Puppy Forum and Dog Forums banner

1 - 20 of 44 Posts

·
Registered
Joined
·
27 Posts
Discussion Starter #1
I've asked this in other forums and have always been interested in what others have done to get their dogs to execute without having a bribe in front of their dog's face. It took me a long time to fade any tangible reward other than praise to get my dog to comply to her commands. If I give my dog a score of 10 when there is a visible bribe, I'd say she is at best an 8 or so when there is no bribe when it comes to her overall performance. Hiding a treat or toy in your pocket doesn't count IMO since the dog knows you have it on you.

Thanks
 

·
Registered
Joined
·
3,334 Posts
I guess it really boils down to how much importance is put on basic "compliance". Personally I like to consider a variety of factors such as speed, enthusiasm and confidence too, among others.

Generally though, if you've experienced a degradation from a 10 to an 8, you're doing something fundamentally wrong. With correct application of training mechanics and proper utilization of reinforcement schedules, an 11 or even a 12 or 13 should be achievable.
 

·
Registered
Joined
·
2,834 Posts
I guess I don't expect my dog to work completely without a reward. For some things, the reward is just delayed. A chain of commands earns some treats. It was worked up to gradually, with the dog understanding each command before upping the criteria. Agility is a delayed reward at the end of the run, but I think simply getting to run is also very rewarding for my dog based on his enthusiasm. He can do long obedience courses without a reward, but he knows something is coming, of course.

But yeah, I guess I don't want to work without compensation, so I don't expect my dog to WANT to, either. Rewards become less frequent and intermittent in some cases, of course, but I don't think I would ever eliminate them completely.
 

·
Registered
Joined
·
1,728 Posts
You have to work up to it. You can't go straight from luring every single behavior to a long, complex behavior chain like an AKC Novice heeling pattern and expect the dog to perform with any amount of precision. You have a LOT of steps in between, teaching the dog that while their reward might be a while in coming, it WILL come eventually. I know of several classes at FDSA that are specific to fading reinforcers, including Bridging the Gap https://fenzidogsportsacademy.com/index.php/courses/203 Unchain Your Performance https://fenzidogsportsacademy.com/index.php/courses/11127 Bye Bye Cookie: Helloo Delayed Reinforcement https://fenzidogsportsacademy.com/index.php/courses/13864 and Cookie Jar Games https://fenzidogsportsacademy.com/index.php/courses/4839
 

·
Registered
Joined
·
27 Posts
Discussion Starter #6
I guess it really boils down to how much importance is put on basic "compliance". Personally I like to consider a variety of factors such as speed, enthusiasm and confidence too, among others.

Generally though, if you've experienced a degradation from a 10 to an 8, you're doing something fundamentally wrong. With correct application of training mechanics and proper utilization of reinforcement schedules, an 11 or even a 12 or 13 should be achievable.
Yes, a slight reduction in " speed, enthusiasm and confidence" along with accuracy are what lowers the score from a 10 to an 8. I will train at times while she carries the tug/ball so she already has the "reward". I thought this might be the best test to see if the dog will work for the least amount of potential tangible reward since she already has it in her possession. She does well but has moments where the "fire" is not quite there.

I'm not certain what you mean by achieving an 11 -13 as her "10" with the promise of a material reward is as good as I expect. I understand that everything can improve but I am looking at this in a relative fashion.

If you wouldn't mind, maybe you could expand on " proper utilization of reinforcement schedules ".

Thanks
 

·
Registered
Joined
·
27 Posts
Discussion Starter #7
I guess my question is, do you work for free? Then why should your dog?
My dog never works for "free". The idea that a dog being rewarded with a piece of food or a toy is the only paycheck that registers with a dog seems foreign to me but I can appreciate how many could see it your way. When praise and physical contact ( slapping her ribs or a hearty petting etc.) after success is given to her, I believe my dog knows she has done well and is ultimately rewarded but just in a different form hence the dog is never working for "free".
 

·
Registered
Joined
·
27 Posts
Discussion Starter #8
I guess I don't expect my dog to work completely without a reward. For some things, the reward is just delayed. A chain of commands earns some treats. It was worked up to gradually, with the dog understanding each command before upping the criteria. Agility is a delayed reward at the end of the run, but I think simply getting to run is also very rewarding for my dog based on his enthusiasm. He can do long obedience courses without a reward, but he knows something is coming, of course.

But yeah, I guess I don't want to work without compensation, so I don't expect my dog to WANT to, either. Rewards become less frequent and intermittent in some cases, of course, but I don't think I would ever eliminate them completely.
Agreed, especially in the light of your comment " I think simply getting to run is also very rewarding for my dog based on his enthusiasm."
 

·
Registered
Joined
·
27 Posts
Discussion Starter #9
You have to work up to it. You can't go straight from luring every single behavior to a long, complex behavior chain like an AKC Novice heeling pattern and expect the dog to perform with any amount of precision. You have a LOT of steps in between, teaching the dog that while their reward might be a while in coming, it WILL come eventually. I know of several classes at FDSA that are specific to fading reinforcers, including Bridging the Gap https://fenzidogsportsacademy.com/index.php/courses/203 Unchain Your Performance https://fenzidogsportsacademy.com/index.php/courses/11127 Bye Bye Cookie: Helloo Delayed Reinforcement https://fenzidogsportsacademy.com/index.php/courses/13864 and Cookie Jar Games https://fenzidogsportsacademy.com/index.php/courses/4839
Thank you for the info. I'll read these over for certain.
 

·
Super Moderator
Joined
·
7,351 Posts
My dog never works for "free". The idea that a dog being rewarded with a piece of food or a toy is the only paycheck that registers with a dog seems foreign to me but I can appreciate how many could see it your way. When praise and physical contact ( slapping her ribs or a hearty petting etc.) after success is given to her, I believe my dog knows she has done well and is ultimately rewarded but just in a different form hence the dog is never working for "free".
Except that YOU don’t get to decide what is reinforcing for your dog. There are dogs for whom praise always be is sufficiently rewarding, but the majority simply don’t find it enough of a reward to work for. There is nothing wrong with using food or toy rewards if that is what your dog finds most rewarding.
 

·
Registered
Joined
·
27 Posts
Discussion Starter #12
Except that YOU don’t get to decide what is reinforcing for your dog. There are dogs for whom praise always be is sufficiently rewarding, but the majority simply don’t find it enough of a reward to work for. There is nothing wrong with using food or toy rewards if that is what your dog finds most rewarding.
Oh, of course there is nothing inherently wrong with using food or toy rewards, heck I use them quite a bit, especially when training new behaviors. I generally try and build some additional food drive in my pups when they are very young. I guess I am more thinking more along the lines of competition dogs such as IPO, French ring, Mondioring, agility etc. where tangible rewards are not allowed during competition. Certainly many of the disciplines provide a reward all on their own as a previous member mentioned. I'm just curious how forum members bridged the gap above and beyond a simple fading of the material reward and still maintained 100% performance.
 

·
Registered
Joined
·
1,728 Posts
Check out Hannah Branigan's Drinking From the Toilet podcast. She recently did an episode on what makes a good reinforcer. Also, the CogDog Radio podcast recently had some episodes on reinforcers.
 

·
Registered
Joined
·
6,227 Posts
When I was doing agility one did a lot of engaging before and after the run that may or may not have included food or toys. If you look at a good agility trainer with a pup it can be difficult to see where the training happens and that play extends over time until dog is playing, running a course and playing after finishing the course. Fenzi Academy does engagement class as well.

I've continued to use that type of training for manners and my current stinker has 'walked' a mile playing the whole time. Scary how well it works at least with this extremely busy dog. I am using food in his case though.
 

·
Registered
Joined
·
7,296 Posts
It's all about primary, secondary, tertiary, etc. reinforcers.

Food, being able to exercise prey drive (chase, bite, hold, fight, disembowel), water, etc. are all primary reinforcers to varying degrees depending on the dog.

Anything that is associated with a primary reinforcer becomes a secondary reinforcer and also becomes, in itself, highly rewarding to a dog. For example, some markers like "good dog!" or the clicker can be secondary reinforcers, but of course not all marker are. Susan Garrett talks about teaching this idea, basically having a signal for the dog that reinforces them when you can't bring food or toys into the ring. Ian Dunbar talks extensively about it too. This is the guiding principle behind Denise Fenzi's engagement training. Pretty much, YOU become a secondary reinforcer. Basically, if you train thoughtfully your dog can from the start work for YOU, not blatantly for a treat or toy or whatever. But YOU are associated with (and come with all the good feelings of) those primary reinforcers. Or, a person can be associated with the anticipation of negative consequence if a lot of corrections are used.

Also, if training is done 'right' then training in general should be classically conditioned as being an amazing thing to the dog.

I don't think there ever should be NO primary reinforcers in the picture though. If a person trains well enough in training situations I think there can be plenty of times, such as in the ring, when no 'obvious' rewards are in sight. But I don't think the goal is to fade out all rewards forever any time in a dog's life. There is the maintenance stage of any behavior.
 

·
Registered
Joined
·
7,296 Posts
I guess it's also about fading rewards in a thoughtful way, which is it's own category.

You mentioned you have a correction phase in training. If that is the case, and you switch from dropping all reinforcement except praise and then add correction when you are proofing a fluent or generalized behavior, I can see why your dog's performance might drop. I don't have a correction phase in my training and I've found that by randomizing my rewards, gradually extending time between rewards, and using ALL things my dogs love, I do not see any drop in performance when I don't have a reward on my person. "Bribe" and "Reward" are two totally different things. I "bribe" when I lure a behavior in the preliminary stages of training some behaviors. The rest of everything I do is thoughtfully utilizing rewards.

My older dog performs best when we are in a field. Reason being, he knows that stellar performance earns him the release cue to sniff (which is what he really, really, really wants). Also the reason why, when I used a lot of corrections in the past and tried to force his attention, I could never quite get the focus I wanted, not even with treats (and he is crazy food motivated). I was always competing with the environment. Using the environment to my advantage changed our relationship.

My Dutchie is very black and white. He wants to tug. Training him is almost mindlessly easy. I just need to think about where the reward is (ie, not in sight until earned), and gradually asking for more before rewarding him. No issues with performance.
 

·
Registered
Joined
·
6,658 Posts
Using food in training is not the same a bribing, unless you present the food and then ask the dog to work.

To teach your dog to work without food, you have to teach them that working makes the food (or toys, play, game, etc.) appear. And, just like you can ask them to wo

Hiding a treat or toy in your pocket doesn't count IMO since the dog knows you have it on you.
So... teach your dog that you can make food (or toys or whatever) appear from ANYWHERE even if you don't obviously have food in your hand or in your pocket.

If you're asking how to teach a dog to work for you for the sheer joy of working, there are two answers: 1) genetics or 2) if your dog doesn't have the right genetic temperament you can use food and/or play to build your relationship with the dog, and teach them that when they work for you, awesome things happen.
 

·
Registered
Joined
·
3,334 Posts
So... teach your dog that you can make food (or toys or whatever) appear from ANYWHERE even if you don't obviously have food in your hand or in your pocket.
Yes. And add to this "at ANY TIME".

Some dogs are inadvertently conditioned to believe that the potential for reinforcement only occurs during specific situations or contexts, which can be a major flaw in any training program and lead to things like a 'what's in it for me' or a 'show me the money' attitude. It's best to condition the dog to think that there are rewards available at *all* times, including times when training might seem somewhat out of place.
 

·
Registered
Joined
·
27 Posts
Discussion Starter #20
My older dog performs best when we are in a field. Reason being, he knows that stellar performance earns him the release cue to sniff (which is what he really, really, really wants).
I like what you are saying and how you got to this point.

I have taken advantage of the building of anticipation coupled with a pending release. Overall, I would say this is the number one motivator and form of reward which appeals to my dog when the task at hand allows for this sequence (release) to exist. There is no food involved, no tug or toy involved, just the primal reward of the dog indulging its drive instincts (mostly prey), I have found nothing more powerful as a motivator even though my dog has a healthy food drive.

I watch many videos of dog training and all too many are just videos of handlers cramming food scraps in their dog's mouth and calling that "success". I'm a fan of Michael Ellis and a true believer in creating engagement and as the saying goes " if you don't have the dog's engagement you have nothing".

My goal is to create the accuracy and crispness across the board which a pending release and built up anticipation creates. The problem is, there are situations where there is no realistic release or opportunity to reward the dog via its innate drives except food and I obviously have an issue with relying on food rewards and calling that a "success".

Examples, I'll put my dog on a down/wait and walk 50 yards away at our training field and via hand signals have my dog go through her position changes, she's perfect, no creep and 100% focus. The moment my arm goes up she charges full speed ears down to a front position and takes her finish tight or park position per my signal. All she gets is positive attitude from me and maybe a slap on the ribs. If I choose, I'll down her anywhere along the way back to me and she hits the dirt with a deliberate action. She'll exhibit this type of accuracy with most of her tasks and disciplines especially tracking or running down an "item" and latching on. I could work her all day with disciplines which involve the task as the reward and never even think about any food or tug reward being needed because it just doesn't even come close in significance.

However, let's take her focused heel and I'll paint a different picture. Without the tug under my left armpit, my dog will certainly stay in the slot but her focus is less than 100% and her gait is less than exaggerated. When I have the tug under my armpit and do the same exercise, she is locked on to me head up and she is prancing with her front legs. I have tied the focused heel exercises to other tasks which involve a release hoping the focused heel sans any tangible reward ( tug ) will create an association which works backwards through the focused heeling exercise, so the focused heel portion becomes a secondary reinforcer but have not seen the same results as having the tug under my armpit. I'd appreciate any advice you might have to offer.

FWIW, at this stage of the game with my GSD, about the only corrections I use is just my negative verbal marker. I have come to find she is a very hard dog when it comes to any physical corrections but rather soft when it comes to verbal reprimand. Because of this characteristic, I go easy on the negative verbal admonishments and also believe this "softness" lends hand to why praise and a positive attitude on my behalf registers with her so well.
 
1 - 20 of 44 Posts
Top