Teaching A Retrieve

[The Spotted Devil]

Awe he is so cute! My original dog trainer had an ESS, he was the smartest dog and picked up new things so easily. Even at an old age and nearly blind he still fronted up with her to training and did demos. Beautiful dogs.

The field-bred ones can be pretty intense - don't let that cute little face deceive you! Although the boys seem to me to be more laid back than the girls. I remember Em being incredibly intense as a puppy (still with a great off switch) whereas Mr B is all "Aw, can I have a cuddle first?"

[kiesha09]

OMG I want that puppy Not only is he incredibly cute(!!) he is a natural retriever solves all my problems :laugh:

TSD because I have been training this on and off for so long now (probably overall about 2 years with LONG breaks in between)inadvertently I've used three different items (tug toy, dumbell & cardboard bin liner roll) in probably about 3 different locations (outside, in loungeroom and in other loungeroom).

I've been doing some youtube research and kikopup has this

which might be a help. The trainer shapes the dog to pick up the dumbell and nose touch their hand. The video explains it better but it could help with her coming forwards not flinging backwards.

[The Spotted Devil]

OMG I want that puppy Not only is he incredibly cute(!!) he is a natural retriever solves all my problems :laugh:

TSD because I have been training this on and off for so long now (probably overall about 2 years with LONG breaks in between)inadvertently I've used three different items (tug toy, dumbell & cardboard bin liner roll) in probably about 3 different locations (outside, in loungeroom and in other loungeroom).

I've been doing some youtube research and kikopup has this

which might be a help. The trainer shapes the dog to pick up the dumbell and nose touch their hand. The video explains it better but it could help with her coming forwards not flinging backwards.

I like the idea of the nose touch. In terms of articles - you can use lots of things at once including pens and household items but I wouldn't use the actual item you want to retrieve (i.e. dumbbell) until you've got the behaviour you want - or at least getting it really strong and valuable. The other thing that has worked for me is incorporating the 2 food game into training. I did it last night with Em when I was nutting out her weaves - just made the session FUN - even though I felt really relaxed and my body language and voice reflected that I could see she was unsure and a good sprint every now and again really livened her up and gave her confidence. I had to fast track Zig's retrieve the last few weeks (neglected for agility) and he loved running for the food also although I used the Ready Treat. The change in attitude really came down to reward delivery.

Edited November 15, 2012 by The Spotted Devil

[The Spotted Devil]

Sorry - more random thoughts!

Does your dog know a really tight "front" position? And are you teaching it standing up or sitting down?

[itsadogslife]

Posted by Wuffles My dog is NOT a natural retriever at all :laugh: It took me 6 months to train her formal retrieve and TSD and others were telling me eventually she would learn to love it, and I didn't believe them... but she did :laugh: It is something she loves because it's been so highly rewarded, and once her reward rate for it drops she starts to lose interest again. I have to make sure I don't overdo the training (2-3 retrieves in one session is her limit) and still play lots of 'grab the dumbbell' games to build its value.

I am sorry in advance if I give offense, I'm really not meaning to, but I find the above to be rather confusing. A dog that loves retrieving is not the sort of dog that needs constant reward. If your dog is losing interest in retrieving once the reward rate starts to drop then I can't see how it is the retrieving that he loves.

My own dog, a Kelpie to whom I have only informally trained the retrieve, would literally run himself into the ground before he stopped retrieving the ball, or any other object I threw. I have never rewarded this behavior at all - other than to throw the ball.

This is not meant as a brag on my training skills or method - I actually did very little in the way of teaching the retrieve - it is simply a observation on the difference between intrinsic and extrinsic reward. For my own dog retrieving is intrinsically rewarding hence there is no need to reward the behavior other than throwing the ball. A dog that needs a constant rate of extrinsic reward in order maintain its behavior is not a dog that I would say "loves" what he is doing.

[The Spotted Devil]

Posted by Wuffles My dog is NOT a natural retriever at all :laugh: It took me 6 months to train her formal retrieve and TSD and others were telling me eventually she would learn to love it, and I didn't believe them... but she did :laugh: It is something she loves because it's been so highly rewarded, and once her reward rate for it drops she starts to lose interest again. I have to make sure I don't overdo the training (2-3 retrieves in one session is her limit) and still play lots of 'grab the dumbbell' games to build its value.

I am sorry in advance if I give offense, I'm really not meaning to, but I find the above to be rather confusing. A dog that loves retrieving is not the sort of dog that needs constant reward. If your dog is losing interest in retrieving once the reward rate starts to drop then I can't see how it is the retrieving that he loves.

My own dog, a Kelpie to whom I have only informally trained the retrieve, would literally run himself into the ground before he stopped retrieving the ball, or any other object I threw. I have never rewarded this behavior at all - other than to throw the ball.

This is not meant as a brag on my training skills or method - I actually did very little in the way of teaching the retrieve - it is simply a observation on the difference between intrinsic and extrinsic reward. For my own dog retrieving is intrinsically rewarding hence there is no need to reward the behavior other than throwing the ball. A dog that needs a constant rate of extrinsic reward in order maintain its behavior is not a dog that I would say "loves" what he is doing.

I see what you are saying - I think it is about which PART of the retrieve the dog finds rewarding. For us lucky ones, it's the going out to the retrieve that is intrinsically rewarding. They bring it back to us because they are rewarded by us throwing it again etc etc. With dogs like my Dally and Wuffles' Ava, the backchaining and shaping means that bringing it back is incredibly rewarding. Thus they go and get it! You can get to the point where they transfer so much value to the retrieve (or tug or agility etc) that it does become self-perpetuating and self-rewarding over time. But if you stop rewarding too early or don't mix up the rewards it can lead to extinction like any other behaviour.

I do like to shape a mad retriever to retrieve simply because it's another tool I can use when I'm teaching more advanced retrieves. My Springer doesn't get sick of retrieving but there's a lot of other behaviours I want surrounding that so I do use other rewards.

Sorry I also meant to say that when you've had a dog that finds the idea of tugging or retrieving (or doing ANYTHING for that matter!) pretty aversive you can get a little carried away with emotive descriptions - I know I do anyway :laugh:

Edited November 15, 2012 by The Spotted Devil

[itsadogslife]

But if you stop rewarding too early or don't mix up the rewards it can lead to extinction like any other behaviour.

Fair enough, I agree. But I am also inclined to think that the point of teaching anything is to bring the dog to the point of understanding whereby the behavior becomes intrinsically reinforcing. The constant emphasis on extrinsically rewarding the dog seems to me at least, to be getting in the way of the dog's understanding. I mean if you saw someone who was constantly correcting their dog, you would rightly say to them that their dog does not understand. The need to constantly be reinforcing the dog is a least to my eyes an indication the the dog does not understand the value of what is being taught. Rather the dog is simply working for the value of the extrinsic reward. And yet whenever the behavior in question starts to drop off the advice always seems to be increase the extrinsic reward.

[Kavik]

But if you stop rewarding too early or don't mix up the rewards it can lead to extinction like any other behaviour.

Fair enough, I agree. But I am also inclined to think that the point of teaching anything is to bring the dog to the point of understanding whereby the behavior becomes intrinsically reinforcing. The constant emphasis on extrinsically rewarding the dog seems to me at least, to be getting in the way of the dog's understanding. I mean if you saw someone who was constantly correcting their dog, you would rightly say to them that their dog does not understand. The need to constantly be reinforcing the dog is a least to my eyes an indication the the dog does not understand the value of what is being taught. Rather the dog is simply working for the value of the extrinsic reward. And yet whenever the behavior in question starts to drop off the advice always seems to be increase the extrinsic reward.

I think it depends on what your aim is? If your aim is competition then you are trying to get a happy and fast performance to a certain standard. Also you may try to tweak small things in their performance, or challenge them or proof them, and to do this extrinsic rewards are very useful as it allows you to increase your criteria to get a better performance or non reward a less than average or poor/incorrect performance. If the action is reinforcing itself (and certainly not all competition behaviours would be able to be made self reinforcing) then it is more difficult to change small aspects of it.

Edited November 15, 2012 by Kavik

[kiesha09]

TSD I'm happy to hear any random thoughts so go for your life

She currently doesn't know a tight front and generally I am sitting down with her in front of me either standing, sitting. I haven't set a criteria for this because at the moment I just want her to pick up the item and put it in my hand and I hadn't given a thought to whether she should be sitting or standing.

[huski]

I wouldn't train a dog to see the dumbell as the reward. There are too many problems IMO that can arise from the dog seeing the exercise as the reward itself.

I taught my beagle to retrieve a dumbell and she has literally no retrieving instinct. I shaped her picking it up and went from there. She has a very clean and enthusiastic retrieve now :) ETA: She always completes the retrieve in food drive.

Edited November 15, 2012 by huski

[The Spotted Devil]

But if you stop rewarding too early or don't mix up the rewards it can lead to extinction like any other behaviour.

Fair enough, I agree. But I am also inclined to think that the point of teaching anything is to bring the dog to the point of understanding whereby the behavior becomes intrinsically reinforcing. The constant emphasis on extrinsically rewarding the dog seems to me at least, to be getting in the way of the dog's understanding. I mean if you saw someone who was constantly correcting their dog, you would rightly say to them that their dog does not understand. The need to constantly be reinforcing the dog is a least to my eyes an indication the the dog does not understand the value of what is being taught. Rather the dog is simply working for the value of the extrinsic reward. And yet whenever the behavior in question starts to drop off the advice always seems to be increase the extrinsic reward.

There is a difference between understanding and motivation. And building value into an exercise depends on the level of precision you want too. Me wanting a perfectly straight sit, dog looking me in the eye, not mouthing the retrieve item and not dropping the article even if I place my hands on it is a lot different to a dog dropping a ball at your feet. As I commented recently in the case of the latter who's shaping who? :laugh: The thing is, whether we are training or not the dog is ALWAYS learning.

Regarding using rewards - I think Bob Bailey put it beautifully when he said "dog training is simple but not easy". You see I talked about the delivery of the reward and there are so many more subtleties I find difficult to pin down. Like the dogs who pick up your frustration and shut down. Or the times we inadvertently reward the undesired behaviour. Not everyone has good timing (let alone great!) The trainers who use positive reinforcement that I admire the most have very stringent criteria and an innate understanding about when to up the ante. Real hard @r$es in fact!

[The Spotted Devil]

Damn - missed all the replies - sorry, on my phone!

Keisha - ok, so maybe you could take the retrieve out of the picture and try the neat front. It might give you both something to work with and that is the end behaviour you want, which is the strength of back chaining. It's so hard to explain over the interwebz!

I don't mind what the dogs find rewarding as long as it is of my choosing. You can't help but transfer value across different rewards - chaining behaviours does that.

Edited November 15, 2012 by The Spotted Devil

[huski]

There is a difference between understanding and motivation.

Totally agree! Just because you use rewards like a tug or food to motivate a dog does not mean the dog doesn't understand the exercise it's being asked to perform. To be honest I find that line of thought a bit strange.

Edited November 15, 2012 by huski

[itsadogslife]

There are too many problems IMO that can arise from the dog seeing the exercise as the reward itself.

You must be joking? I am really not sure I understand this. I presume Kelpies and Borders Collies love herding - where are the problems with this?

[huski]

There are too many problems IMO that can arise from the dog seeing the exercise as the reward itself.

You must be joking? I am really not sure I understand this. I presume Kelpies and Borders Collies love herding - where are the problems with this?

I was talking in regards to the dumbell retrieve, but I also think it can be applied to a lot of other things too. Why would I want my dog to learn that the exercise is the reward, as opposed to the exercise is what you need to do to earn the reward from me?

Edited November 15, 2012 by huski

[itsadogslife]

Totally agree! Just because you use rewards like a tug or food to motivate a dog does not mean the dog doesn't understand the exercise it's being asked to perform. To be honest I find that line of thought a bit strange.

You must have missed the part where I said the need to constantly reinforce the behavior or else it drops off leads me to the belief that the dog doesn't understand.

[itsadogslife]

Why would I want my dog to learn that the exercise is the reward, as opposed to the exercise is what you need to do to earn the reward from me?

As the exercise requires my participation, the reward is self evidently my participation.

[Weasels]

There are too many problems IMO that can arise from the dog seeing the exercise as the reward itself.

You must be joking? I am really not sure I understand this. I presume Kelpies and Borders Collies love herding - where are the problems with this?

Ever tried taking one to a trial? Where there's stock and other dogs working all day and they only get one or two 10 minute runs :/ Learning a bit of self-control and the value of other rewards in the presence of stock is merciful to everyone within earshot.

[The Spotted Devil]

Sometimes that's also a trap of not realising what you're rewarding. I've noticed my dogs have Eureka! moments when teaching something new - they'll anticipate or throw in another behaviour - almost trying too hard. When I withhold the reward it's like "duh! Now I get it" and then I can work on improving performance, increasing distractions etc.

Also, dogs tend to be crap at generalizing. Big trap!!!

Another also! Every dog sport is stylised nowadays, even field trials and retrieving. I mean, honestly, what hunter would give a crap which order the game was picked up in or whether the dog cheats the water. Admittedly being steady to the shot is pretty desirable!!!

ETA: sorry again - in response to itsadogslife

Edited November 15, 2012 by The Spotted Devil

[itsadogslife]

Ever tried taking one to a trial? Where there's stock and other dogs working all day and they only get one or two 10 minute runs :/ Learning a bit of self-control and the value of other rewards in the presence of stock is merciful to everyone within earshot.

Sorry, I don't really see the relevance. Are you saying sheepdog herders would do better if herding wasn't so intrinsically rewarding to the dogs?