While writing my last post about value building in Crate Games I was reminded of a post I answered about using a food toy such as the Tug It toy or other container with food in it. One reader on a forum asked;
Why not just throw the toy on the reward line and then treat from your pocket when she comes back with it? It sure is easier than dealing with a treat bag?
Here was my response;
It is all about value . . . what is and isn’t valuable to an animal and how to get what I refer to as the “transfer of value” from the reward to the work.
How many people readying this, if you were walking to their car with your arms full holding grocery bags and saw a single penny on the ground would go to the effort of bending down to see if it was really only a penny or possibly something more? I would guess less than 1% of us would bother checking out the penny (perhaps the odd person that had superstitious reason for doing it), but from a “relative value to you” point of view, it is very likely we all would just keep walking with your arms full of groceries.
Now what if your arms were still full holding groceries and you saw a roll of $100 bills lying in the street? Would you then go to the effort of adjusting your load so you could pick that up? I would think the vast majority of us would do so, possibly 99.9%? Even if you were a billionaire and the roll of hundreds were “chump change” to you, you would still likely pick it up. Why? Because even if you are wealthy money would carry great value.
So why go to the effort to reward from the throw toy? Currently your dog has tons of value for food and the food comes from you so when you are training the dog holding food you have tons of value– you are the roll of hundreds.
If you throw a toy (which the dog has no value for — it is a penny) the dog may investigate the roll the first time you throw it but quickly recognizing nothing of value comes from it (ie the food only comes from you) the dog would quickly stop noticing you throwing the valueless item and just stare at you to feed them.
Training becomes more than a little difficult in this scenario.
Today I am grateful to be on my way back home — gearing up for antics of our upcoming puppy camp!
Getting a dog to interact with a toy.
This is how I managed to get my dog hooked on toys.I hope it may help others.
I teach and play the “two food game” with the dog.it was based on the ” two hose game”which is explained in the book ” Schutzhund obedience Training in Drive” by Gottfried Dildei and Sheila booth.
I just slightly altered the “two food game”by replacing food in one hand with a toy and retaining some food treats in the other.
I then throw the toy, the dog went to investigate I “mark” this then run and throw a treat in the opposite direction.
The dog runs to get the treat.Immediately I run pick up the toy and throw it again.Bit by bit as the dog caught on to the game became more excited and motivated.Soon it was retrieving the toy running back to me( at top speed) dropping the toy at my feet then tearing off in the opposite direction to get the thrown treat.
The value had transfered onto the toy…to the extent that the dog went nuts whenever I produced it.So it was not just motivating the dog but also served as reinforcement and reward too.( thats how I got my dog to tug as well….she didn’t want to let go)
I can now get any toy I want to get the dog to value by playing this game.
I then just phase out the food and just throw the toy.( the food is replaced by two toys)
You can get really inovative and come up with all sorts of variations of this game.
I still use food and toys and just be a little unpredictable… the dog, she loves it.
I now have incorperated this into training along with crate games.
Just another tool in the dog tool box.
So if you have a dog that is not real toy crazy like mine was, try this and just be consistant and have a bit of patience.
Thank you so much for sharing how you transferred value to the toy. I have been struggling with this for a long time.
I think what Susan is saying is that to transfer the value of the food to the toy you have to feed the dog the food from the toy. If you feed the dog the food from your pocket you won’t transfer any value to the toy. As the toy becomes more valuable to the dog it can then be used as an effective reinforcer for the weaving behavior but until the dog sees value in the toy it won’t do much good to throw it.
What a sensational blog.Really enjoying the read.Great educational experience.
Makes you wonder where would we all be with out some “value”, influencing our lives one way or the other and nicely summed up by Lee Carr 🙂
Nothing of value comes easily 🙂
thanks Dawn and Clyde for your responses. All help is appreciated. I’m now a couple of steps closer to ‘getting it’. Lots of repetition and increasing criteria needed for her rewards – that’s what I’m hearing.
Leslie,
I think it goes back to learning theory and classical conditioning. The dogs salivated at the ring of a bell even though no food was present (Pavlov). She should be enthused even though there is no roll of hundreds because the value of the roll of hundreds has transfered to weaving.
Unfortunately, I’m one of those dog trainers who find it difficult to “get it”. I’ve been reading everything on transfer of value to the obstacle, but still don’t really understand how to do that. If my dog loves a certain thing or activity and I make her do an agility obstacle before getting her reward, why will she suddenly think – awe, I really love doing weaves – won’t she continue to just do it because she knows she gets the reward after she’s done it. If I don’t have the “roll of hundreds” on the reward line after the weaves, she is not too enthused about working. She obviously is having trouble ‘getting it’ as well. No one ever said this was easy but I remain optimistic.
Hi Leslie,
I understand the confusion. One point to remember is to continue to increase criteria. Also, when are you placing the reward? The reward should not be offered until the behaviour is done. After your dog is offering weaves at the level you want, you can start to add other obstacles so the predictability of doing the reward is unknown. Same as playing the slot machine. Another thing is transfer of value comes with repititions and not first time association. Hope this helps.
I know I sound like a broken record, but DO enter the recallers e-course, you’ll get a ton of idea on how to shape things using tranfer of value concept.
Yes, mechanically your placement of reward is important: thrown (when and where) or in your hand. Of course there has to be real passion for the rewards you are using. REAL passion, not just so so passion. Then the transfer does happen. For some dogs it take a lot more short and sweet sessions, but you should see progress.
Take my young sheltie for example. We’ve had a lot of snow this winter, so I only had a small are in my front yard with no snow. I was stuck with one jump. I free shaped it, purely him offering it and me tossing a toy. Adding angles, different positions. I started noticing he would load himself staring at a jump – now that’s waht I call real joy and passion. He was loading himself just by staring at a jump, the same way he stares at my older dog when he starts chasing her. I was able to start doing rear crosses, something I didn’t really understand before – but how could I, when there was no focus and passion for the jump? No it’s like he is being sucked into the jumps. I can test the value for the jump by trying putting him in a sit or stand before a jump: he does break his position sometimes, but it’s good! It shows he does want nto do a jump, whereas before I could not do even the smallest send. So now we are proffing sits and stands in front of a silly little jump (it’s a good type of naughty though).
Katarina
Susan,
I hope you know how very much I and, I am sure, many others find this blog 🙂
Thank you,
Barrie
P.S. I know certain friends of mine have sworn to stop speaking to me if I say “you just need to attach more value to X so the dog/cat/boyfriend/husband/child will stop wanting to do Y” again!
Forgive me for this stupid question, but I’ve never had a dog like this before. What if you have a dog who will take treats but not eat them. He hold them in his mouth and spits them out when there is more action. He is truly working for the tug, and the retrieve that precedes it. I really can only get him to eat a treat in obedience/clicker/control work. At the agility yard, no way. (okay yes, this IS a Border Collie, how did you guess.) 🙂
Answer 1: So just use toys, no big deal.
Answer 2: Does your dog get over the top excited around agility equipment? Then he’s spitting the food out because he is too excited, just like you probably have problem eating food right before an important job interview. Work on relaxation exercises and he will be able to take food again (and probably think more clearly on the course).
1) Nope, we actually do want to work with food sometimes, even in situations dog finds high drive. For some behaviors we want to train with both food and tuggy (go between them).
2) We still want the dog to be fairly excited during training and still take food. So a question of transfer of value: this dog is crazy for toys, but doesn’t allways take food. What to do?
Same issue with a dog that only takes food, but is underaroused to play with toys.
Tarcika, I prefer the answer 2 as well.
But: not all dogs have problems switching between food and toys and the question states the dog DOES take food, but he spits it out. That’s what dogs do when they’re too aroused (scared, excited) to eat. I’m not saying one needs to kill all the excitement – just teach relaxation on cue so that it can be brought down a notch.
Teaching the dog to relax on cue won’t make him loose motivation for the game, just as teaching Leave It won’t decrease dog’s motivation for the food.
I don’t think a relax cue is endorsed here:) Self-control yes, but relax not really. Neither is leave it as a cue:)
I repeat my question about transfer of value: dog is crazy for toys, but can’t take food, what do you do to balance these out, so that dog WILL take food in ANY environment and in any stage of excitement?
Oh, I see… we’resparring with words here. Not doing that, thanks.
No no, it’s not wording, these are real training concepts, very pivotal and they were really core lessons in recallers e-course (explained in depth and covered in reals lessons). It was completely new to me, but it really made a difference for my motion crazy dog.
Andreja and Tarcika,
I think I may have found solution #3. Put dog on diet. Don’t feed before class. Bring pepperoni. He still spit it out once, but when I know he is excited I just “treat” with praise and tug.
It is good to have both reward tools available for different situations, and that’s why I’m trying to work with it.
Honestly though, I have had Labradors for 25 years and I have never ever had a dog that wouldn’t eat. When Labs don’t eat, there is something badly wrong.
Mikey,
I know this is months late, but my BC also could care less for a ‘treat’ when retrieving… it’s all about retrieve and tugging with her. She loves the treats when we do our clicker training sessions though….
Yes, BCs are all about work, it truly is their play.
I know activities the dog loves can be used to reinforce behaviour but if the dog always needs to be given a cookie for retrieving is retrieving actually that valuable a behaviour to the dog? I mean if I used an activity that was really valuable to my dog ie go for a swim I wouldn’t need to also give them a cookie to be able to keep using it as a reinforcment.
For weaves we want the value to be associated with the poles and want the reward to be on the reward line and close to the poles. If retrieving is the reward it seems to me that the transfer of value to the weaves is not going to be as clear and efficient as it could be and why would we choose to do that?
Of course it seems to me that when we throw the toy the dog is retrieving it to us too but I suppose since they are then tugging with the toy the toy keeps a ton of value because it is the primary reinforcement???
I understand the need for clarification and I believe Andrea did a very good job with her explanation. I’d like to add on, based on what I learned about transfer of value, and specific to the example of weave training.
First of all, when I finally understood the concept and power of ‘transfer of value,’ Susan literally blew my mind! Imagine the countless ways you could create value for something using this concept. Transfer of value might initially have to be set up when you first start working with a young dog, but after a while, it is happening fluidly all around you and your dog.
For example, my 2yr old dog, Tricky, had low food drive as a puppy. I mean what dog doesn’t lose her mind over cheese? Tricky didn’t. But she was a nut about tennis balls and tug toys. I literally used to feed her cheese and then offer her the tug toy. It felt as though I was rewarding her with the toy for eating the cheese. Very quickly, I was able to ask for a simple behavior (such has a hand touch), reward with cheese and then offer the tug toy. Now we have fluidity where we have a large hierarchy of reinforcements. Everything valuable to Tricky is transferring to one another when we train–that’s what I mean by fluidity.
Now, back to weave pole training. The ultimate goal is to create strong weave understanding–which means enteries, driving, low head, striding, and independence. There must be passion for the poles for all of these criteria to be met. Passion = value. So if my goal is to build value for the poles, I want to be able to immediately reward the dog for whatever challenge I present in weave training. And placement of that reward helps to build value for the behavior I am training.
For example, if I am working on independence on 12 poles, I will reward at the end of the 12 poles regardless of my position, because it’s about the independent poles–not me. If my dog has to retrieve the reward to bring it back to me (such as a treat pouch), yes my dog is being rewarded, but the timing of the actual primary reinforcement is delayed. While building passion, I want my dog to receive the primary reinforcement immediately.
Same goes for a tug toy. Toys often have value in and of itself. We build more value into it by tugging with our dogs. The primary reinforcement is actually diving on the toy and playing with it. In fact, Tricky would love to take her toy on a victory lap if I didn’t train her to retrieve it instead.
You might consider tossing a bowl or lid with something yummy smeared into it so that your dog can be immediately rewarded: this will meet the timing and placement aspects of training weaves and will send a more powerful message of value to your dog.
Hope this helps!
Because with a dog who is not into toys at all there won’t be any tugging after the retrieve.
But this is an interesting question you asked: will there be any value transfer for the weaves? Yeah, there might not be…
Hmmm… I can understand how the trainer with treats is more valuable than the thrown toy (even a toy full of food), but isn’t it possible to transfer the value to the fetch first and then use it in training?
So if one were to take additional steps to first make fetching really fun for the dog (still with the food reward at the end), wouldn’t it be possible to use this sequence: dog offers behavior – a toy is thrown – dog retrieves the toy – gets food as a reward?
The only real problem I see is that if the trainer clicks after the behavior, the dog would most probably come back for reward instead of fetching the toy first. Unless the trainer would be using the freestyle-kind of clicking, where single click is a keep going signal and double click marks the end of behavior chain…
What do you think?
Seems to me that the dog would be getting rewarded for retrieving, not weaving and I do remember that using a clicker for weaving is not recommended. Are you clicking the entry, staying in the poles, the dog posture, the foot patterning, etc?
Yes, the dog would ultimately get rewarded for retrieving, but since retrieving would carry some positive feelings on its own (assuming that transfer of value occurred), the retrieve would be the small reward, and food would be the big reward.
Same principle as backwards chaining, except in backwards chaining you teach each behavior of the chain well before combining them. And this is where the thing might fall apart… the first behavior (weaving or whatever) is in this case not taught separately, without using retrieve.
Because of backwards chaining it seems that it should work, but I really don’t know, that’s why I’m asking. It works with my dog, but he likes toys, so he’s not a good test candidate here (though I did have to work hard to make him like retrieving and he still needs some kind of reward for that with most toys).
Hi Andreja,
the problem with clicking anything concerning toys or feeding after interaction with toy is that the toy becomes work and not joy in itself. In training we want passion and drive and toys usually shift dogs to a different state of mind that just food. And we want that high state of arousal, desire, so value transfers to the thing we are training.
Also, the reward to bring the toy back or even incentive to bring a thrown toy is coming back to us and tugging (not feeding). So this is the real reward, tugging, playing afterward, not the retrieve itself. The throw of a toy is just an extension of our arms.
My sheltie boy was very very hard to get tugging and retrieve was nonexistent. Things improved dramatically after doing the recallers e-course here and I really got to understand whole different levels of dog motivation, I cannot recommend this course enough. Tugging was paired with amazing drive-developing games and his toy drive rose exponentially. Games such as cookie in the corner and all reinforcement zone chases were groundbreaking for him (Or to me really). Being a sheltie, his highest drive was with running and chasing me and when I paired toy with chasing me, things clicked for him. So value of high-octane chasing transferred to toy. Had I used just food rewards to reward toy interaction, it would be slow, mechanical and not rewarding – it would be work to him.
You most definitelly could teach the retrieve using food rewards afterwards, but it would be just a taught behavior, nothing else. What Susan wanted us to achieve was some squeeling-crazy passion when playing, not just doing behaviors, so that this state of mind will later on be transferred to behaviors we are training. It’s not about whether you get something done, it’s about passion and desire in the process.
Now my sheltie passionately works through hardest challenges in 2×2 weaves training, no food, just thrown toy and tugging. During tugging I tell him how clever he is and what a passionate boy he is, while he relieves some of the stress of working. He still isn’t really into retrieving just for the sake of it, but he is crazy for it when we are training things, because the association of speed and toy makes it work for him.
Also, tugging should be used as a stress-break (or recharging break if you will). During training with food rewards (like tricks), I think the rule is tugging after 3 to 5 treats? Maybe it seems a lot of tugging, but boy, it brings training into a whole new dimension of passion.
I am from Slovenia too and I know how clickers are (over)used in aspects of agility training and from what I observe, toys are not being used to their potential or used properly (not to mention too much value for doing too many stationary tricks next to the traininer). Since beginning of the e-course I put my clicker away and haven’t used it since. I DO freeshape everything for agility, but no need for clicker and mostly just with a toy. It’s the placement of the toy that really does it.
Anyway, I am grateful Susan put out her recallers e-course, we have definitely gone to a new dimension of training (not to mention fun and pure joy at doing so).
Katarina
Ha ha, it’s really funny discussing this in English with someone from Slovenia 🙂
Thank you for recommending the recallers e-course. I was itching to try it, but realized that I already had too much to work on at that time, so I skipped. I hope Susan will repeat it. I think a dog can never be too motivated to work with me 🙂
I am talking about a hypothetical dog in my question above (my dog’s favorite game these days involves retrieving and no food, so he’s not a good subject to study), the kind of dog for whom Susan’s article was written, who doesn’t care for toys and hasn’t been trained to love tugging (otherwise, why would you go get the food-stuffed toy and give him a treat from it).
So, the dog doesn’t find playing/tugging rewarding, and Susan suggests that the trainer should go to the thrown food-stuffed toy every time and treat from it instead of treating when the dog returns with the toy. Well since we know that back chaining works I’m wondering couldn’t we make this work as well by transferring value to the retrieve? I would still treat from the toy of course, not from my pocket, to emphasize where good things come from.
BTW, are you active on pesjanar.si as Taar?
Yep, taar in Slovene forum. I also appreciate the irony of us discussing dog training here:) Your greyhound looks fabulous in training, so focused!
I think recallers e-course might be repeated in spring (sign up for the newsletter to get announcements). We also covered exactly HOW to use food to make it extra exciting, for instance in beginning stages of some games.
Susan did write on 2×2 weaves and food motivated dogs here:
http://dogagilityblog.wpengine.com/2009/01/2×2-with-a-food-only-motivated-dog/
I see what you mean now. Yes, dog could be trained some degree of retrieving by rewarding with food after the retrieve, sticking to clear criteria and making a task of it. The value for the toy however would be very low in my mind, purely a job to be done. Then to use this ‘job’ as a reward to shape 2×2 weaves or any agility related training? That would be the real problem. Dogs will try very hard for very long time, even going through some failures and distractions if our rewards are sky high. 2×2 is a very complex behavior and at every stage dog goes through a few failures. He is just learning, but if rewards are low, he will just give up easily. Also, he will still have to be rewarded with food after each retrieve probably. He will still get rewarded for the retrieve and he won’t get any reward for the 2x2s (in his mind). Even initiating 2×2 will become nuisance to him.
I think it would be okay if the dog liked running for tug-it, (food pouch) then you open it up and give food. Hopefully the trainer would invest time in playing interactive games with tug-it, most dogs do love to sink their teeth in raw food stuffed inside and their pray drive and possibly tugging will kick in (and maybe trainer can move on to other tug toys). So we want the dog to grab the thrown tug-it, maybe even start bringing it back to us and THEN we feed. Before ever trying to use tug-it in 2x2s, I would want to see great passion for it.
In Susan Salo puppy video, she covers this for dogs that are food motivated only – they must somehow interact with the toy or food pouch (not just go in general direction of it and then look at you), otherwise training won’t work. Dog will still try to watch you and he won’t have value to drive along a line of stride regulators.
In recent agility seminars I attended I observed how handlers used toys to reward sequences. It was after e-course (so I was already cursed) and I just couldn’t stop noticing how many typical mistakes handlers were doing. Some had great toy drive, but were being rewarded way too late or in inappropriate position, not giving the dog information of what handler wanted. For instance, in a curve where a handler should toss the toy so that dog would keep on the curve or in his RZ but still on the curve, they got the toy out long after they ended the sequence and tugged in front of them (dog was rewarded for coming to their front, nothing else).
Others had poor toy drive that started to collapse at first signs of failure in agility sequences (when ideally toys should be used as stress-breaks). They would just spit out a toy and what did handler do? Go into his pocket and give food. After that toy had to go away (because dog got reward for spitting out the toy) and dog was lured with food (being rewarded to stress). Other dogs that had pretty nice toy drive, handler stopped a sequence and tossed the toy in a very random direction, dog went after the toy, came back with it and got food from the pocket for the retrieve. After being rewarded like this a couple of times, dog started to run only half way to the tossed toy and went back to the pocket. Then less and less and in the end the tossed toy was just redundant. Total disrespect of the value!
Katarina
I think you’re missing the point. By reinforcing tugging or retrieving you are building value for tugging/relieving so you won’t *need* to reinforce that behavior with food… since the value is now in the tugging/retrieving. Not to say you never reinforce it, but you don’t have to do it much once you’ve made that behavior important to them.
I have a BC who from day 1 didn’t like toys but following Susan’s program I built so much value in toys and retrieving that I have to build value back into my release the toy behavior 🙂
For some dogs it takes longer but they get it. I dogsit for extra cash and I teach my dogsitting dogs Crate Games and nose touches and toy value.. the #1 goal is to have the best pet you possibly can.