So, behaviour should happen in response to a cue, and not happen without the cue....stimulus control..... Just working this through.... Some behaviours have stimulus apart from the human cue - shake is a good one. What are reasonable expectations for "no shake"? Ever unless I give a cue? And just because I've got shake on cue, that doesn't stop him shaking without it! ;D ;D ;D Er...help with that one... I do seem to be struggling with this. I put my placeboard down, Charlie jumps on it. My trainer says "get that off cue". Right. I guess I've got to stand there, treats for not jumping on the placeboard until I say place.... Gosh, I've got SUCH a lot of this to do! Any tips? And how do you fit all this training into your lives?????? I must be hopelessly inefficient....
Re: Getting behaviour "off cue" Your training sounds very complicated to me! My mind would be in a terrible muddle. My dogs shake on command but it wouldn't worry me if they shake without a command, I just need them to shake when they get out of the car or out of water, so that is when I ask them to do so.
Re: Getting behaviour "off cue" It's pretty standard, Stacia, a behaviour isn't under stimulus control if it happens without a cue (although a coat full of water is a cue to shake, I guess ;D ;D ;D). All steadiness is is fetch off cue... But maybe it's more important with clicker trained dogs - they offer behaviour all the time. If Charlie doesn't have anything better to do, he'll run through the entire set of everything that's ever been rewarded in the hope something "pays".... : And positive training involves rewarding a lot of uncured behaviours....so yes, maybe it is a bit more complicated.
Re: Getting behaviour "off cue" I think I might be a bit confused sorry. What does your trainer mean by "get that off cue". Is it because Charlie sees you putting the place board down as a cue to get on it. Could you just use your wait/stay cue, put the board down then release him with the cue 'place' so he goes to the board. That might not be what your getting at sorry. When you say all behaviour should happen in response to a cue and not happen without the cue do you mean a cue from you or a cue in general ? I think all behaviour occurs because of some sort of cue.
Re: Getting behaviour "off cue" [quote author=JulieT link=topic=9954.msg144652#msg144652 date=1424379793] It's pretty standard, Stacia, a behaviour isn't under stimulus control if it happens without a cue (although a coat full of water is a cue to shake, I guess ;D ;D ;D). All steadiness is is fetch off cue... But maybe it's more important with clicker trained dogs - they offer behaviour all the time. If Charlie doesn't have anything better to do, he'll run through the entire set of everything that's ever been rewarded in the hope something "pays".... : And positive training involves rewarding a lot of uncured behaviours....so yes, maybe it is a bit more complicated. [/quote] That is what I find confusing "all steadiness is a fetch off cue", I have unpicked it and understand what it means, but I really cannot see we need to think that way, mudding the waters for me ;D
Re: Getting behaviour "off cue" Well, maybe the forum isn't the right place for these discussions. I think we need to think like this because that's the language of dog training - or at least it is in my world. My trainer means "get that behaviour properly under stimulus control". Pippa explains much better than I can what that means here: http://totallydogtraining.com/what-is-stimulus-control/ "It means: A behaviour is said to be under stimulus control when we can accurately predict a desirable outcome to our cue. We can truly say this has happened when four conditions have been fulfilled: The dog always performs the behaviour when you give your cue The dog does not perform a different behaviour in response to that cue The dog does not perform the behaviour in the absence of the cue The dog does not perform the behaviour in response to a different cue"
Re: Getting behaviour "off cue" I would have thought the forum was the place to discuss this. Just because we approach the subject from different angles doesn't negate a good discussion. I don't think a dog will always perform a behaviour when giving a cue (though this may be desirable and the aim). Dogs do perform a different behaviour in response to a cue etc. I can see that this is what we aim for but I am afraid no dog is a robot. I do understand that it is the scientific principle of how dogs learn. Not all dogs have read the book I just feel you are approaching this from a scientific principle of learning, which is fine, but to me, dog training is much looser than that and my dog does behave relatively well and does most behaviours on cue. I cannot say that for no 1 dog who was brought up differently. Maybe I do train in a scientific way, but don't know I am doing it! SORRY I DIDN'T MEAN THIS TO BE IN BOLD! I ONLY MEAN THE WORD 'THE' AS IN ' THE' FORUM WAS THE PLACE TO DISCUSS THIS. CAN A MODERATOR CHANGE IT FOR ME PLEASE? [size=8pt]Fixed the text so it's not all bold (Rachael)[/size]
Re: Getting behaviour "off cue" I think I understand what you're getting at, Julie. Right now with Maisie, if we're in a training session, I'm working very hard to differentiate behaviours for her, as she is so smart (like Charlie) she will run through her list of behaviours in the hopes of getting it right. After all, if you're working on three things, sit, down, stand, then she has a 1/3 chance of getting it right just by doing the first thing. So even though she knows "sit" right now, and I could start enforcing that sporadically more, I tend to reward it each time she gets it right because sometimes when I say sit/hand signal (although less so with hand signal) she'll sit… but if I'm not quick enough to mark it she starts laying down from the sit to see if I really meant "down"… Of course this isn't quite what you're getting at, as you mostly want the behaviour only offered at the cue, but I do see how the trainer is framing it. I am working on the second and fourth parts of the four part list you posted - getting the right behaviour and not mixing up cues, essentially. I understand the desire to have the behaviour only offered when the cue is given… most people would probably understand this when thinking about the "shake a paw" cue. A lot of pet dogs only have a few behaviours, including sit/down/paw, etc… so if you ask them to sit for a treat, a lot of the times a dog that knows "shake a paw" will also paw at you because they know that is a behaviour you ask for. However, if you're not prepared to shake the paw, the dog might end up scratching your arm because it is swinging its paw up for a shake (especially if it needs a nail trim). I had a very well mannered, lovely golden x lab cross growing up, who would never bite/scratch/hurt a soul in any way (after puppyhood anyways), but sometimes she would accidentally draw blood by her enthusiastic paw shaking! So if you were able to fully master the shake a paw such that the behaviour is ONLY offered when you give the cue, you would decrease the likelihood of that happening (or a paw being given when it's not nice to have it be given, such as you're sitting on the couch and the dog gives you a paw just to get attention/possibly a treat - it can be annoying). Does that example resonate with anyone? (Although perhaps I'm totally off on what Julie means as well!) My response to the paw dilemma though was to not teach "shake a paw". I ask for her paws when doing nails or wiping them, but in two days after I started trying to teach "shake a paw" she started pawing me while we were doing training… I could probably have trained that out but I decided however cute shake a paw is… it isn't worth it. We learnt "bow" instead as a cute trick to show off. Many people probably would love to have a dog that fits all four checkmarks on the list but perhaps we just don't have the time to fully proof that or work on it, it's admirable to have a dog that would behave so well though and if you and your dog both enjoy doing the training and working on that, then I don't see anything the matter with trying to get the behaviours locked down like that.
Re: Getting behaviour "off cue" Getting a paw shake not to happen unless you cue it, is a good example of needing to make sure something doesn't happen unless you give the cue. I need to get a lot of stuff reliably under control but more so the behaviours that have been completely shaped and clicker trained. Charlie offers these behaviours a lot. The placeboard is a great example. Left to his own devices, he'll hurl himself at any placeboard on the training field. ;D ;D ;D Which means the behaviour is strong, but I really need it only on cue now! I do think this might be a bigger problem for positively trained dogs - whereas with dogs traditionally trained a commmand means "do this or else" and proactive offering of behaviour to earn rewards doesn't really come into it so much, I suppose.
Re: Getting behaviour "off cue" I think it is more a problem that during clicker training it wasn't specific enough. The dog was allowed to offer too many random behaviours before a click. Whereas if heel work, or 'on the box' etc was the aim, then there would be a narrower area of opportunity to express other behaviours. I do know that when first teaching via the clicker we wait for the dog to offer something we like and then click, but we should move on quickly.
Re: Getting behaviour "off cue" Interesting, and I can see how the confusion may set in especially as the two phrases 'on cue' and 'off cue' seem to mean almost the same thing (i.e. doing the behaviour only after the cue) although possibly 'on cue' doesn't prohibit the behaviour being done without the cue. Maybe. Am I close here?! Thing is, the dogs have to make independent decisions some of the time as to what the most appropriate behaviour might be, because they must be doing something at any given moment, and might not have been given any instructions yet! That said, gundogs won't go far wrong with 'sit quietly at the handler's left side' as a default behaviour ;D And I think that might be the key. I can't see your trainer suggesting that 'sit quietly at my left side' must be 'off cue' and that you need to sort it out if it happens spontaneously. So both handler and dog need to be clear about what the desired 'default behaviour' is, i.e. what special behaviours will be rewarded even if not cued!
Re: Getting behaviour "off cue" Interesting, Kath - yes, well those are the things I'm thinking about. I don't think "no behaviour", or even a constant default behaviour, is an objective. Indeed, that would be extremely frowned upon (where I train, anyway), and I think context is everything. So perhaps for me the relevant context would be between "ready" (we are working) and "go free" (we have stopped working). If a dog is told "go free" (do what you want), then that's a different context than after "ready" (working now). I don't care if Charlie does various behaviours that I have on cue after "go free" (run around for example) but I don't expect them to happen uncued while he is working. Although some behaviour I think I never want unless cued. Fetch is one that I don't want to happen even if the dog has free time, I don't want him sitting on a placeboard uncued either, I think, anyway.
Re: Getting behaviour "off cue" I think the phrase 'get that OFF cue' was confusing and ambiguous. I would have said 'make sure that behaviour only happens ON cue' Other than that I think it's crystal clear what is meant by 'put it on cue' or 'put it under stimulus control'. It means 'train that behaviour so it happens when you ask for it, and does not happen when you have not asked for it'. [quote author=Stacia link=topic=9954.msg144668#msg144668 date=1424383553] I do know that when first teaching via the clicker we wait for the dog to offer something we like and then click, but we should move on quickly. [/quote] This is free shaping, which is one technique that you can use to train a behaviour (using a clicker, any other marker, or even just food used without a marker although that is less effective). Sorry to sound like a pedant, but 'clicker training' doesn't necessarily involve free shaping. You can use a clicker all your life and never, ever use free shaping. Also, while I am riding this hobby-horse, training with a clicker (or any marker) can be really slow - it is not all fast paced, behaviour-evey-two-seconds stuff. The term 'clicker training' has acquired so much baggage, sigh. Anyway....ideas to stop Charlie leaping onto the placeboard.... Basically the problem is that being on the placeboard has become far more rewarding (more likely to lead to treats, games, dummies, whataver) than being off the placeboard. So it wil be helpful to make being off the placeboard much, much more rewarding than it currently is, and being on the placeboard less rewarding than it is. Start using more boring rewards for your placeboard training. Start using awesome rewards for not approaching the placeboard as it appears and is lowered to the ground. I'd make this a two person effort - one wrangling the placeboard and the other rewarding (food, dummies, whipit, whatever) for not going near the placeboard. Also, obviously never reward uncued sits on the placeboard (but I'm sure you already have that covered!) We had this problem when we first crate trained Obi. He was always dashing into the crate and looking delightedly at us for a treat. But that was pretty easy as we just kept the door shut when we didn't want him in it....
Re: Getting behaviour "off cue" Free shaping of what for too long Stacia? And how long would you say is enough? Sorry, I'm being cheeky. ;D ;D ;D If you mean that too many behaviours were rewarded on the way to an end outcome, and so what I am suffering from is the repetition of behaviours that were rewarded as intermediate steps - no, that's not the case, I am efficient in the degree of choice I allow in free shaping. If that were the case, what I would see is a different behaviour occurring, not the exact, perfect behaviour that also occurs in response to the cue. Charlie sits uncued on his placeboard, as well as he sits cued on his placeboard. Charlie is, in his offering of behaviour, just displaying that he is a dog that seeks to work for rewards and I have to now teach him that unless there is a cue, there is no reward for finished behaviour - then I will have the behaviour under stimulus control. He will offer new behaviour when his shaping cues are present - which is me sitting down, in a certain way, with my pot of treats in a certain position and so on.
Re: Getting behaviour "off cue" Well, he's definitely a dog who has learned to try lots of different behaviours to see what gets rewarded. That's a good thing, because it shows that he's very confident and has no fear of any negative consequences during training. It's very hard to train a dog that won't experiment a bit... I definitely agree that a lot of free shaping will encourage a dog to be like this. For that reason, free shaping is a good way to build up the confidence of a dog that's a bit lacking in that department or who has come to not enjoy training. Can a dog be TOO willing to offer behaviours? Well, it can make it hard for the handler to stay a step ahead But I don't think a dog can ever be too willing to try new things I don't think that Julie's current issue with the place board is because Charlie has had too much free shaping (I don't think you can have too much, as long as the dog is making progress, and handler skills are good enough). It's just because getting on the place board is hugely rewarding and has been well learned and heavily reinforced. Also, it's possibly the case that training with the place board always happens right after the appearance of the place board. That is, the appearance of the place board is quickly followed by exercises involving sitting on the place board. So Charlie is just cutting out the middle man and anticipating the request by sitting on it without being asked. It'd help to break the strong correlation between the appearance of the place board and being asked to sit on it. Having the place board always accessible but rarely being asked to sit on it would help, so the sight of the place board is no longer such a strong predictor of being asked to sit on it. For a time I'd probably leave it out in the house all the time, but never or rarely use it. And do the same in different locations.
Re: Getting behaviour "off cue" [quote author=Oberon link=topic=9954.msg144760#msg144760 date=1424429530] Also, it's possibly the case that training with the place board always happens right after the appearance of the place board. That is, the appearance of the place board is quickly followed by exercises involving sitting on the place board. So Charlie is just cutting out the middle man and anticipating the request by sitting on it without being asked. [/quote] Ah, yes. This is the case. What about "shake" then? So, he'll shake on cue, although a bit of water around helps. But of course he shakes all the time. Coming out of water, I'm trying to get the "hand it over then shake". But I still get shake, hand it over (if I asked him to shake again, he would). Guess I've just got to wade in a bit more...might wait for spring for that. And walk off lead at heel - that's one I struggle a bit to sort out. As he offers that behaviour all the time, and don't mind at all if he wants to trot along at heel without being cued..
Re: Getting behaviour "off cue" With the water/shake thing, I'm guessing teaching 'shake' is to try to encourage him to deliver first and shake after, not something you're really keen to have 'off cue' for its own sake. The holy grail for Charlie is of course getting him keener to hand the dummy over in the first place, and this situation is no exception. Would a switch retrieve, one into water then one behind onto land be any use do you think (ignoring the whole 'shake' issue, just providing a bigger stimulus than a soggy coat)?
Re: Getting behaviour "off cue" I wasn't really thinking of Charlie's eagerness to get on the place board but you mentioned that he offered a whole repertoire of behaviours to elicit a click at other times Maybe you should put him in a sit first and then ask him to get on the placeboard, so that he doesn't 'jump the gun'!
Re: Getting behaviour "off cue" [quote author=Merla link=topic=9954.msg144785#msg144785 date=1424439623] With the water/shake thing, I'm guessing teaching 'shake' is to try to encourage him to deliver first and shake after, not something you're really keen to have 'off cue' for its own sake. The holy grail for Charlie is of course getting him keener to hand the dummy over in the first place, and this situation is no exception. Would a switch retrieve, one into water then one behind onto land be any use do you think (ignoring the whole 'shake' issue, just providing a bigger stimulus than a soggy coat)? [/quote] Ah, these days, Charlie has no trouble handing the dummy over coming out of water - but that could be because I don't have new water dummies. : All I've got left on water retrieves is the shake thing....well, no shake is in context, I think. I think when we are doing water retrieves he shouldn't shake until told to do so. I'd be interested in what others do about that. Or is it just a narrow "hand the dummy over before you shake" thing? In which case, is having shake on cue much help anyway?
