CockersOnline Forum
Cocker Specific Discussion => Behaviour & Training => Topic started by: ips on March 28, 2017, 11:39:50 AM
-
I have never really given much thought to "methods" as I have just muddled along using various ideas from various sources and made much of it up myself. This "none method" has possibly hindered proceedings and I may stick to one particular method next time. Anyway there seems to be so many different "methods" out there that it is very confusing for many (including me) of all the different ways the one that seems closest to the way I do things is "shaping" but without the clicker. This brings me too the question, is it feesable to introduce the clicker at this stage in order to polish a few things as in a nearly three year old. I would not be so bold to call her fully trained but she has worked two seasons beating and has picked up and is generally biddable.
Any thoughts ???
-
It's never too late to introduce the clicker. It just gives you one more way to teach new behaviours and keep it interesting. ;)
-
Agree with Joules, never too late to introduce the clicker :003:
I started out by reading a ton of books on positive reinforcement training and clicker training. For me that theory is the only way to go. But I realised that no two dogs are the same, so you actually can't follow one method to the letter. I sometimes overcomplicated things in my efforts to be the perfet positive trainer. Sometimes common sense works better because it suits your own style and the dog responds to you being comfortable with what you're doing.
In my opinion really good trainers follow a theory based on modern learning psychology, but they adjust the method to fit the dog. So probably similar to what you allready do ;)
The theory behind clicker training is brilliant, I always base all training on positive reinforcement. But I rarely use the actual clicker anymore. It is a really helpful tool in training new behaviours, but the important thing is the theory and the timing. I only bring the clicker out for teaching something entirely new or for fun now.
I think the best way to go is to learn the theory of clicker training and then with that knowledge as a base do it your way. Being flexible enough to "muddle along using various sources" is probably a good thing :003:
-
Thanks joules.
I am very interested in trying it. I think I avoided it early on due to a perceived idea that it was a bit mamby pamby for gundog trainers 😁 with hindsight I think it could be a useful tool. I understand that timing is critical (obviously) and that one must first "load the clicker". My only concern is that once you start are you then not stuck with a "clingy" dog expecting treats for every action ?
I trained initially with food treats but have all but faded them out now that the behaviours are learned, is this the same for clicker training is must it be ongoing ?
-
Agree with Joules, never too late to introduce the clicker :003:
I started out by reading a ton of books on positive reinforcement training and clicker training. For me that theory is the only way to go. But I realised that no two dogs are the same, so you actually can't follow one method to the letter. I sometimes overcomplicated things in my efforts to be the perfet positive trainer. Sometimes common sense works better because it suits your own style and the dog responds to you being comfortable with what you're doing.
In my opinion really good trainers follow a theory based on modern learning psychology, but they adjust the method to fit the dog. So probably similar to what you allready do ;)
The theory behind clicker training is brilliant, I always base all training on positive reinforcement. But I rarely use the actual clicker anymore. It is a really helpful tool in training new behaviours, but the important thing is the theory and the timing. I only bring the clicker out for teaching something entirely new or for fun now.
I think the best way to go is to learn the theory of clicker training and then with that knowledge as a base do it your way :) Being flexible enough to "muddle along using various sources" is probably a good thing :003:
Great post Mari.
That makes a lot of sense and has answered some of my thoughts.
I fully agree that one must adapt to the dog in front of you or more importantly the "relationship" and how you "interact" with your dog. 👍 maybe my "muddling along" method is not quite as haphazard as I thought 😁
-
My only concern is that once you start are you then not stuck with a "clingy" dog expecting treats for every action ?
Nope, because you only reward for every repetition in the beginning, during the learning phase. Once the behaviour is linked to the command and the dog is doing it reliably every time you ask, then you fade out the rewards. They don't need the click every time when they know they are doing the right thing. So you reward more sporadicly and fade out rewards just like with other positive reinforcement training. And it sounds like you already are familiar with that concept :) The clicker just helps with the timing in positive reinforcement training, theory is mostly the same.
-
Thanks for clarification.
So really the method I currently use remains the same. Shape the behaviour but use the clicker to mark the wanted behaviour more precisely than a "good girl" any attempt at what you want gets a click and a reward then only mark the more "perfect" behaviours untill eventually fade out over time by only treating"exceptionally" well carried out wanted behaviours ??
If I am correct then it should be a doddle as the above is fundamentally how I approach training 😉
-
Yes, that is the basics of it :)
I've always been bad at shaping, so I mostly capture behaviours and finetune them in the way you describe. Or I "cheat" and use a target stick/target pad or strategically placed rewards. But it's so much fun to see a dog that is trained using a lot of shaping, you can see their brains working really hard to figure out what the trainer wants rather than constantly waiting for cues from the trainer. Very cool method!
-
I have used clicker training with both my previous dogs with great success. Not jus to teach basic behaviour but specific tricks too (pointless but great fun) With Diesel were still charging the clicker as he's more focussed on the treat coming so very focussed on me anyway than relating it to the actual click at this point.
-
Ok, so. One of the things I want to polish is heel. She comes to heel no problem BUT occasionally she will start to pull forward (especially on a beating line with temptation of scent and birds) traditionally pooch would be dragged back but I don't do physical correction. I want heel to mean heel until I release. At the moment in these scenarios of high temptation it takes constant nagging. Would the clicker be of any use in training or rather improving heel and if so how would I start. ???
-
That is a tough one, way too difficult for my level so hopefully someone else has better and more specific advise. I have mostly worked with fear agression in dogs, and teaching them to trust the world and behave like civilized members of society :005:
But one very general thing I've learned is that positive reinforcement training is about training up to something. It has one disadvantage and that is that training takes time, effort and patience. In the moment, when adrenaline is high, it is almost impossible to reach a cocker. So the reward based training must be started on an easy level and then you work up to the final goal.
Training heel in different scenarios with less temptation should help. When heel is perfect without disturbance then add something that is distracting, but not too distracting. Someone playing with a ball or running around shouting maybe. Keep adding distractions until she is so good at heel that she can do it in every situation.
Oh, and obviously with such a difficult goal the reward needs to be high value and action filled :)
Added: I never trained that kind of heel. But I have used clicker to train walking without pulling. And by the end she was actually heeling so I had to train less because that wasn't my goal. What I did was change directions alot and rewarding her for following with a loose lead between us. So if she wanted to earn clicks she had to really focus on my movements. I also rewarded any time she offered eyecontact. Started out with no distractions and moved to more challenging enviroments. Not sure if this helps at all :)
-
You are right that the clicker is basically a highly effective marker of the behaviour you want, but when you see it in action, you'll realise how powerful it is. For one, the sound of the click is much more attention grabbing for Henry than anything I can do with my voice. It means a precise thing and always sounds the same (like a whistle). Second, once you get the timing, marking the EXACT moment the dog does what you want is incredibly reinforcing.
With the heel work, you can click the instant she is exactly where you want her, and, if you chose, intermittently when she stays there. I haven't trained heel work, but I've used the clicker to really sharpen behaviours we already have. Henry had a 'lacksidasical' sit - he'd do it when he wanted, in his own sweet time. A few sessions with the clicker reinforced that I wanted an immediate sit - at first I clicked and treated every reasonably prompt sit. Then I only clicked and rewarded quick ones. And then only the instant ones. It's a great way to fade out behaviours that are only approximations of what you want but the dog has decided are good enough!
-
I use the clicker every now and again. I did sometimes find it difficult managing three things - leads, whistle, treats etc. But I do like it for shaping and watching the dog try to work out what you want them to do!
The clicker also carries no emotion so nothing can be transferred via your voice. Our trainer advised me to use a clicker with Maddie's fear aggression as my voice was full of tension and 'transferring down the lead'
-
I trained heelwork with my last cocker with the clicker. As people have mentioned its a slow process. 3 steps, click reward, 3 steps click reward, 4 steps etc. I found doing a slow jog was more interesting for the dog to keep up with than walking initially.
-
Thanks, that is all good info.
To date I have trained by changing / shaping the behaviour. For instance the break away from a bird became a break from flush and ultimately shaped into a sit or stop to flush. Her retrieves and directions are good, she loves the three card trick 😁 her hunting is good.. So most things are good (to my eyes anyway) I just need to get on top of this pulling forward from heel when temptation of the flush point is upon us. It is very very difficult without using aversives I fear. Stop is another thing that "can" fall apart not always but again if the temptation is too much.
Maybe my expectations are too high, I am by nature a perfectionist and also impatient which means I am rarely fully satisfied with a job.
I think I may give the clicker a go but how effective it will be for what I need remains to be seen.
-
I find that anything that requires slowing down is extremely difficult to teach a cocker, it takes so much longer to teach them to stop and stay than to do anything active. But I'm thinking it could help if you train impulse control in different settings too. Imo that's the sort of training that helps all other training go easier. Just doing sessions to improve impulse control and help them to cope with frustration (frustration of having to wait/go slow). I'm sure you'll get there, didn't you post the impressive video of her sitting perfectly while you throw dummies?
Hope this doesn't sound bad, because I'm a huge fan of positive reinforcement training and I would never use aversives. But I have realised that it is possible to correct behaviour without actually punishing. I would never yank the leash or yell or anything like that. But having a command that means "wrong" or actually physically guiding the dog, allthough not really clicker theory, can be quite useful. I would always go for the clicker way first, but as you say, this is difficult stuff, so much instinct is involved.
But of course, I'm just talking general training methods, I've never trained a hunting dog. I just think this is really interesting :)
-
Mari
Yes she is steady as a rock with dummies (yes was my vid) and generally reliable. It probably sounds a bigger issue than it is. These misdemeanours are not "that often". I think you have hit the nail on the head re the impulse control. Now that everything is pretty much trained and understood I am concentrating purely on trying to calm her (yes I laughed when I typed it, good luck with that I hear you all say 😁) and get her able to control her instinct when under a lot of pressure. As you know its all well and good being steady on dummies and steady to flush but at the end of a drive particularity early season we are in huge amounts of birds and huge amounts of scent and the general level of excitement of a shoot. The dogs at this specific moment are under enormous temptation and pressure.
I heard a saying a while ago that I shall plagiarize "my dog never breaks said no man ever" in other words dogs instinct is always simmering away waiting to boil over.
I will try anything once BUT I have and never will raise a hand to her, I have no issues with growling at her when she has blatantly ignored a command that I know she fully understands. She also knows from my demeanour if I am less than impressed 😁
-
Calming a cocker :rofl1:
Mari
Yes she is steady as a rock with dummies (yes was my vid) and generally reliable. It probably sounds a bigger issue than it is. These misdemeanours are not "that often".
It doesn't sound like issues at all. Sounds like she is a really good dog and gets as much training as a dog can dream of. I just like talking training theory :005: Sometimes someone will mention a method that I have forgotten about that is really useful.
-
Me too Mari, I like trying to work out what works, or not as the case may be and I enjoy discussing it with other more experienced than I. I am beginning to think that a hybrid method can work and that one does not necessarily need to sign up to one specific regime. Its as though one method works well for one specific thing but not for another so it a case of thinking on your feet so to speak. For instance I tried many different methods to steady her to a thrown dummie and was getting no were fast after six months, so I threw the hypothetical book away and did something else which included something that your not supposed to do, but it worked.
Ps
The not supposed to do thing was to get after her when she ran in and read her the riot act as she picked the dummy. This supposedly is punishing the retrieve (the last act before the reprimand) but in my head she knew by now what I wanted but chose not todo ut so I reckon she knew what she was in bother for despite my poor timing 😁
-
Training a "look at me" is very effective with the clicker Ian. Humphrey tends to be a step in front when walking to heel, I've been getting him to "look" as soon as he starts to move too far forward, I then click and treat holding my hand where his head should be, so that he has to stay in line to get the treat. Don't know if that might help you?
We went to a dog club on Sunday, by the way, and were welcomed with "ahhhh yes, a cocker spaniel, you can reckon on him being fully trained by the time he's about 12!!! " :005: - I actually found that quite reassuring as at last there's someone there that's not got any unrealistic expectations of us, so I don't feel under quite so much pressure!!
-
Thanks BL yes that is helpful .
12 ...ruddy ell, I better get a spurt on 😁
-
We went to a dog club on Sunday, by the way, and were welcomed with "ahhhh yes, a cocker spaniel, you can reckon on him being fully trained by the time he's about 12!!!" :005:
I have also heard that "Labradors are born half trained: spaniels die half trained" :rofl1: :rofl1:
At least they are never dull ph34r :005:
-
Yep I have heard that 😁
-
Thanks, that is all good info.
To date I have trained by changing / shaping the behaviour. For instance the break away from a bird became a break from flush and ultimately shaped into a sit or stop to flush. Her retrieves and directions are good, she loves the three card trick 😁 her hunting is good.. So most things are good (to my eyes anyway) I just need to get on top of this pulling forward from heel when temptation of the flush point is upon us. It is very very difficult without using aversives I fear. Stop is another thing that "can" fall apart not always but again if the temptation is too much.
Maybe my expectations are too high, I am by nature a perfectionist and also impatient which means I am rarely fully satisfied with a job.
I think I may give the clicker a go but how effective it will be for what I need remains to be seen.
Perhaps you are a bit too much of a perfectionist 😜. Eze sounds fine. My dog Pip is very good in the beating line - very responsive to turn and stop whistle and has never ruined a drive. However sometimes we have to wait right near a flushing point where birds are going up and then falling and Pip gets super excited watching as he is mad for retrieving after shot. So I put him on lead. As I feel it's a huge amount of pressure on him expecting him to sit there watching loads of birds coming down in front of him with no prospect of an immediate retrieve... I think we see these super well trained trial dogs looking impeccable and think ours should be the same. But really, most of the trial dogs aren't exposed to that extreme pressure/temptation all day. I also just put pip on lead between drives if he can't control himself and just hares about hunting. It's the pressure/excitement again. I feel putting him on lead stops me nagging him and allows him to relax a bit. I feel it's huge discipline and strain him having to be obedient while beating or waiting to pick up. Heeling in between off lead is further stress that he probably could do without. I also think that many of the dogs you see on shoots just milling about behind their handlers most of the day aren't shall we say 'high drive' types. Dunno if any of that makes sense?! Just my opinion.
-
Thanks, that is all good info.
To date I have trained by changing / shaping the behaviour. For instance the break away from a bird became a break from flush and ultimately shaped into a sit or stop to flush. Her retrieves and directions are good, she loves the three card trick 😁 her hunting is good.. So most things are good (to my eyes anyway) I just need to get on top of this pulling forward from heel when temptation of the flush point is upon us. It is very very difficult without using aversives I fear. Stop is another thing that "can" fall apart not always but again if the temptation is too much.
Maybe my expectations are too high, I am by nature a perfectionist and also impatient which means I am rarely fully satisfied with a job.
I think I may give the clicker a go but how effective it will be for what I need remains to be seen.
Perhaps you are a bit too much of a perfectionist 😜. Eze sounds fine. My dog Pip is very good in the beating line - very responsive to turn and stop whistle and has never ruined a drive. However sometimes we have to wait right near a flushing point where birds are going up and then falling and Pip gets super excited watching as he is mad for retrieving after shot. So I put him on lead. As I feel it's a huge amount of pressure on him expecting him to sit there watching loads of birds coming down in front of him with no prospect of an immediate retrieve... I think we see these super well trained trial dogs looking impeccable and think ours should be the same. But really, most of the trial dogs aren't exposed to that extreme pressure/temptation all day. I also just put pip on lead between drives if he can't control himself and just hares about hunting. It's the pressure/excitement again. I feel putting him on lead stops me nagging him and allows him to relax a bit. I feel it's huge discipline and strain him having to be obedient while beating or waiting to pick up. Heeling in between off lead is further stress that he probably could do without. I also think that many of the dogs you see on shoots just milling about behind their handlers most of the day aren't shall we say 'high drive' types. Dunno if any of that makes sense?! Just my opinion.
Perfect sense. I see a lot of eve and I in that, very similar to how we work the trouble is she is worse on a lead but I am working on that too. As for FT dogs you are correct from what I understand those dogs will never have the pressure of a beating line.
Cheers
-
Pip is appalling on lead. So I do the lazy thing and make a figure of 8 with slip lead to go over his nose ;)
-
There is an old saying "The way to ruin a good dog is to take it shooting" All of my dogs over the last 40 years have been perfect until about 18 months and they went shooting.
It really does depend on what you want out of your dog. Most of my shooting was walked up rough shooting, rabbit shooting and wildfowling.
My dogs all ran in to the fall of a bird and chased rabbits. Shooting ducks in the dark I relied on my dogs to know when a bird was hit and to retrieve it asap so they learnt to run in they knew when a bird was hit more often than I did. If they had waited for a command birds would have been lost in the tide of the estuary.
A great deal of my shooting was rabbits in fact for 3 years I worked as a professional rabbit and pigeon shooter making most of my money from selling game I had shot. The ground I worked was mostly hill covered with gorse some patches of gorse could be 30 or 40 yards long dogs would go in one end and rabbits would come out the other end normally with a spaniels 20 yards behind.
I once asked Peter Moxon a very famous trainer at the time how to stop my dogs from chasing rabbits his first answer was don't take them rabbit shooting I explained the sort of rabbit shooting I had and he said " so you want your dogs to chase rabbits 30 or 40 yards through gorse and the magically stop when it suits you, good luck mate" he said.
So my dogs would not win trials, and at the posh shoots they were on leads at times but they were great shooting partners, I often gained shooting through my dogs reputations they were hard bold dogs even Henry the show cocker
I rarely ever used a whistle or gave them recall we shot at least 4 times a week and sometimes 6 days a week they learnt their jobs and I trusted them to do the job I needed.
-
There is an old saying "The way to ruin a good dog is to take it shooting" All of my dogs over the last 40 years have been perfect until about 18 months and they went shooting.
It really does depend on what you want out of your dog. Most of my shooting was walked up rough shooting, rabbit shooting and wildfowling.
My dogs all ran in to the fall of a bird and chased rabbits. Shooting ducks in the dark I relied on my dogs to know when a bird was hit and to retrieve it asap so they learnt to run in they knew when a bird was hit more often than I did. If they had waited for a command birds would have been lost in the tide of the estuary.
A great deal of my shooting was rabbits in fact for 3 years I worked as a professional rabbit and pigeon shooter making most of my money from selling game I had shot. The ground I worked was mostly hill covered with gorse some patches of gorse could be 30 or 40 yards long dogs would go in one end and rabbits would come out the other end normally with a spaniels 20 yards behind.
I once asked Peter Moxon a very famous trainer at the time how to stop my dogs from chasing rabbits his first answer was don't take them rabbit shooting I explained the sort of rabbit shooting I had and he said " so you want your dogs to chase rabbits 30 or 40 yards through gorse and the magically stop when it suits you, good luck mate" he said.
So my dogs would not win trials, and at the posh shoots they were on leads at times but they were great shooting partners, I often gained shooting through my dogs reputations they were hard bold dogs even Henry the show cocker
I rarely ever used a whistle or gave them recall we shot at least 4 times a week and sometimes 6 days a week they learnt their jobs and I trusted them to do the job I needed.
That was a great read Jeff and certainly puts things into perspective 😊
-
Update
Haven't tried the clicker yet but having researched "capturing" and "shaping" it seems that I have been doing a version of these all along without even knowing it. So to put it to the test I have been training her a new "behaviour" but with more thought in what and how I am doing it.
A few times in the evening I sit at the kitchen table to smoke a cigar (terrible habit I know) eze brings tv remote then glasses case then slippers etc etc. I have never told her off which is why she will retrieve a pheasant, partridge, woodcock with vigour but now that this live retrieving is well and truly tested and proofed I thought I would use the scenario to train something else that I want which is eze sitting, lying or just mooching around at my feet whenever I stop on the beating line, so. I am now four days in and its already very close to being reliable (in the kitchen anyway)
I started to ignore the household items and not give her a reward ie biscuit or any verbal at all. The first few occasions she brought a couple of items then realised something wasn't working and she sat, I instantly gave a "good girl" a few times she lay down which also got a "good girl" the amount of items brought got less over a few days and the last six or so occasions I think only once did she bring an item all other times she has sat or lay down (went to sleep twice) at my feet and waited patiently until I had finished my cigar and then every other"ish" occasion treated with a biscuit.
Cant believe how quickly this method works and I can really see now how and why you could add a clicker into it.
-
A clicker worked very well for Blue, but not for me. I have a connective tissue disorder which affects my hands and I just didn't have the manual dexterity necessary to operate the clicker consistently at the right moment to mark the behaviour, so I ended up reinforcing all the wrong bits! I've been told that some use a click of the tongue or a key word instead, which would certainly work better for me.
-
I have always given a lot of "good girls" in fact I have been pulled up for it from gundog folk saying I should not and that its too often however having recently researched clicker, capturing and shaping I now realise that in effect my verbal praise is a "clicker" so in my humble inexperienced opinion I wouldn't worry about it I would just use your usual verbal praise, as long as its given instantly 👍
-
I know people with deaf dogs that use flashlights or small laser thingies instead of clickers to mark correct behaviours. It's all the same theory :) I use the clicker because I tend to say "good girl" in situations where I don't reward too. But I could have chosen a word as a clicker if I remembered not to use it for anything else. Some feel it's easier to get the timing right with clicker, but others don't like the extra gadget and prefers a word or tongue click or even a whistle. Whatever works :)
Eze sounds like a clever dog, perfect for clickertraining!
-
I know we all think the same about our dogs and so we should but yes she is very very clever. Once she "gets it" then its generally reliable from then on. Her only issue is being lumbered with a novice inexperienced trainer who doesn't always explain things very well to her 😞 if someone with experience had trained her I know for a fact that she would be capable of field trial championships no question about it. Her FT pedigree is stunning and I feel honoured to have had the opportunity to learn from her.
My only concern with the clicker is doing it wrong and mucking up what we have. I rarely seem to do things text book I seem to always just middle along, I am a very awkward and stubborn man....apparently 😁
-
My dog has had the same problem :D It's hard to reach your potential with a novice teacher. But it sounds like you guys are doing really well so far! Learning together is half the fun imo. Of course, I've never competed in anything with my dog so that's easy for me to say, everything we do is for fun :P
-
It is indeed a lot of fun. 👍
I particularly enjoy the mental side of it, trying to work out how to train it or untrain it as the case may be. 😁
-
You may be a novice, ips, but you have great instincts. The way you have taught eze to settle at your feet is exactly the way I saw in a training video by a great trainer (Kikopup on YouTube if you're interested). I find the behaviours I get most reliably are the ones where Henry has had to work it out and offer it for me to capture.
-
Thank you LG you are most kind. Yes interested in seeing that, have you got a link ??
-
Here you go: https://youtu.be/wesm2OpE_2c
-
Brilliant. 👍
I like that its exactly my kind of approach. Learned some good tips from it.
Thanks for the link 👍
-
All her videos are excellent, well worth browsing.
-
Two occasions so far this evening I have been sat in kitchen both times she has sat then lay down in front of me, no big pile of items on the table, nothing brought in. I am yet again amazed how quickly she picks things up. This capturing / shaping is a powerful tool 😁
-
Brilliant! Wish I had been clever enough to capture that behaviour. 'Settle' has always been a difficult thing for us! She didn't get it until she was at least 5 (years, not months) :005: And I like kikopup too, good channel!
-
Well, it may be my imagination but over the last few days since doing the capturing thing I have noticed a general calmness about her. Is this coincidence or tied up with the capturing and or the lead work I have been doing. There is definitely a slight change in her demeanor !!!!