Positive & Traditional Training Techniques Discussion

[Allyson]

If you control access to all the resources a dog wants, you are then able to reinforce appropriate behaviour by permitting that access so you can train by

Is this aversive training? or positive?.or both?
It would appear to me ,that is what I base all training on.
I was in the military , it worked .

I think the quadrant in OC we are looking at in this scenario is negative punishment is it not!

There is no such thing as "purely positive" ie no sentient being can be trained without the use of some punishment.

Some people are not concerned with the amount or type, others are.

[Stacia]

Peter, I would be interested to read further about your regime. At the moment I am in the middle and see both sides. I think the most important part of working with a dog is that the dog has confidence in you and from that confidence, all the rest grows as you work together. I don't actually agree with shaking a dog if it moves from the sit position as that usually then causes some anxiety and the dog cannot concentrate and learn its lesson. Far better (in my opinion) is to simply put the dog on the lead and calmly lead it back to the spot from whence it moved, reward it there by whatever means, mine would be "good dog" so the dog understands that that is where is should be. Also, if the dog does move, it means the dog hasn't been taught sufficiently that "sit means sit until I tell you what else to do" and that training starts very close to the dog, if the dog moves, then the trainer has moved on too far.

I happen to be just watching police dogs working at an airport and their reward is a tennis ball. My GSP loved food, however, once she was in the shooting field she would not entertain such a treat, her work was her all.

It all looks so difficult to me and I think dog training should be simple, I think that is what Judy said, she said it so much better than I have

Jelinga

[suerey]

I have really enjoyed reading this thread! I was new to gundog training six years ago when I got my second cocker (the first one was a pet, although working bred) and what little I had trained her (recall wasn't bad, she would sit, stay, fetch and return and then play tug) was done through positive reward based methods. When I started trying to train them both for gundog work, I read some gundog training books and went to training classes, went to another trainer all who used the traditional aversive training methods. This never did sit well with me. My younger cocker was a very sensitive, yet pushy and very fast dog and shied away from any sort of verbal or touch parise when we were training, if I scruffed her or was too dominant to her she would just sulk and refuse to do anything, so I found it very difficult to get anywhere with her. But at that time this seemed to be the only accepted way you could train a gundog. After a couple of years of trying this and not getting very far (behaviours were there but would fall apart at the first sdign of any distraction), and giving up any hope of being able to train them to work, I started reading up and finding out about positive training and clicker training. Since then I have used reward based training (food, chase a ball, retrieve, hunt) and although I will never now be able to get them to a stage where I can work them (too much done wrong in the past) , my relatinship with them has vastly improved and they now show me the respect and have bonded far more since using these techniques than they ever did before. They are now learning what I try to train them and both they and I enjoy our training sessions.

I know many people can train dogs well using the traditional aversive techniques, but I think more harm can be done to a dog by an inexperienced handler using these techniques (as I have found), than by using positive reward based methods.

[beth.james]

I don't actually agree with shaking a dog if it moves from the sit position as that usually then causes some anxiety and the dog cannot concentrate and learn its lesson. Jelinga

There's a big difference between shaking a dog and being aggressive with it, and scruffing it and silently leading it back to where he should be, which I think is what Peter was suggesting, and certainly the way I deal with any movement. The dog doesn't get distressed, I just calmly walk up to them, usually making sure I don't look them directly in the eye to intimidate them and then as I say, lead them back etc - to clarify, I'm holding them by the scruff, but they're on all four feet (probably mainly cos my dogs are labs I think with my springer I would lift him but he was at most 12 months old)

As with Jelinga, I would be very interested to hear any more of your regime Peter.

Maybe I'm being too narrow minded but I can't help but agree with Peter, Naj, in that that system seems to go all round the houses in order to get to the same point as Peter would be, but he'd have probably got there quicker. And just because it might be quicker I don't think that necessarily means it's worse or not as instilled in the dog. At the end of the day those commands and drills will continue to be run through and if the dog doesn't do it right, he'll be corrected and 'retrained' at any stage, whether it be 9 months old or 3 years old. I think if you had two say, 2 year old dogs, one trained with positive rewards and one trained in aversive training (I think those are the right phrases...) I'd have thought they'd both be at the same sort of level (obviously assuming the trainer's were of similar standard etc) and I certainly couldn't see any grounds for one dog being the 'happier' or more 'content' because of the way it learned things. Also, surely it costs more to be buying all these 'rewards'

Also, the cues you mention at the end of your message - well my dogs do that too... Not because they were given food or left to their own devices to learn it but because during their training they registered patterns and what was expected of them and so gradually you don't need so many commands. Things such as, if walking to heel and the trainer stops - the dog should stop automatically also, without a word.

At the end of the day, there is no right or wrong. If there was we'd all be training our dogs to the exact same timetable and they'd all be the same standard etc.

I find it very interesting to hear people's views on this sort of thing, I think it's how we learn. At the end of the day all our knowledge for training our dog's comes from other people and then our own experience.

[Wendy M]

Pete – I can feel your confusion!! When I first started reading about this way of training it seemed to be turning everything I knew about dog training on its head! But the more I read the more it made sense to me.

Dogs are perfectly capable of working through a problem (“what do I need to do to get something I want?”) IF you allow them to explore different options. It doesn’t take long for even “cross-over” dogs (dogs that have been previously used to correction based training) to work it out.

Food is just one thing you can use (I mainly use cubes of cheese about a centimeter square or chunks of sausage if I'm throwing it - for training the stop/sit at distance) but if you stop and think about it your dog will be motivated by a whole multitude of things. One article I read recommended drawing up a heirarchy of the things that motivate your particular dog so that you can see how you can use them in different situations to help your dog learn.

For example: (and obviously different for different individual dogs)

BRONZE:
kibble
opportunity to go through doorway
opportunity to get into/out of car
cheerios
having leash put on
physical contact (pats)
praise
SILVER:
natural balance/red barn food roll
Wellness Pure Rewards
opportunity to smell interesting scents on ground
cue for an established behavior
fetch
hot dogs
GOLD:
liverwurst
peanut butter
canned dog food
meat balls
opportunity to swim
PLATINUM:
opportunity to greet other dogs
opportunity to greet approaching human
opportunity to chase squirrels
mashed potatoes and gravy
steak
raw liver/hamburger
tug
deer poop

When the dog has fully learned the behaviour, you do not need to go “back” to praise, as the dog has learned that by performing the desired behaviour good things will come . i.e. Heeling, brings the opportunity to sit, which brings the opportunity to be sent away, which brings the opportunity to retrieve - the end reward for the whole chain of events.

If behaviour starts to break down at any point. Such as your example, pete, of the dog creeping on the stop, we go back to training, take it back to the point you need to work on and build it back up again using an appropriate motivating reward. This is where a Clicker or some other event marker is helpful to mark the exact moment the dog performs the bit that you want. I was just working on creeping with my dog the other day. I made the exercise a bit easier – less distance from me – then began clicking and throwing a reward for less and less creeping till we got back to immediate stops then extended the distance again, bit by bit, until the dog “got it” again.
Then we put it back into the chain of heel, sit, send, stop, cast, retrieve.

You asked how long it might take to for a dog trained in this way to be ready for the shoot – well my dog has just turned two and I will have her out this pheasant season. Obviously still a work in progress but as you say – a gundog is never fully trained! and incidentally, my father-in-law trains/trials labs using traditional methods and has my dogs brother out of the same littler and we are at similar stages of training.

I'm NOT saying this method is RIGHT and yours is WRONG – far from it! All I’m pointing out is that this is a method that does work.

And of course I’m interested in how you train – I’m interested in how everyone trains and how dogs learn.

Wendy

[drumkeen gundogs]

Final Final response Dog training is simple its common sense its practical its physical but all in good ways believe me dogs respond more to positive action rather than negative action , when people bring dogs to me for training I can see straight away if a dogs has been hit or badly treated that has such a detrimental effect on a dogs ability to learn and respond positively , I just can't seem to get my point across or I am not explaining well enough, to be successful at training dogs you have to build a confident bond with the dog some people have this natural ability and others have to work at it to achieve, that bond has to be like parent / child relationship you are teaching a dog specific skills that are required for its role in life, for the dog it should be fun and happy times but controlled my regime of training works so damn well because it takes the dog from one success to another ( maybe regime is not the right word thats people keep thinking it sounds like a military exercise ) the dog grows in confidence from day to day , there are times in the early days you have to correct the dog but in a calm and controlled manner eg teaching Bob to sit and stay whilst I climb a fence half way Bob moves and runs under the fence so I have to climb back over call the dog back sit him again with the whistle and give firm voice command of STAY flat hand also, if he moves again you have to do it again untill he doesn't patience and consistancy if you hit shake the living daylights out of him he wont relax and thinkabout what he is supposed to be doing, all my dogs when reaching a fence or style etc sit down and wait without me having to give the command because they have learnt and are now using their intelligence .
I can't stress enough how important that connection between the dog and you is I have a phrase when I'm talking to clients " switched on " try this on your own dog and make your own mind up if your dog is " switched on to you " then you will know if your training is working when I take my dogs out of the kennel and begin to train the dog or dogs hang on every movement / gesture and so important eye contact watching all the time for what does he want us to do next , that is a special feeling for any owner / trainer , you can always tell an owner and dog where there is no bond no switched on partnership the dogs mind begins to wander so quickly and they break the command given in seconds rather than minutes, mostly they don't pay attention at all and struggle to learn because that bond and understanding has not been achieved by either party and no amount of treats / food etc will change it as I said I feed my dogs in the comfort of their kennels when we go training or shooting its fun and very enjoyable.
Remember if a dogs tail is not wagging its not confident or happy
Best Wishes Pete

[beth.james]

i.e. Heeling, brings the opportunity to sit, which brings the opportunity to be sent away, which brings the opportunity to retrieve - the end reward for the whole chain of events. Wendy

But this is how my dog's interpret things already... without the need for food rewards... They quickly learn that if they sit, there's the 'opportunity' to do something more exciting - leading up to the retrieve. I'm sorry, I just don't see the need to complicate things with adding the need for food rewards. The only time I think I would consider using food rewards would be for a food motivated, nervous dog that needed to come out of its shell and gain confidence and even then it'd be the last resort. Although, after reading all this I might do it one day with one of my dogs just to see, but since so far my methods seem to have worked, I don't have that push to change so drastically.

I'm certainly not saying your method is wrong, and I don't believe you are either, it's just a discussion and a very interesting one at that

[Wendy M]

i.e. Heeling, brings the opportunity to sit, which brings the opportunity to be sent away, which brings the opportunity to retrieve - the end reward for the whole chain of events. Wendy

But this is how my dog's interpret things already... without the need for food rewards... They quickly learn that if they sit, there's the 'opportunity' to do something more exciting - leading up to the retrieve. I'm sorry, I just don't see the need to complicate things with adding the need for food rewards. The only time I think I would consider using food rewards would be for a food motivated, nervous dog that needed to come out of its shell and gain confidence and even then it'd be the last resort. Although, after reading all this I might do it one day with one of my dogs just to see, but since so far my methods seem to have worked, I don't have that push to change so drastically.

I'm certainly not saying your method is wrong, and I don't believe you are either, it's just a discussion and a very interesting one at that

I'm not saying you NEED to use food rewards! Use what ever motivates you dog. Its just that food is sometimes easier in the beginning when starting to train behaviour. and sometimes its easier to use food as you can just give it to your dog and get on with the next repetition or next exercise quickly without having to wait for your dog to go out get the ball/dummy etc and bring it back.

In heeling for example, to get the message across that the correct position is the only way to get a reward this is going to be much more time consuming if you throw a ball and wait for the dog to come back for every correct step in the beginning than if you can click and give a cube of cheese as you are walking.

Wendy

[cornishman]

I like to keep it simple

[beth.james]

I have to admit, it's starting to make sense to a degree. You're good at getting your point across

With heelwork, of course, throwing a ball every time he gets it right isn't going to work - as you say, very time consuming, and dependant on the breed, counterproductive giving him so many retrieves. With the young pups I simply tend to for example, only give them a little stroke when they're walking in the right place. It's not a 'training session' so to speak, it's more like conditioning general conduct. Then when they're ready for 'training', it's on the lead (if they're starting more or less from scratch - ie not had them right from a pup) teaching them to heel, then when they're happy there, dropping the lead across their back or wrapping it around their neck (not strangling them just so they have that weight there and there's the 'handle' of the lead coming out the top for you to hold if they pull on too far, thinking they're off the lead) then no lead... et voila If you progress at the right time you shouldn't have to go back a stage. The dog also 'learns' as you say, that the only time he'll get any 'fun' ie retrieving, or hunting dependant on the breed, is when he's walking calmly at your side.

If you're giving a dog treats when training heelwork, do you not find that the dog pulls a bit too far forward and sort of 'curls in' towards you in front of you, looking for the next reward? I know watching you can be good, but you need the dog in the right place and to progress to looking forwards 'marking' anything in front of you.

edited to add (cornishman posted while I was writing) yes that's my point exactly, I just feel that it's over complicating things and going round the houses to reach the goal. If you've got the time and inclination to go about it that way that's great, everybody has their own methods, I just can't see myself ever going down that route voluntarily. But it's always interesting to hear about