Hi there, I'm having trouble fading rewards...even though I get the whole psychology of the gambling effect and I do use premium, basic and bonus rewards....I can't seem to stop myself giving a reward when Welly my 7m lab does something right. And here's why...when he does something well I want to reward him, when he doesn't do it well...I'm worried if I don't reward him when he finally does it he might not do it at all next time. Also as I'm currently working on proofing things as I'm struggling know how 'hard' any given situation might have been for him so I constantly reward good behaviour. For example, how does one know when proofing is ever finished? Surely there could always be a set of circumstances (smells, sounds, location, other distractions) that come together to create a unique situations all the time and given the dog can't generalise and may never faced an element or that combined situation before I keep thinking I had better reward to embed the behaviour and help him learn....hence my dilemma Anyone help with this? L
Re: Having trouble fading rewards It is actually totally find to use a high rate of rewards But do avoid giving rewards for anything below the standard you know your dog can routinely do. At or above standard - reward that every time if you're happy with where you've got to. That is fine! If you do want to gradually shape a better performance though (faster, longer, straighter, closer, further away, more distractions, whatever) then gradually shift the 'band' of acceptable (rewardable) behaviours up a gear. Just wait for a little more before you reward. e.g. I might want my dog to sit straight, looking straight ahead, by my left side when I say 'heel'. Sometimes he might sit a bit crooked, with his bum out a bit far, so I don't reward those. I ask again for a heel and if it's straight enough I reward that. I can reward every single straight one if I want to - just not the crooked ones. And, nope, proofing is rarely finished I reward heavily in new or challenging situstions. So, don't be afraid to dish out the treats - but just have the standard you want clear in your head. And remember that when something gets harder on one dimension (like, the environment is new or busier then previously) the standard must be allowed to drop on all other dimensions (so, lower the bar for speed, accuracy, distance, duration). Hope that makes sense
Re: Having trouble fading rewards What Rachael said, and in addition, I know several force-free trainers that don't do much at all in the way of fading rewards, and who claim that they still get excellent results.
Re: Having trouble fading rewards I work with a trainer who doesn't fade rewards, although we don't always use food, and we always reward to a certain standard. No reward is a hugely useful bit of information for the dog. Increasingly, we use the environment and yesterday we used access to distractions as rewards. Eg water can be a reward for Charlie.
Re: Having trouble fading rewards [quote author=LisaB link=topic=9643.msg139306#msg139306 date=1422178939] when he doesn't do it well...I'm worried if I don't reward him when he finally does it he might not do it at all next time. [/quote] If your dog makes 2 mistakes in a row, stop and change something. Eg if you've tried twice to get a certain behaviour and he's got it wrong, or hasn't got it to the standard you want, make it easier. Also, you can ask for something else that he can definitely do and reward that to keep his confidence up. My default is a hand touch, which my dog can just about always do. So if I've got myself in a place where my dog's performance is too bad to reward, I ask for a few hand touches and reward those. That's enough to keep him engaged and working normally.
