Training A Reliable Recall

[jesomil]

I have been reading the thread about training a reliable recall at the dog park and want some opinions.

I have had a few discussions with people about methods used and i have come to my own conclusion that you cant get a reliable recall using a reward method. I have found that the only dogs who recall very well most of the time have been trained with aversives.

Maybe i have only come across people using the reward method ineffectively?

There are 3 "professional" trainers that i can think of that have dogs that still require a food pouch and treating where ever they go and their dogs recall is good under heavy distraction only if they throw in the recall the split second before they are off.

I am talking about recalling off very heavy distraction, not just a basic recall.

I would be interested in others opinions on this. I am very happy to be corrected if i am wrong. Basically i have never seen or heard of a dog who recalls most of the time off anything that has been only positively trained. Would be interested to see if it is possible.

Thanks.

By the way, i am not asking for advice on how to teach the recall myself. My method works very well and i am happy with it. Just more questioning others methods and their effectivness.

[Working_Setters]

Interesting thread. As I guess is widely known after recent threads I train with both positive and negative.

I've been to 3 different obedience schools/classes (1 in Oz, 2 in USA) and in every case my dogs have recalled better than the instructors. I had one class in the USA where the instructor couldn't let her dog off leash out of her yard as no recall

I agree, I've never seen a reliable recall from a positive only trained dog. Now it could be me, or the positive only people I've met to date, but IMO clicker etc is great for pups and great for teaching/shaping all sorts of interesting behaviour, stuff that would be really hard to teach in other ways. However the down fall of these methods IMO is distractions, when the dogs would rather continue what they are doing than get the reward, what do you do? No enforcement means, (in my limited understanding of positive only training) that if a command is ignored and you can't find a more desirable reward, then you are SOL.

Would be very interested in hearing from positive only trainers that have achieved reliable high distraction recalls. How did you do it? Do you think there are advantages over more “traditional” (use of some negatives) methods?

[dogdude]

From what I have read about other peoples experiences with working with prey drive, they say that that method gets reliable results. Basically, they work on getting a prey item as the dogs greatest distraction and building the dogs drive so it will do almost anything to satisfy its frenzy.

They build it to a point where it is valued by the dog more than anything else.

I dont know alot about it but K9 is the expert. I get the impression though that not all breeds or individuals are responsive to this method. Maybe he can clear things up? I am very interested in finding out about the level of success across individual dogs.

Let me just say, I use a combo of both positive and negative. Positive to teach and aversives to proof, and where necessary, back up the negative with a positive.

I personally believe from my own experiences to date, that you have to use aversives to proof a dog properly under varying distraction to get good reliability. Mind you, I dont use any forcefull form while proofing initially, first show them gently, and only when the dog has made a clear choice to ignore the command I will add a pop on a check etc. When you test a dog by proofing I think it is very important to make sure that the dog is not confused before using force. You are in effect trying to trick him into making a mistake and force would be unfair.

Edited December 12, 2006 by dogdude

[squeak]

I have had a few discussions with people about methods used and i have come to my own conclusion that you cant get a reliable recall using a reward method. I have found that the only dogs who recall very well most of the time have been trained with aversives.

I'm curious to understand what you mean by "aversive". When I read it the first time, I thought that you meant something like beating the dog with a stick if it didn't obey the command. Now that I've reread it, do you mean an aversive to be something like a "correction" with a check chain?

[wheres my rock]

come down to kcc and see my dogs i have three with excellent recalls all trained with positives and no i dont have to have food on me for them to respond.

asl shoemonster last night i had our bc and our new addition a german spitz whos only been with us two weeks running off lead and playing with several dogs they came everytiem they were called i can think of three times where my shelti ignored my frst call to come and they were all in those teenage years and he did come but took a few seconds to think about it

[shoemonster]

Yeah, sad but true, Stella's dogs make mine look like delinquents!

[wheres my rock]

oh they do not

yours are coming along in leaps and bounds

you one of the most natural trainers ive met in a long time

[helen]

Wheres my rock obviously has learnt how to apply the theory of positive reinforcement in a manner that allows a reliable recall to happen, like so many people can and have done (myself included).

Have you read the book Don't Shoot The Dog by Karen Pryor? Might answer some of your questions

[wheres my rock]

actually the dvd really reliable recalls is very good too i wntched it the other day

thanksyou helen years the many years of studying and experimenting have paid off a bit lol

[Working_Setters]

Wheres my rock obviously has learnt how to apply the theory of positive reinforcement in a manner that allows a reliable recall to happen, like so many people can and have done (myself included).

Hi Helen - If your dog was out free running, say 50 meters from you, and put up a rabbit, could you reliably stop/recall the dog mid-chase? What reward can you offer a Brittany (or other gundog) that is more desirable that chasing a rabbit?

[wheres my rock]

my whippet would stop on a chase and come reliably everytime

a because we didnt let him off to chase until the recall had become a reflex action

and b he it often meant he would get to go again on something else

i think you did the same called dog of rabbit then cast them back out in a different direction

my sheltie will stop on sheep and then come on recall

to add my gsp that did do retrieveing also had an ecellent recall and when we got her she was a livestock chaser but after trianing she would ignore the world around her could swim through duck for her dummie and always came when i called her

Edited December 12, 2006 by wheres my rock

[wheres my rock]

it is my belief that anything you can train wit aversives can be taught with positives you just have to know how and belive in th methods

[Working_Setters]

deleted - cause it's more trouble than it's worth.

Edited December 14, 2006 by Working_Setters

[wheres my rock]

hehe i remember when started retrieving trianing a guy who was trinaing with us had a dog with a crap recall he marched out grabbed th dog dragged it back to the spot where he called it from left it in a stay and walked back to where he started called it wt happy body language and the dog ran off into the bushes took half an hour to find it and get it back dogs know when someone is shitty and he was shitty he couldnt fool the dog

[Working_Setters]

my whippet would stop on a chase and come reliably everytime

a because we didnt let him off to chase until the recall had become a reflex action

and b he it often meant he would get to go again on something else

i think you did the same called dog of rabbit then cast them back out in a different direction

YES - I agree that is a huge part of it, stop/recall can't always mean end of fun, dog must know that there is an excellent chance that recall will be a brief pat and then back to more fun, but not recalling means no more fun

[Working_Setters]

my whippet would stop on a chase and come reliably everytime

a because we didnt let him off to chase until the recall had become a reflex action

So the key to a positive trained reliable recall is to practice over and over until it's a reflex action? Then it's not a question of the reward out weighing the pleasure of not responding (continuing to chase the rabbit for example), since the dog is so well conditioned to the command that there won't be any thought of not obeying?

If the whipet did ignore a recall, how would you correct it? Are there no corrections? Just ignore and go back to a lower level distraction environment and more practice until response is better ingrained?

Edited December 12, 2006 by Working_Setters

[wheres my rock]

umm if he ignored he be so far away you wouldnt be able to go after him hehe he was that bloody fast

id nick off and hide and then he'd have to find me or id run away in the other direction chasing something else when he found me we'd have a big play for coming to get me then i'd pop the long line on and go back to the drawing borad whilst hitting myself in the head for overestimating his level of training

but thast my point never happened because i train and trian under huge distraction in a controlled environment before the dog is allowed off willy nilly.

for eample rabbit skin on a lure heeling next to the race track where he knows he gets to chse then letting him have a run remote control cars ae a lot of fun but you have to have a good driver they hurt when they whack you in the ankle. we spend most of the first year of trianing working on finding different distraction to train with then the second year polishing for the ring.

we train very strong focus so the dogs become less aware of the environment around them it becomes unimportant to them as the best thigns in life come from me and if i see a bunny and sedn them off to chase it i als get to say when thast enough but its ok because you may leave the bunny but something good is always around the next corner maybe another bunny

[Arya]

Good thread! Good to read other people's ideas and see what other people do. I tried to use positive only for months. I must say it didn't work for me with the dog I have now. Often, as soon as I turned my back to walk away the dog was up and off. Or she'd run right past me. If she had a middle finger she would have been flipping me the bird on the way past LOL! (I can laugh about it now...) In the end, against everything I'd been taught, I set her up to fail on a long line. I took her to an area of high distraction. i was very careful as the longer the line the greater the force on the dog's neck. I wasn't out to be cruel. Then we practiced... For every instant, good 'come' she did she got heaps of praise and rewards. For every glance away before she came (this was a regular thing - check out if there's something more interesting to run to before making the decision to come to handler), or attempt to run, she got a good hard, quick sharp correction (sorry to those who don't like this!). Very black and white. All xmases came at once or all hell broke loose. Her choice. Guess what she chose. After two short five minute sessions, months of struggle were over. I simply showed her the consequences of her actions, good and bad, very plain. About a month later we had another single run away. Back on the long line for two minutes. It hasn't happened again. Now I have a dog who loves the recall. She rockets in like a bullet, ears up, mouth open. Knows a big reward is coming. Purely positive now because she doesn't do the wrong thing. Under heavy distraction she is just as reliable, with or without a food reward. Good and bad, consequences of actions. I think each individual dog is different though and a correction like this would have sent my last dog into a nervous breakdown.

[Working_Setters]

Calling a whippet off a rabbit is quite a feat, sighthounds are not noted for their responsiveness, particularly during a chase - I'm impressed :D

During the first year, when you are working on building that focus, what type of exercises are you doing, are they food, praise, play based, or something else? Do you have a core of general exercises you do with all your dogs, or is it very much dog specific?

In a way our training isn’t all that different, neither of us EVER lets the dog get away with not obeying a recall, it’s just that you have the patience to work on a long line for a year, where as I cut them loose earlier, but expect to have to run them down (e-collar correct) a couple of times, to enforce recall.

[Sardog]

Our National SardaUK (parent organisation)standard requires the search & rescue dog to be "stock proofed". In the UK, Scotland and Wales the dogs are required to search the "moors" with stock grazing most of the time.

We stock proof our pups with the negative value attached and we are lucky enough to be able to train in a very "rabbity" area.

One example of this was put into practice was a few months ago we had our sponsors - Bayside Companion Dog Training School, 8 of their instructors come to spend the day with our team training at a disused school in GlenWaverley. We demonstrated the disaster discipline of our work which required the dogs to locate victims placed inside and outside buildings. Both operational dogs had very difficult searches and we, as handlers/trainers were very pleased with their expertise and accuracy and was hoping the onlookers had the same opinion. Sardog "Gus" was searching the outside of the buildings and was searching a small shed located in a remote part of the school grounds. Suddenly a fox shot out of the shed with Gus "right on his tail"...it was the element of surprise that made the dog chase and his handler blew his whistle and the dog immediately spun around and back to the handler. The dog then continued on his search.

At the debrief, the instructors all made complimentary comment...not about the accuracy of the dog's search work, but "being able to call the dog off the fox!"....oh well, it was great anyway!