Any Tips For Keeping Attention During Distractions

[SpecTraining]

That's the reason you will find most if not all high level performance trainers use a combination of both positive and negative reinforcement. Positive only methods have been tried and failed training police dogs, security dogs and Schutzund sporting dogs and one's that claim to have trained in positive only methods, no one would ever know the truth or tell the difference in the dogs performance once trained. If positive only methods worked best to achieve the greatest reliablity and performance, the people dedicated to train and win in these disciplines at high level competition wouldn't be training with prongs and E Collars as they do and would be concentrating on positive only methods which they don't is the way I see things.

Interestingly the top level agility trainers don't use physical corrections

Top level as in world level or club level here???. Most use E Collars nowdays if you research it

[Aidan3]

Top level as in world level or club level here???. Most use E Collars nowdays if you research it

Either, and no they don't. I'm sure you will find some if you research it. The same as you will find top level SchH, police service, customs, dual-purpose, Mondio, SAR, herding, obedience, tracking, and personal Service Dogs who are clicker trained (what you might erroneously refer to as "purely positive" training) - if you look hard enough.

This debate is dead, and can only be forcefully revived by selective reporting and rigid opinions. Let's discuss something that isn't painfully obvious and actually matters.

[JulesP]

Not sure how this pissing contest is helping the OP. And well it has been done to death. Having the strong urge to use an ecollar and not on a dog!

[SpecTraining]

That's the reason you will find most if not all high level performance trainers use a combination of both positive and negative reinforcement. Positive only methods have been tried and failed training police dogs, security dogs and Schutzund sporting dogs and one's that claim to have trained in positive only methods, no one would ever know the truth or tell the difference in the dogs performance once trained. If positive only methods worked best to achieve the greatest reliablity and performance, the people dedicated to train and win in these disciplines at high level competition wouldn't be training with prongs and E Collars as they do and would be concentrating on positive only methods which they don't is the way I see things.

No-one was arguing otherwise. My apologies to the OP for letting this get off-topic by my participation. I had only intended to answer specific questions related to the original post and had hoped that this would not turn into a method debate. Least of all a debate of extremes ("purely positive", police dogs, high level competition dogs etc). Somehow Neville (SpecTraining) seems to always bring it back to either an imagined "purely positive" or "all GSDs must have SchH titles" debate, which is a shame because I honestly believe that Neville has more to offer than that. The original question he posed was insightful, unfortunately the forest seems to have been lost for the trees.

Some of you may remember the transition from aversive based training to positive reinforcement was largely based upon aversive training methods assumed as being cruel on the dog and had little to do with achieving better performance and reliability, it was based on training without aversive methods being used and training on a more humane platform.

You specifically asked for reasons other than these. What was happening back then is way out of context with what is happening today.

This discussion has nothing to do with GSD's and Schutzhund titles George (Aidan), it has to do with what has transpired in the thread from my suggestion to add a correction and the "purely positive" people wanted to voice an opinion in opposition to using a correction in the OP's situation and the reason for opposing a correction is what I asked in hope of an informed and sensible discussion which is relative to what the OP has asked for, "tips to re-gain focus".

My point of mentioning high level competition and police dogs is that reliability is paramount, so if purely positive methods worked best to attain the required reliability the trainers working in these disciplines would use only positive methods, yeah???. They wouldn't use aversives and jeopardise the training of a dog that is worth a lot of value in that case if the get it wrong, and purely positive methods is not how these dogs are trained "fact". They are trained in a combination of both methods in the same fashion as explained by Steve Courtney from K9 Pro in the link I provided, did you read that???.

What happens today regarding purley positive training is still the same reasoning as it was intitially, injuring the dog, handler fallout, cruelty, etc as a basis for why aversive methods should be avoided, and my question irrespective of this still remains unanswered where no one yet has provided a reason in terms of reliability how positve reinforcement is assumed to provide a more reliable result than a correction

[huski]

SpecTraining. Have you missed the part where Aidan has said that he is not totally opposed to using corrections? And has actually used them himself? And that the point he was making was not that he would never use corrections or thinks that they should never be used but simply that he would not necessarily use them in this instance?

[Mrs Rusty Bucket]

Are some of us confusing "negative re-inforcement" with "physical correction"?

Incidentally the World Championship level agility trainer whose semimar I'm going to as a spectator, would evict any handler that uses the word "NO" at their dog. So I might be in trouble. But yelling "NO!!!" at my dog has never been effective at stopping any behaviour of hers or even getting her attention.

I think positive and negative are like training a dog to find something (what you want them to do), using "hot" (good) and "cold" (oops). Ie "yes" or "good" or "click" means she's getting it right and she's going to get a reward (pat/food/play etc) a positive re-inforcement. And "oops" or "nope" or "try again" means she's not going to get a reward for that behaviour, try something different. Not using the word "cold" means finding the right behaviour or hidden treat takes much longer. Imagine playing the game using only "hot" if a step in the right direction is made and not "cold" if a step away is made.

Like my recall training. Recall had recently gone to hell. Dog comes out of oval hedge with roast lamb shank bone (arrgghh). Self rewarding. But I have uber treat. So eventually I get her attention without rushing her into swallowing bone, and give the release command (eg "thank you") and throw a handful of highly visible uber treats in her general direction. She has to drop the bone to get the treats (fresh cooked liver and bread cubes). I win. She wins but she doesn't know it exacty. Today I practice recall with the liver. She comes when I call, she gets some liver. If she doesn't - I "hide". Which is slightly more traumatic for her than "oops", she gets really upset when she doesn't know where I am so she tends to be more responsive and quicker next time I call, and that gets her fresh cooked liver treats, which she remembers, and so does every other dog that gets some. I now "own" about five dogs with perfect recall...

And I didn't have to use an e-collar once. But I did use a negative reinforcement once.

Incidentally another dog trainer, I know - described training his dog with an e-collar for obedience comp and winning but when he looks back at the video now, he isn't proud of it - because that dog is clearly uncomfortable and unhappy, its tail and head are down, and it's cringing the whole way round the course. Where as the ones trained with loads of positive reinforcement - have their heads up and their tails wagging the whole way round.

It is possible to get fast results with an ecollar as a very reliable delivery of "oops", but it's also possible to over do it and get a cringing unhappy insecure dog.

Oh and for keeping a dogs attention in the face of distractions - one dog I know will be ultra attentive if the owner has blue vein cheese under her fingernails. Is that cheating?

[huski]

MRB, this is getting OT now, but I think what's really important is how you use positive reinforcement. I have see PR/clicker trained dogs working like crap in the obedience ring - no drive, focus, bored etc.

Edited October 4, 2010 by huski

[Tiggy]

Not sure how this pissing contest is helping the OP. And well it has been done to death. Having the strong urge to use an ecollar and not on a dog!

[Diva]

I find that hardly any DOLers are what I would call 'purely positive'. I know there are a few out in the 'real world' but I don't hear them on DOL. But I see many many posts arguing against this invisible cohort. So many that they seem to derail a lot of threads and ruin a lot of opportunities for real discussion.

Plenty of DOLers may argue against the use of corrections in some specific situations, or against some specific sorts of corrections, but I am struggling to think of too many who say they'd never correct or think others should never correct.

So who are these purely positive souls?

Edited October 4, 2010 by Diva

[Mrs Rusty Bucket]

Huski

I saw dogs trained in drive with mostly positve re-inforcement, and traditionally trained dogs doing all sorts of not quite right stuff (failing) in the competition ring yesterday. I think one class had a single pass or no pass at all. Even one dog which won the day before, failed yesterday. And it looked superb for most of the heel work, head up, tail wagging, stuck to the handler's leg. But broke the stay before recall. Apparently it was upset by a sibling's barking. Sigh.

Other dogs, just like you said, no drive, focus and bored.

Though my fave was one that was working really well but decided to do the broad jump by going around to the side and then tip toeing between the slats to the owner instead of jumping over. And this is a dog that likes agility. Butchering an agility element in the obedience ring. It looked so pleased with itself too.

I blamed the weather. Hot and windy - first ikky day this season. So many dogs were affected. With so many different training methods failing. Or dogs are not robots.

PS - I know a delta trainer face to face. She says I should not say "no" or "oops" to my dog, and should not squirt water at or near her by way of distraction when she's barking and shouldn't be. Delta trainer's dogs are really laid back and well behaved. I've seen her out training a puppy and owner recall with a long line. It wasn't going to happen instantly but the puppy was having a lovely time. But I'm pretty sure not getting a treat is still negative/adversive for the puppy even if it doesn't get told "oops" when it does the wrong thing.

Edited October 4, 2010 by Mrs Rusty Bucket

[SpecTraining]

SpecTraining. Have you missed the part where Aidan has said that he is not totally opposed to using corrections? And has actually used them himself? And that the point he was making was not that he would never use corrections or thinks that they should never be used but simply that he would not necessarily use them in this instance?

Yes, I did get that Huski and asked why he wouldn't use a correction in the circumstances which boiled down to the same "purely positive" perspective which IMHO is nonsence for the most part, and I share the same opinion regarding the side effects of correction in the same way described by Nekhbet in this thread and K9pro's link I provided.

Dogs that have learned no consequence for their actions shows up in other behavioural traits. A loose leash walk can established with positive reinforcement and the dog can walk beautifully until it reaches a major distraction and will shoot the end of the leash and pull. A dog trained to loose leash walk from aversive methods for example Koehler leash work, knows what happens at the end of the leash and they don't display the same behaviour with any distractions, there is a marked difference between the two methods in reliability terms. If you have ever walked a dog trained in Koehler leash work and walked a dog trained in positive methods, you would know the difference I am referring to when distractions are elevated, it's like chalk and cheese

[corvus]

I'd love to know, Diva. I keep looking for them, but dammit, I just can't find any.

To the OP, just be aware that increasing distance between your dog and a distraction they want to get closer to may increase their frustration, which I always find to be quite detrimental and avoid almost as much as I avoid punishments. A frustrated dog can get more aroused, and that just makes it harder to get their attention. I think it's personal preference, but I tend to prefer to wait them out. My dogs have nice default behaviours. Usually if they want to go somewhere and they aren't getting there they'll offer a sit or down within about 20 seconds. It's much easier to get their attention, then, and I'd heel them back some with lots of rewards for paying attention and heeling. For us this is mostly for surprises, like the cat that jumps out of a bush and dashes across the road or something like that. Setups I try to keep really easy. If they are too distracted I'd wait for the default behaviour and get them farther away again and reassess whether this was in their capabilities or not. I figure if I ask them to do something in a situation they are not practised at doing it in, I'm essentially asking for something almost brand new.

I also like the rapid fire treats. At agility training if we're not working our dogs they are to lie down at our feet. It wasn't easy for my Erik to begin with as he just wanted to get up and do stuff, but a high reward rate convinced him to stay put and over time I lowered the reward rate. We bumble along.

[huski]

Yes, I did get that Huski and asked why he wouldn't use a correction in the circumstances which boiled down to the same "purely positive" perspective which IMHO is nonsence for the most part, and I share the same opinion regarding the side effects of correction in the same way described by Nekhbet in this thread and K9pro's link I provided.

Dogs that have learned no consequence for their actions shows up in other behavioural traits. A loose leash walk can established with positive reinforcement and the dog can walk beautifully until it reaches a major distraction and will shoot the end of the leash and pull. A dog trained to loose leash walk from aversive methods for example Koehler leash work, knows what happens at the end of the leash and they don't display the same behaviour with any distractions, there is a marked difference between the two methods in reliability terms. If you have ever walked a dog trained in Koehler leash work and walked a dog trained in positive methods, you would know the difference I am referring to when distractions are elevated, it's like chalk and cheese

Far be it for me to speak for K9 Pro but I've never seen him use a Koehler style leash correction, and I've seen him train quite a few dogs to walk reliably on a loose leash.

I also do all my training on flat collar and have since day one... and am yet to be advised to use a Koehler style correction, but I still teach my dog there are consequences to certain actions without giving a big leash correction, a leash correction is not the only way to show the dog there are negative consequences to specific actions.

[SpecTraining]

Yes, I did get that Huski and asked why he wouldn't use a correction in the circumstances which boiled down to the same "purely positive" perspective which IMHO is nonsence for the most part, and I share the same opinion regarding the side effects of correction in the same way described by Nekhbet in this thread and K9pro's link I provided.

Dogs that have learned no consequence for their actions shows up in other behavioural traits. A loose leash walk can established with positive reinforcement and the dog can walk beautifully until it reaches a major distraction and will shoot the end of the leash and pull. A dog trained to loose leash walk from aversive methods for example Koehler leash work, knows what happens at the end of the leash and they don't display the same behaviour with any distractions, there is a marked difference between the two methods in reliability terms. If you have ever walked a dog trained in Koehler leash work and walked a dog trained in positive methods, you would know the difference I am referring to when distractions are elevated, it's like chalk and cheese

Far be it for me to speak for K9 Pro but I've never seen him use a Koehler style leash correction, and I've seen him train quite a few dogs to walk reliably on a loose leash.

I also do all my training on flat collar and have since day one... and am yet to be advised to use a Koehler style correction, but I still teach my dog there are consequences to certain actions without giving a big leash correction, a leash correction is not the only way to show the dog there are negative consequences to specific actions.

Perhaps I didn't write clearly enough Huski my reference to K9Pro was his approach to training in both negative and positive methods, nothing to do with Koehler in that vein.

If you have never trained in Koehler leash obedience Huski, how do you compare the results with your methods to determine which is most effective???.

[SpecTraining]

I'd love to know, Diva. I keep looking for them, but dammit, I just can't find any.

To the OP, just be aware that increasing distance between your dog and a distraction they want to get closer to may increase their frustration, which I always find to be quite detrimental and avoid almost as much as I avoid punishments. A frustrated dog can get more aroused, and that just makes it harder to get their attention. I think it's personal preference, but I tend to prefer to wait them out. My dogs have nice default behaviours. Usually if they want to go somewhere and they aren't getting there they'll offer a sit or down within about 20 seconds. It's much easier to get their attention, then, and I'd heel them back some with lots of rewards for paying attention and heeling. For us this is mostly for surprises, like the cat that jumps out of a bush and dashes across the road or something like that. Setups I try to keep really easy. If they are too distracted I'd wait for the default behaviour and get them farther away again and reassess whether this was in their capabilities or not. I figure if I ask them to do something in a situation they are not practised at doing it in, I'm essentially asking for something almost brand new.

I also like the rapid fire treats. At agility training if we're not working our dogs they are to lie down at our feet. It wasn't easy for my Erik to begin with as he just wanted to get up and do stuff, but a high reward rate convinced him to stay put and over time I lowered the reward rate. We bumble along.

Purley out of interest, what is your reasoning for avoiding punishments Corvus???. Is it the humane, handler fallout, cruelty, purely positive patter???.

What you have achieved sounds good and semi reliable, but geez, there is a lot of unnecessary work IMHO to achieve the goals you have reached. The difference when leash trained with corrections properly, the default behaviour is that the dog stays by your side and responds to any known verbal command. The dog doesn't have to be proofed to individual scenarios like cats dashing across the road as it has two choices, stay put and enjoy tranquility or chase the cat and reach the end of the leash with an uncomfortable outcome

Edited October 4, 2010 by SpecTraining

[bedazzledx2]

Nor do the top level Obedience competitors (speaking for WA)

That's the reason you will find most if not all high level performance trainers use a combination of both positive and negative reinforcement. Positive only methods have been tried and failed training police dogs, security dogs and Schutzund sporting dogs and one's that claim to have trained in positive only methods, no one would ever know the truth or tell the difference in the dogs performance once trained. If positive only methods worked best to achieve the greatest reliablity and performance, the people dedicated to train and win in these disciplines at high level competition wouldn't be training with prongs and E Collars as they do and would be concentrating on positive only methods which they don't is the way I see things.

Interestingly the top level agility trainers don't use physical corrections

Edited October 4, 2010 by bedazzledx2

[Mrs Rusty Bucket]

My dog currently prefers to chase the cat and somersault at the end of the lead like she's so excited she doesn't even feel the (automatic) correction.

Much better to spot the cat (or horse or possum) first and teach her an alternate behaviour that is rewarding like sitting and eating (uber) treats. So she associates sitting and watching cats with pleasure. Instead of the thrill of chasing them.

Again a conditioned recall using positive methods seems to work much better for me than one involving yanking on a long line and dragging the dog back to me. I looked up Koehler methods and apart from yanking on the lead to get a dog back to heel - I couldn't figure out what it is. And yes, I've trained using that method, but luring with treats worked better for me.

[m-j]

The dog doesn't have to be proofed to individual scenarios like cats dashing across the road as it has two choices, stay put and enjoy tranquility or chase the cat and reach the end of the leash with an uncomfortable outcome

I wouldn't describe being quiet and obedient while having to suppress an innate response through worry of retribution as tranquility, by any stretch of the imagination. Not from what I have observed/experienced anyway.

Nor do the top level Obedience competitors (speaking for WA)

And they are top level too, from what I have seen.

cheers

M-J

[SpecTraining]

The dog doesn't have to be proofed to individual scenarios like cats dashing across the road as it has two choices, stay put and enjoy tranquility or chase the cat and reach the end of the leash with an uncomfortable outcome

I wouldn't describe being quiet and obedient while having to suppress an innate response through worry of retribution as tranquility, by any stretch of the imagination. Not from what I have observed/experienced anyway.

Nor do the top level Obedience competitors (speaking for WA)

And they are top level too, from what I have seen.

cheers

M-J

The dog doesn't supress anything in that situation, it simply makes a choice and chooses the most comfortable outcome which is to stay put and is more than happy to comply.

[SpecTraining]

My dog currently prefers to chase the cat and somersault at the end of the lead like she's so excited she doesn't even feel the (automatic) correction.

Much better to spot the cat (or horse or possum) first and teach her an alternate behaviour that is rewarding like sitting and eating (uber) treats. So she associates sitting and watching cats with pleasure. Instead of the thrill of chasing them.

Again a conditioned recall using positive methods seems to work much better for me than one involving yanking on a long line and dragging the dog back to me. I looked up Koehler methods and apart from yanking on the lead to get a dog back to heel - I couldn't figure out what it is. And yes, I've trained using that method, but luring with treats worked better for me.

I don't think you did figure it out either Mrs Rusty Bucket because if the dog's doing backflips at the end of the leash, you are not applying the correction correctly As the dog bolts forward, you turn around and walk in the opposite direction which teaches the dog to focus on the handler at all times, then the focus is reinforced with positive reward

Edited October 4, 2010 by SpecTraining