I would use FREE for both situations. I also often jump up in the air when I say FREE, mostly when we're outside. Saying FREE very rarely means I give a treat. For us, as we were taught, it means the dog can now go do whatever the heck he feels like doing. If that's go eat his supper that's his choice. Maybe he will jump up on me for a hug, or go running off down the trail or head to the sniff I made him pass up.
For boundary games, I use "OK" and it's almost never a release to get a treat, it's simply a release from the boundary. That's by design - I really want to build value in the release cue, so I make it high energy and make sure they respond to the cue before the reward for the release is given. This, in turn, builds value in the boundary. If I am using treats as proofing, I generally pick them up and feed them on the boundary if I want them to have it. Of course, there are exceptions, but that's how I plan to do it most of the time. I have plans to bring the same level of energy into the release for regular sit/stays, too, in order to strengthen the sit/stay whilst keeping it fun. I don't cue a sit for when I give their food - I expect them to wait until released to take it. With W&S I just use "take it", which is a fluffy nothing phrase, but with Luna, I'm working on her marking the bowl, so I give her the "mark" cue as I place the bowl and I want her to keep watching it until I give her the "get it" cue.
Thank you! Sorry just a few more questions...do you mean that you rarely release them to a treat already on the floor, but that you give the release then once they have left the boundary you reward from your hand? Why exactly does that build more value in the release cue than releasing them to a treat on the floor? In both situations, they release themselves, on cue, then get rewarded, so I am not sure why one would create more value than the other?
I use "go place" to send Inky to a boundary place where I expect him to stay until released. Sometimes I feed him on the boundary but generally I expect him to stay there and make it harder by rolling balls in front or food or jump about. I use break when he can get off, he can then get the food on the floor if there is some or else I treat him as I want him to come to me from the boundary. I'm not sure why I do this, I think I read it somewhere. Over Christmas with more visitors to the house, I'm going to really work on keeping him on the boundary when people come in. I reckon break and then letting him go to greet will be reward enough! Otherwise I use free to let him move from a sit or a lie or get his dinner, although thinking about it, he's quivering by then in anticipation and sometimes I say get it, I'm not sure he cares what I say, maybe it's the time of my voice!
Sorry, I wasn’t very clear. I mean that when I throw food around, I normally then pick that up and feed it to them while they remain on the boundary; a bit of Premack with the low probability behaviour of staying on the boundary bring reinforced by the high probability behaviour of eating. The dog not breaking the boundary to get the food means they get to eat the food. This reinforces staying on the boundary. But, I also want the dog to come off the boundary eagerly when I release them. Since I’ve built a lot of value in staying on the boundary, they can be a bit hesitant in coming off. So a “release to do what you want” would actually have them staying right there. That’s why I need to build value in the release (or cue to get off the boundary) so that that’s really rewarding, too. Then, you’re in a situation where the dog loves staying on the boundary and also loves the release cue. Simply breaking the boundary isn’t rewarding at all because they lose the opportunity to earn the rewards for staying on the boundary AND they miss out on the reward for being cued to leave it. Because there is so much value in the release, they are more likely to stay on the boundary until they are released. As ever, something that is well reinforced becomes rewarding in its own right in time. To go back to the timing, I want the dog to explode from the bed when I give the release cue. If I’m just releasing them to food, that food is a bribe to leave the bed (remembering that the bed has a lot of value). So it’s important to me that the dog comes off the bed before the reinforcer is presented, whether that is food or a toy. Again, Premack has been playing a real part in my training of this, with me running around them like a lunatic, sprinting past them, towards them, away from them while they stay in the boundary, then releasing with “OK” and only then presenting a tugger or throwing a ball or a chunk of food. I hope that made some sense!
Yes! Thanks that makes a lot of sense and I guess you don’t bother then marking the release too, you just chuck the reward after they have released, right?
I mark when they’re learning the behaviour, but you don’t need to keep marking once the dog understands what it means. A mark is communicating what bit you’re rewarding. Once the dog makes the association between cue and behaviour, marking isn’t really necessary any more. I still do sometimes, but only with a verbal marker as a “yay!” sort of response
Ok i think I’m almost there, sorry for all these questions! I think the final key question for me is this: (And now I realise that this thread should have been called ‘use of mark’ rather than ‘release cues’). So I reward a behaviour in different ways. I either 1) mark and treat from my hand, 2) I throw a treat for them to chase (normally after a check-in on a walk) or 3) I allow them to go to a treat on the floor/food bowl from a held position. I use my mark word in situation 1), but then use different words in 2) and 3). My question is, does it make sense to use the mark word (that marks the desired behaviour, e.g. a check-in, sitting nicely for dinner, or whatever, and then releases them) for all of them? Or is it necessary to differentiate between ‘reward by me’ and ‘reward away from me’? It seems like others do have different words for these situations but perhaps we don’t need to? It would be much simpler if the mark could be used regardless of mode of reward presentation? This is assuming I do not actually want to reward the release, but the behaviour before the release. In the limited number of situations I want to reward a release cue (only in boundary games?) I would use a separate OK cue, which is a ‘release but keep engaged’ cue. And finally, a pure release cue that is ‘ok we are done, go do your own thing’. Does that sound ok?
We only have one: "Vapaa" ("Free"). This releases him from whatever we were doing beforehand. He usually takes this as an opportunity to fold his ears down and sprint off like a lunatic.
I think it's clearer to the dog to have distinct markers which determine how she behaves in each situation. Whether that's "stay there and the reward will come to you" or "break your position and collect your reward". This is discussed in one of the early episodes of the "Drinking From the Toilet" podcast. You reward the release if the release is not inherently rewarding. So, within boundary games, because the value of staying on the boundary is high, you need to reward the dog's release in some manner, because otherwise it can be punishing - the dog is losing the opportunity to earn the rewards she was getting by staying on the boundary. This could absolutely be the case in a simple sit/stay, too, if you have build enough value in the sit that she doesn't want to break it. So, it depends on the dog and the context at the time as to whether I need to reward a release. But that's just simple proofing against distraction (the distraction in this case being the desire to stay in the previously cued behaviour). It sounds like you're on the right track. It's nice to be clear about it in your head if you're a thinker, but as long as your dog is doing what you expect in that situation, I wouldn't get too bogged down in it all. We're giving so many other unintentional cues with our body language that the dog will probably read that more to determine whether to work through the mark or not, rather than really listening to what comes out of your mouth
Haha when I am not dog-wrangling I am an academic so over-thinking is my job and it spills over into the rest of my life ok thanks for explaining about rewarding the release, that makes perfect sense. I certainly have different words for "stay there and the reward will come to you" or "break your position and collect your reward". Was just wondering whether you need different words for the second one depending HOW the reward is collected (eg from my hand, thrown, already on floor). But I am guessing it’s ok to have the same word for all of this (my normal mark)?
Hope you don't mind me asking a question on this thread too. Reading Snowbunny's thread about boundaries, could I ask you a question! Am I right that you put treats on the floor to help proof the boundary and then feed those same treats on the boundary to reinforce the value of the boundary? When you release, for me the word is break, you then reinforce the value of the release by tug or a ball or sometimes food. Is there a reason why you don't use food all the time? When I've put him on the boundary I always use the release cue, but reading above is this right? Also if he goes there on his own I don't tend to either reinforce the boundary or use release cue, is this right? Finally I get inky to sit or lie many times through the day, sometimes as a default behaviour instead of say jumping up. Should I use my release cue every time I ask him to sit even if it's for a few seconds or a default behavour?
I proof with all sorts of things; food is a good place to start because it’s convenient, but I’ll also proof against tuggers, balls and me sprinting around like a hooligan. Releasing to join in the party when they’ve stayed steady long enough, or else bringing the party to them. With food, I pretty almost always feed to the boundary by picking it up. An exception is breakfast time when I scatter feed in the field; I scatter then release them from their boundaries to get it. I don’t use food all the tine because food isn’t the biggest distraction or motivator for my dogs. In the early stages I reinforce the boundary whether I’ve cued it or not, but I don’t release unless I want to get the dog off the bed! I always release (or give another cue) from a cued behaviour because otherwise you’re not being clear about whether it’s ok to break or not. I don’t release from an uncued behaviour unless I need the dog to move and they’re on a boundary. I hope that helps!
Thank you. I've noticed he's slipped on staying on the place so I'm thinking I need to reinforce that again and/or proof more. He's enthusiastic on the release so could I start fading the rewards ?
I would work on reinforcing the stay and reinforce the release. That will make the dog want to stay and not break until released. Fading treats is a concept I believe most people get wrong. You want to end up transferring from inconvenient (to you) rewards to convenient (to you) rewards, which generally means transferring to real-world rewards like releasing to sniff, to play, to work etc. In actual fact, these tend to hold more value for the dog, they are just less easy to utilise in the early stages of training a behaviour. There isn’t any sense in fading to nothing because you will lose the value of the behaviour - would you work for free? No? Then why should your dog?
A reinforcer delivered after a marker reinforcers everything (at least) from the behaviour you mark to the point the dog finishes with the reinforcer. So there is a very good reason to use your marker as a release and that's when you want the dog to change position at the end of an exercise and you want that reinforced - the best example of that is when you want a behaviour to end with a release cue. You swap your mark for a release cue. Steadiness before asking the dog to run out in a straight line to something he wants....say a retrieve.... is a good example of this. Sometimes there are very good reasons why you shouldn't use your marker as a release and that's when you do not want your dog to move and you want THAT reinforced. Heelwork is a good example of this - you rarely want the dog to move when you mark doing heel work. You get to decide when your marker is a release. So, if you want to mark the dog sitting still for its dinner AND you don't want the dog to move when you mark, you would take the dinner to the dog and if the dog moved, you would back up a bit until the dog stayed still. If you wanted to use the click as a release for the dog to get its dinner, you mark and then (if necessary) encourage the dog towards the bowl to take the reinforcer. In no time at all, the dog will stay still or move on the click/marker just depending on how you have delivered the reinforcer.
Thanks! Yes, I don’t use my normal ‘yes’ mark (which has been trained to release) if I want her to stay in the same position - when I am heeling or doing a long stay and want to give her an interim treat I use ‘thaaaat’s it’ and she knows that is not a release but she still associates it with a reward. I’m happy to keep that as a separate word. What I actually wasn’t sure about was the use of my ‘yes’ mark (which does release her) - normally after that mark, I give her a treat from my hand (she has released from her position to take it, which is fine with me). I just wasn’t sure if I needed a different word if I wanted to throw her treat after mark/release, or release her to a treat already on the ground, for example. I wasn’t sure if one mark word should be used regardless of how the treat was delivered - from my hand vs. away from me (assuming it is releasing the dog for all situations).
I wouldn't use a different word, because I think it would dilute my mark, and also my mark is usually my clicker, which is much more effective than my marker words (unsurprisingly). I suppose you could do that, it seems overly complicated to me though and, actually, I think it feels a tiny bit lazy because if you were consistent with your reinforcers you would have no need to do this. The same marker can be a release or a signal to hold position depending on the exercise. You just have to be consistent with how you deliver your reinforcer, that's all there is to it. You can have a click mean release from a sit, and a click mean hold position on heel - the only difference is how you deliver the reinforcer. If you don't want to train this difference in the same marker then you could use two different markers but you still have to train what both mean - so what's the point? Perhaps it makes it easy for the human to remember, I suppose. So if you use 'good' for heel, but 'nice' for sit, you still have to train the dog not to move on 'good' - if you don't do this, the dog will swing towards the reinforcer, even if you don't really notice it (you might not care though).
