What's So Bad About Using Force In Dog Training?

[Kavik]

Yes my experience with NDTF is probably not that relevant now as it was 16 years ago and pretty much all lecturers were security guards or ex military. We were shown all training equipment and shown how it was used, when and why. They did not shy away from corrections. I met some people and saw some training that I would not have otherwise.

Edited June 11, 2016 by Kavik

[The Spotted Devil]

Thanks Kavik and SC. Honestly I was not going to post in this thread but if just one person thinks "there must be another way" like I did 20 years ago then that is worth the effort. I'm just trying to pay it forward as a way of thanking those that have inspired me.

[Willem]

I'm not sure what your last post is supposed to mean Kavik but as a current lecturer on the NDTF course who has been heavily involved in content changes over the last few years, I'd just like to advise anyone here that the course has seriously evolved and changed in recent times.

I'm a bit tired of debating training techniques so won't weigh in too heavily- Willem, you might be interested in my recent blog post.

http://underdogtraining.com.au/dont-be-a-tool-the-dog-training-equipment-debate-debacle/

I agree with Bob Bailey's position on punishment and novice trainers generally.

thanks for the article...also I would have liked to read more about the 'genuine skills' - my objective is to establish a communication with my dog that allows me to guide her without any force. That's the big objective for me in dog training. And, for me, the most important skill to get there is gaining the trust of my dog. Once I got her trust, the learning curve (in agility and obedience) increased dramatically. Positive reinforcement, force...that are only tools for the basic training in my toolbox, without the trust any of these tools are difficult to use. And - this might sound confusing - IMO these tools don't help regarding gaining the trust, they are really only useful for the basic training in agility and obedience. For gaining the trust some other skills are required - can't find the right wording, has something to do with the attitude towards the dog...

[Mrs Rusty Bucket]

willem

Your method for loose lead dog walking - complete fail for my dog. I had used it "successfully" with other dogs but not with this one. She'd just shut down completely and I'd have to carry her home or drag her along the ground. That's why I had to learn about reward based training.

The trouble with force or aversives - is fall out. The dog doesn't always connect the force with the behaviour you're trying to "stop". It's much much easier to train a behaviour you do want with something the dog likes as reward. And the studies are showing that this is also faster and more reliable and lasts longer than training by force.

If you stuff up your reward based training - there's much less trauma associated with fixing your training.

The agility dogs that win - need to be fast. They can't be fast if they're worried about what will happen to them if they stuff up (and they don't even know what it would take to stuff up). You can see the ones that are scared of what will happen if they knock a bar down or miss a contact - they're slow and very careful.

He (Bob Bailey) is also a huge fan of Susan Garrett.

I think that fan club goes both ways.

And after last year's SG August Seminar - I think Steve White has become a huge fan. He's been looking for ways to train without using force since he started and got mauled by his first assigned service dog for trying to force it to drop. How much faster, gentler and safer that task would have been if he'd used marker and rewards.

Edited June 11, 2016 by Mrs Rusty Bucket

[Mrs Rusty Bucket]

there is no difference, not even a subtile one - or you would prove Newton (actio = reactio) and Einstein (relativity) wrong, which would be quite interesting.

Fact is that the dog can't apply a force on its own if you don't apply the force on the other end of the leash!

You want to move, but the dog doesn't results in tension in the leash, the dog wants to move, but you don't results in tension in the leash, you both want to move, but in different directions or with different speed results in tension in the leash. The difference between you and the fence post is only that you can move, the post can't.

For the training there is no difference regarding force respectively who applies the force: you and your dog walking in different directions will result in tension in the leash. Saying that it is the dog that applies the force (and that the exercise is therefore 'force free') because he doesn't follow you is ignoring your existence and intention - but you are the reason why there is the tension in the leash! The purpose of the leash / lead is to 'force' the dog to follow the handler respectively to respond to the handler's movements, not the other way round - so who applies the force?

If you want to do a force free training: use a string instead of a leash...(and no collar grab games).

Eta: the problem of the 'force-free trainer army' is that they mistaken the objective of the training for a suitable method to get there.

Might as well deal with this in here.

in order for the dog to be in motion - the dog applies force to the ground and its own body...

in any case - physics aside - the motion of the dog is initiated by the dog. I'm not moving. I am as the fence post. I don't use the lead to yank the dog in my direction and I don't do that fast change direction and yank that some recommend. I just wait for the dog to be paying attention to me and then encourage her to come with me - and reward that. I do tell her when I'm changing direction and she responds well to that.

Most of our loose lead walking is done without a lead where it is safe to do so. So our on lead training is mostly to stop her nicking off to eat stuff out of the hedges or stop her from scaring the crap out of poodle crosses and their owners at the park - ie my recall there is not perfect so she stays on lead. When we're on the footpaths mostly it's entirely loose lead walking - unless we spy a cat. And then it can get exciting. But there's no way I initiate the force in that. She's pretty good at not launching at cats but sort of indicating instead. And then I can reward her staying calm enough to pay attention and there is no launch - hooray but an ongoing training challenge for us.

If I beat her or yanked her around for trying to chase a cat - I think she would blame the cat and become more aggressive when she saw one.

[Willem]

willem

Your method for loose lead dog walking - complete fail for my dog. I had used it "successfully" with other dogs but not with this one. She'd just shut down completely and I'd have to carry her home or drag her along the ground. That's why I had to learn about reward based training.

The trouble with force or aversives - is fall out. The dog doesn't always connect the force with the behaviour you're trying to "stop". It's much much easier to train a behaviour you do want with something the dog likes as reward. And the studies are showing that this is also faster and more reliable and lasts longer than training by force.

If you stuff up your reward based training - there's much less trauma associated with fixing your training.

The agility dogs that win - need to be fast. They can't be fast if they're worried about what will happen to them if they stuff up (and they don't even know what it would take to stuff up). You can see the ones that are scared of what will happen if they knock a bar down or miss a contact - they're slow and very careful.

agree with everything, except the red highlighted text - there might be sometimes less trauma, but it can just be the same.

wrt agility: ...my dog is too fast (for me) and doesn't give a damn if she knocks something over :D ...on the other hand I could send her through a burning tire (because she trust me).

When you read my last post (about the trust etc.): I guess no matter what tool you pick, you need the trust of your dog when exercising this particular tool - it might be easier to get the trust for a positive reward based system with yummy treats - the proof will be in the pudding when you do the recall without the yummy treat in your hand :) . There are also other examples where the reward system just don't work as you would expect it based on your experience with other dogs: in our class there is a friendly guy with a friendly dog, nothing wrong with the guy, nothing wrong with the dog - except I just can't see any bond between the two, and the dog is not really food driven. Consequently the communication between the two is somehow handicapped when it comes to recalls and distance training.

[Simply Grand]

Really? Do you think my dog could see whether or not there was a treat in my hand when I recalled her away from a young child (her favourite thing in the world, better than any food) approximately 200m away at the park today?

There actually was no treat in my hand and her reward was a pat, which was enough as she has the conditioning and reward history that means when I call "Quinny come!" She does.

ETA she and the child were 200 metres away from me, she was right next to the child when I saw them and called her back.

Edited June 11, 2016 by Simply Grand

[The Spotted Devil]

Correcting agility dogs causes them either to take more off course obstacles and/or bark/spin/nip in frustration OR causes them to slow down. Neither is desirable.

[Simply Grand]

Actually now that I think about it, Willem how did you train recall?

[Willem]

there is no difference, not even a subtile one - or you would prove Newton (actio = reactio) and Einstein (relativity) wrong, which would be quite interesting.

Fact is that the dog can't apply a force on its own if you don't apply the force on the other end of the leash!

You want to move, but the dog doesn't results in tension in the leash, the dog wants to move, but you don't results in tension in the leash, you both want to move, but in different directions or with different speed results in tension in the leash. The difference between you and the fence post is only that you can move, the post can't.

For the training there is no difference regarding force respectively who applies the force: you and your dog walking in different directions will result in tension in the leash. Saying that it is the dog that applies the force (and that the exercise is therefore 'force free') because he doesn't follow you is ignoring your existence and intention - but you are the reason why there is the tension in the leash! The purpose of the leash / lead is to 'force' the dog to follow the handler respectively to respond to the handler's movements, not the other way round - so who applies the force?

If you want to do a force free training: use a string instead of a leash...(and no collar grab games).

Eta: the problem of the 'force-free trainer army' is that they mistaken the objective of the training for a suitable method to get there.

Might as well deal with this in here.

in order for the dog to be in motion - the dog applies force to the ground and its own body...

in any case - physics aside - the motion of the dog is initiated by the dog. I'm not moving. I am as the fence post. I don't use the lead to yank the dog in my direction and I don't do that fast change direction and yank that some recommend. I just wait for the dog to be paying attention to me and then encourage her to come with me - and reward that. I do tell her when I'm changing direction and she responds well to that.

Most of our loose lead walking is done without a lead where it is safe to do so. So our on lead training is mostly to stop her nicking off to eat stuff out of the hedges or stop her from scaring the crap out of poodle crosses and their owners at the park - ie my recall there is not perfect so she stays on lead. When we're on the footpaths mostly it's entirely loose lead walking - unless we spy a cat. And then it can get exciting. But there's no way I initiate the force in that. She's pretty good at not launching at cats but sort of indicating instead. And then I can reward her staying calm enough to pay attention and there is no launch - hooray but an ongoing training challenge for us.

If I beat her or yanked her around for trying to chase a cat - I think she would blame the cat and become more aggressive when she saw one.

..physics aside???? ... you might have taught me a lot about the right semantics wrt operant conditioning etc., but that doesn't give you the right to tell an engineer 'physics aside'!!!! :D

it doesn't matter who initiates the tug of war - you (as the pinnacle of evolution and superior representative of mammals) are responsible for the inflicted tension, whether you are playing the fence post or not. You, respectively your choice, provides the anchor required for the tension in the leash.

What I did (and still do) - and I assume it added significantly to her good recall - is using a long leash when walking her. That allows her (if I allow it) to walk further in front of me. Now I trained her (while she was just trotting and not distracted) to respond to an 'EH' plus hand signal and / or a gentle tug to go back into heel position (which got rewarded). Similar to clicker training most of this training happend without distraction on daily walks, and now it is pretty deep imprinted. Plus training the take-it/leave-it game with the flirt pole (her favorite game). When I walk her now on-leash, all I need is a gentle tug to get her attention (with hand signal or verbal cue for heel position), if off-leash an 'EH' will do the same (she can be 20 meters in front of me)...and if we bump into cats, magpies or cattle an 'EH' plus leave-it will hold her back.

[Willem]

@SG (your last 2 posts): ...of course you don't need to have a treat once the wanted behaviour / response to a cue is imprinted - that the whole purpose of positive reinforcement training. The proof is in the pudding and you proved it.

Wrt your second question / post #29: I guess I answered it with my response to MRB.

[Simply Grand]

Yes you did answer it, thank you,but how are you going to graduate from a recall from 20 metres on a long line using the "EH" to recalling from 200 metres away?

ETA also, now I think about it, I don't think I ever showed a treat in my hand with Quinn when training recall. By the time I got her I had a good understanding of the concept of conditioning a recall, which doesn't involve "show the treat and lure them in".

Edited June 11, 2016 by Simply Grand

[Willem]

Yes you did answer it, thank you,but how are you going to graduate from a recall from 20 metres on a long line using the "EH" to recalling from 200 metres away?

the long leash I use for walking and training the recall is actually only 5 meters - it seems it doesn't need to be longer as the reflex the cue triggers seems to be independent from the distance. I have a longer leash / line, but rarely used it as it always got tangled.

When I wrote '20 meters': that's the typical distance she trots in front of me when off-leash, sometimes it might be 50 meter, depends on the line of sight and how much distance I allow her. Wrt '200 meter': not sure whether my voice is loud enough to call her over such a distance (would have to use a whistle, we trained this too, but not over 200 meters) - I don't think she ever took off so far. The oval from the club trainings ground might have 150 m in length - so that would be the longest distance I could test her recall (while she was running with other dogs).

When we bump into walkers respectively if we see them in the distance she stops, turns the head to me and waits for instructions :) - I call her, she comes, heel position / sit...walkers pass (often with some positive comments about her behaviour), she gets her reward and 'go'.

[Willem]

Yes you did answer it, thank you,but how are you going to graduate from a recall from 20 metres on a long line using the "EH" to recalling from 200 metres away?

ETA also, now I think about it, I don't think I ever showed a treat in my hand with Quinn when training recall. By the time I got her I had a good understanding of the concept of conditioning a recall, which doesn't involve "show the treat and lure them in".

...not required, and even better than luring - you just reward the wanted behaviour that's triggered by the cue and not the lure. Luring involves the risk that the treat becomes the visual trigger.

[Simply Grand]

Yes the high pitch tone of a whistle or call does help cross distance, the recalls for my three ("Quinny come!" "Saxsaxsaxsaxsax" and "rileyrileyriley") are all delivered in a high pitched tone that is horribly annoying even to my own ear. Works though.

I start all recall training with a high pitch "pup pup pup pup pup" which all dogs seem to instinctively respond to and head toward from as soon as they can hear and walk. Then the next step is to always, always make coming to me a rewarding experience. So distance and ability to physically control the dog don't really come into play.

[Simply Grand]

Stuffed up my quoting. See next post.

Edited June 11, 2016 by Simply Grand

[Willem]

Yes the high pitch tone of a whistle or call does help cross distance, the recalls for my three ("Quinny come!" "Saxsaxsaxsaxsax" and "rileyrileyriley") are all delivered in a high pitched tone that is horribly annoying even to my own ear. Works though.

I start all recall training with a high pitch "pup pup pup pup pup" which all dogs seem to instinctively respond to and head toward from as soon as they can hear and walk. Then the next step is to always, always make coming to me a rewarding experience. So distance and ability to physically control the dog don't really come into play.

aaah!...negative reinforcement via removing noxious stimuli following correct behaviour :laugh: ...only kidding...

[Simply Grand]

Yes you did answer it, thank you,but how are you going to graduate from a recall from 20 metres on a long line using the "EH" to recalling from 200 metres away?

ETA also, now I think about it, I don't think I ever showed a treat in my hand with Quinn when training recall. By the time I got her I had a good understanding of the concept of conditioning a recall, which doesn't involve "show the treat and lure them in".

...not required, and even better than luring - you just reward the wanted behaviour that's triggered by the cue and not the lure. Luring involves the risk that the treat becomes the visual trigger.

Hang on, but you said in your response to Mrs RB that the test was in how a dog trained with positive reinforcement performed when the treat wasn't visible? if the treat being visible is a requirement for the dog to perform the behaviour then positive reinforcement training is not being done properly, same as if a handler has to repeatedly jerk a collar correction based training isn't being done correctly.

[Willem]

Yes you did answer it, thank you,but how are you going to graduate from a recall from 20 metres on a long line using the "EH" to recalling from 200 metres away?

ETA also, now I think about it, I don't think I ever showed a treat in my hand with Quinn when training recall. By the time I got her I had a good understanding of the concept of conditioning a recall, which doesn't involve "show the treat and lure them in".

...not required, and even better than luring - you just reward the wanted behaviour that's triggered by the cue and not the lure. Luring involves the risk that the treat becomes the visual trigger.

Hang on, but you said in your response to Mrs RB that the test was in how a dog trained with positive reinforcement performed when the treat wasn't visible? if the treat being visible is a requirement for the dog to perform the behaviour then positive reinforcement training is not being done properly, same as if a handler has to repeatedly jerk a collar correction based training isn't being done correctly.

..not sure whether I got this (brain gets tired)...positive reinforcement is adding an appetitive stimulus following correct behaviour - you don't show the stimulus to trigger the wanted behaviour - that would be luring.

Eta: ah, got it, yes, you are right (proof is still in the pudding, but is shouldn't matter whether she has the treat in her hand or not :)...)

Eta: to put this into some context: what I had in mind was the example I mentioned in the same post that treats as reward doesn't achieve always the same reliable outcome. Take a dog that is less food driven and it will be much harder to get the wanted reflex imprinted.

Edited June 11, 2016 by Willem

[corvus]

"Force free" is not a literal term, Willem. ;) That's pretty much the end of the story.

I train for positive emotional states and improved welfare. How I get there depends on what I've got to work with, but adding aversive stimuli is typically counter-productive for these goals. I am not a brilliant trainer, and so I don't always get the results that I hope for. It is a risk when making decisions for another being that you cannot easily communicate with that you let your own shortcomings dictate the options available to you and feel justified in whatever decision you make as a result. It is easy for this to become a cycle you never break out of. I used method x, because I could see no way to use y, and x worked well. I will use x again in future because y is too hard. To grow as a trainer, you have to be prepared IMO to always push the envelope and seek ways to use y, and when you do use x instead, that is not where things end. You revisit again and again as your skills grow and think "Was there a way to use y after all?" Why y? Because it's the fastest, easiest route to positive emotional states and improved welfare of course.