Reaching into a pocket for a treat or toy should only happen during training when luring shows the dog what is required (ie look at me, play with this instead etc). Once that very early stage of training is achieved, the lure drops and becomes a reward, ie is given after the dog has achieved what's required. From there, the level of reward drops for the most part to 'good boy/girl'. However, a good trainer will produce the most valued reward every now and again to keep the training keen. Imagine playing a one-armed bandit - the thing that keeps you feeding in the money is the possibility of a reward. Now think about a sweet machine - although we get rewarded every time with whatever packet of sweets we have selected, we don't get any incentive in keeping feeding money into the slot despite getting that reward every time.

The 'jumping up/traffic light' scenario. Let me turn this around a bit. You are taken to a factory, placed in front of a machine, given a pile of components and are expected to put them together. You would more than likely get it wrong.

BUT, if you are taken to a factory, taken to a machine with a skilled operator who has a pile of components at the side who trains you how to assemble them, you are much more likely to get it right when you are taken to you own machine and given your own pile of components to assemble

.........................

And when it's your turn to try under supervision to assemble the machine.. any mistakes will be corrected by the use of the word.."no" perhaps?
"No Ian, that part goes here not there" a negative response with positive results... IMO



I would have said that if mistakes were made then the training wasn't as effective as first thought so back to training . I think this scenario is exactly what we encounter with dogs. We assume that we have trained them well and that they 'know' what is expected, when obviously the training hasn't been as thorough as initially thought.

Question: When dogs get it wrong, they very often receive a jerk on the neck from a choke chain, a slap, are shouted at. How would you feel as the trainee if, instead of telling/showing you where you were going wrong by further training, the trainer smacked or shouted at you for getting it wrong?

Just to add, which do you think would be the most effective way to get you to want to work, the correction, or being shown again what to do before trying again?

Totally agree with what your saying. And would NEVER jerk, slap or abuse my dog in any way. But during training mistakes will be made. And in my opinion thats when "no" comes to use. I don't shout it, I say it and my dog responds. In some circumstances I may have to say it a little more sharply (usually in an instance where the result of not saying it quickly and sharply could cause harm to the dog). As a new puppy I would say "no" while lifting Jake away from something that he shouldn't be at. From that he has learned that no means no... I don't think that is cruel or harmful, I think it is helpful and neccissary.
So going back to the title of the topic...
"100% Positive Reinforcement training, does it really work?"
I believe showing the negative (in form of "NO" or any other command you want to use) and rewarding the positive... is the best way for me Just like teaching kids really... Teaching right from wrong...

It was Dogwise

Thank you. I'm a little brain-dead tonight :smt002

We can always rely on Wysi for the book stuff

Very good discussion guys

Well, I am undecided. I've learnt so much since starting Dogsey, that I never say never and will always have an open mind.

With me, I think I do my best to try and learn their language though, as much as I try to learn 'our ways' to train 'them'.

Unfortunately, it's not possible to be with my dog 24 hours a day, so while 100% positive is certainly an attractive proposition, I'm not sure if I could manage it. For eg, as with Ian's example of the fire situation, what happens if you leave the room for a minute and the dog goes to it? 100%+ might coax him away if you're there, but without 'being told' it's wrong beforehand, how does he know he shouldn't do it when you're not?

From my experience, if Rocky does something that's not desired, like chew the door handle I will say 'no'. Not too loudly I might add, as I would rather reserve the really loud 'No's in case he was doing something really bad (but even then use it more as a distraction/attention grabber than anything else aka aversion training).

In the past, if a simple 'no' didn't work I kind of did a no-cross-growl sound - a bit like a gurgling scary monster, saying Noo oOo ooO oo. For some reason, he stopped to take notice, and regardless of whether he went back to doing what he was doing or not I 'think' he took it as me not being happy with whatever it was that he was doing. Like I said, I no longer do this, and can only remember doing it once or twice, as *touch wood* he's as good as gold and rarely do I need to stop him doing something (or if I do a simple 'no' or 'down' or 'slow' is suffice, depending on the situation). I mention it because similarly, I will 'hiss' at the cat if she goes near the tv wires etc (240v shock wouldn't be nice!). The hiss has immediate effect, she takes notice, where as you could be shouting at the top of your voice and she wouldn't bat an eyelid! Yet one small hiss and it does the trick. It's true what they say, dogs come when called, cats take a message and get back to you! (but not when a hiss is involved!)

As the growl is a sort of 'I'm not happy' in dog language, the hiss means the same for cats. Now I wouldn't suggest anyone try this, it is something I have done in the past but just wanted to mention it as it is relevant to my experience which this post is about. I have also done alpha rolls, as, that's the kind of thing I was reading up on when Rocky was a pup. Interestingly, after an alpha roll, you notice a difference in his behaviour - it's really quite interesting, it's almost like it makes him 'super happy'/ excited? He becomes very playful.... but again, I haven't done this for a long long time. Now he will roll-over and invite tummy rubs at every chance he gets.

It has certainly been a learning experience for me and by no means am I an expert - but thanks to people like SB, Mini and many of you guys I have learnt, and continue to learn new stuff everyday. Incidentally the hiss/growl thing wasn't learnt from anyone in particular, it was from me reading up on animal behaviour (and living with them since a child) and applying that in what came naturally to me, and what felt right in my situation given what I understood.

So through my experience, a combination of things have worked; pack hierarchy, clicker training, treat based, praised based etc, and to be honest, I don't think I personally could have got Rocky to the stage he's act in the time we've had together - using 100% positive, where it seems you have to be there to 'counter' undesirables with desirables a whole lot more.

But can it work for others? Well according to at least 2/3 people here it can... maybe we need to pick their brains more often (Get writing some articles guys! ps Wysiwyg, I haven't forgotten yours!)

Please note: Anything I have written in this post should not be attempted by a novice or anyone who doesn't fully understand what they are doing, I am just adding my bit to this discussion. Some of the things could be dangerous if not applied correctly.

I also think it's a brilliant and very interesting discussion. And I don't think I disagree with anything I've read here. It's one of those diff strokes/ different folks type things.

But... always a but eh... imagine a senario where there is no human intervention...

Lets take my two, Jake and Charlie. Charlie attempts to take Jakes bone/toy whatever... Jake growls or maybe barks a warning to Charlie... A lesson learned from a "negative action"? Now in my mind that is no dif to my "NO" warning, infact I'd say it's a little more severe.

Or again with no human intervention... Charlie decides to have a sniff at the fire, gets too close and burns his nose. Again he has learned something from a "negative action" and again a little more severe than my "NO".

I believe in order to learn whats good/acceptable bahaviour, a dog needs to learn whats not. And if I can teach him that with a gentle telling off for the negative and a bucketload of fuss and praise for the positives... I reckon he'll be doing ok

Yep And like you said, it's different strokes for different folks really.

'Negatives' are a part of our natural life, it's all about survival - feeling, for example stops us hurting ourselves by 'teaching' us from an early age that some things hurt! (like fires!) Screaming out loud when someone trods on your toe, tells the other person they need to get off!! But I guess people into 100%+ feel that when it comes to training dogs, they can get by without any negatives.

Is it tougher? I would think so.
Does it take longer? Again I would think so.
Is it better? Well, that's a matter of opinion - as I don't think we know enough yet to easily form that conclusion... or do we?