanji 4 months Recall:

training with the ecollar using positive reinforcement, not for negative reinforcement -- that he might get a painful shock after the sound if he continues the behavior.

For clarification, that would be "positive punishment" not negative reinforcement.

[DonC]

@sanjibasenji said in Sanji 4 months Recall:

Qualifications: I never even try to recall when he darts off the path to chase something. I don't need to. He always comes back looking for me within a minute or so. I mean always. I'm very very comfortable going off leash in parks (would never do so in town). He gets along with all people and dogs. (The trainer we worked a few months ago for swimming lessons said he's the nicest Basenji she ever met.)

This qualification is the problem IMO. Yes a Basenji will stay with you and/or consistency recall ... until they don't. The prey drive can be too strong. We've had a Basenji with what seemed to be great recall chase a deer and only "come back" when the deer circled back.

[eeeefarm]

@donc said in Sanji 4 months Recall:

a Basenji will stay with you and/or consistency recall ... until they don't.

True of many dogs. If you want to be sure of your recall, it takes a lot of distraction training to proof it. And, IMO, there has to be a consequence, however mild, when the dog does not obey. Even with a biddable breed, like a Border Collie or German Shepherd, you really need to experiment and increase distractions until you are sure of him/her. A rock solid recall keeps your dog safe, but is definitely not an easy thing to teach. And a dog that is "perfect" while young may not be so great when puberty hits and they feel more self confident.....just like teenagers entrusted with your car!

[sanjibasenji]

@eeeefarm said in Sanji 4 months Recall:

For clarification, that would be "positive punishment" not negative reinforcement.

Not sure it's that. The vibration and sound is not punishment, the electrical stimulation is not really punishment either, though it causes a degree discomfort if set appropriately: at the lowest setting that the pup notices.

In contrast, at present, I have the setting for the electric fence collar to near max shock to cause pain, and that is definitely punishment. But the initial stages of training for the electric fence are similar to the ecollar with the remote in that the setting is set low while the pup learns the boundaries, doesn't feel pain, and is rewarded for moving away. I suppose that this would qualify as "positive punishment." The setting is gradually increased to the point where the pain of transgressing the boundary is not worth doing so. There's little "positive" at this point, except for praise after the yelp for staying in the yard. Fortunately, by this time this happens rarely.

[sanjibasenji]

@donc said in Sanji 4 months Recall:

We've had a Basenji with what seemed to be great recall chase a deer and only "come back" when the deer circled back.

Pup individuality may be a factor. Sanji doesn't seem to have that same drive. He only darts after squirrels for a few yards when were in the park or single track mountain biking, and then ceases when they're up the tree, and returns to me. When he sees squirrels, rabbits, and cats in our yard looking out our large rear window, he just stares. No chase whatsoever. My prior dog Kai -- a cockapoo - would dart out the doggy door after them faster than lighting. And when Sanji first sees another dog, cat, or deer, in parks our public lands, he sits and watches for a bit, more so as he ages. It seems odd to me, but perhaps he's just wisely assessing the situation. When he recognizes it's a friend (dog or human), he waits for them to get at least half way before he proceeds.

[sanjibasenji]

@zande said in Sanji 4 months Recall:

Main thing is to start them young enough and it stays with them for life.

I concur with that.

[eeeefarm]

@sanjibasenji said in Sanji 4 months Recall:

Not sure it's that. The vibration and sound is not punishment, the electrical stimulation is not really punishment either, though it causes a degree discomfort if set appropriately: at the lowest setting that the pup notices.

It's a terminology thing. Operant conditioning is well defined. Positive means to add something that effects behaviour, e.g. positive reinforcement you give a treat or whatever to reinforce the wanted behaviour. Positive punishment, you add a stimulus of whatever kind that makes the behaviour less likely to occur. If it works to decrease the likelihood of the behaviour it is "punishment". It does not have to be severe, it just needs to have an effect.

Negative reinforcement, you are removing something, i.e. taking away a stimulus when you get the desired effect, negative punishment you take away something the animal wants, which decreases the likelihood of the behaviour.

[sanjibasenji]

@eeeefarm said in Sanji 4 months Recall:

It's a terminology thing.

Interesting. Yes, well defined as you reference, but that is specialized terminology that most people don't use. I used the terms according to colloquial usage, in which case, "punishment" is to cause "suffering, pain, or loss that serves as retribution."

[eeeefarm]

"Negative" combined with reinforcement or punishment is not a term one generally hears in everyday speech, it's specific to operant conditioning, so it's useful to understand the terminology. I agree it is used incorrectly often, because people don't understand the meaning of the terms correctly. This is the result of the hyping of "all positive" training, which in practical use is not what trainers actually do, since most "positive" trainers use negative punishment, but never mind.....I just wanted to clarify the meaning, so that people don't keep getting it wrong.

[elbrant]

@sanjibasenji said in Sanji 4 months Recall:

"punishment" is to cause "suffering, pain, or loss that serves as retribution."

I don't think that these collars are meant to be punishment, positive, negative, or otherwise. I think the slight vibration is only meant as an indication that they are near the established perimeter. A warning, at best, to let the dog know they should go further. Same as our clicker, or a vocalization, "uh - uh", would be. There is no punishment intended. They haven't done anything wrong. There is no pain (not that punishment should be painful in the first place). Just a simple, "learn your boundaries", or "I expect you to stay in your yard."

[eeeefarm]

"Punishment" in the Operant sense, is anything that reduces the likelihood of the behaviour being repeated. It is the opposite of "Reinforcement", which means anything that increases the likelihood of the behaviour being repeated. While we are at it, "positive" means something is added, "negative" means something is taken away. I know I am belabouring the point, but the terms are quite specific and incorrectly used more often than not by both trainers and lay people. Which wouldn't matter except it causes confusion when people use the term incorrectly and so are misinterpreted by others who do understand the definitions....

[sanjibasenji]

@eeeefarm said in Sanji 4 months Recall:

"Punishment" in the Operant sense, is anything that reduces the likelihood of the behaviour being repeated. It is the opposite of "Reinforcement", which means anything that increases the likelihood of the behaviour being repeated.

I don't think you're belaboring the point; I appreciate the specialized terminology. The above makes sense to me intuitively.

BTW, "negative reinforcement" and "positive reinforcement" is something people in the US do use in every day language. I assume that training experts here also use the same scientific terminology that you are explaining. The difference is interesting. Negative reinforcement in the colloquial usage here doesn't mean here to remove something, but on the contrary, to add a negative experience, as you describe "punishment" in the scientific usage in order to dissuade behavior. Positive reinforcement, as with "reinforcement" in the scientific definition, means the same: to add a reward to encourage the behavior.

So, the following doesn't make sense to me given the above:

... "positive" means something is added, "negative" means something is taken away.

You did explain this however. "Negative reinforcement, you are removing something, i.e. taking away a stimulus when you get the desired effect, negative punishment you take away something the animal wants, which decreases the likelihood of the behaviour."

I'm not certain of the difference is between the two. Is the former to reinforce repetition of a desired behavior? For example, taking away a toy that a dog wants until it sits (desired behavior)? And would negative punishment be taking away a toy that a dog wants until it stops barking (undesired behavior)?

[eeeefarm]

@sanjibasenji said in Sanji 4 months Recall:

. Is the former to reinforce repetition of a desired behavior? For example, taking away a toy that a dog wants until it sits (desired behavior)? And would negative punishment be taking away a toy that a dog wants until it stops barking (undesired behavior)?

Not exactly. Negative reinforcement usually refers a stimulus that compels the behaviour, removing it when you get the behaviour. e.g. with horses it's very clear.....apply pressure to the reins or the sides of the horse, remove the pressure when the horse responds. An example with dogs would be e-collar training, for instance when teaching the recall, where you use a low level stim which goes away when the dog comes to you. In practice, you let the dog walk around on a long line, apply the stim at the lowest level he can feel, then use the line to draw him in to you while you turn off the stim. Very quickly the dog learns that that annoying sensation goes away when he comes to you. As you can see, the reinforcement is taken away when the animal responds correctly.

Negative punishment is taking away something the animal desires when he doesn't behave as required. Example, not giving the treat he knows you have because he did not respond to your command. Another example, removing his freedom (putting him in a crate) when his behaviour is unacceptable. A "time out" is negative punishment. In both of your examples you are describing negative punishment, as you want the undesired behaviour (refusing a command, unnecessary barking) to cease.

[sanjibasenji]

@eeeefarm said in Sanji 4 months Recall:

not giving the treat he knows you have because he did not respond to your command. Another example, removing his freedom (putting him in a crate) when his behaviour is unacceptable.

This was illuminating. Thanks!

As I understand it, they are flipsides of each other. Giving the treat positively reinforces the desired behavior, not giving it negatively reinforces the non-responsive behavior.