User:Lindsay Ridgeway/Reward-based Field Training for Retrievers/Problem Solving
- 1 Problem Solving
- 1.1 Barking
- 1.2 Dropping the Article
- 1.3 Not Returning with the Article
- 1.4 Slow Returns
- 1.5 Beaching
- 1.6 Cheating
As discussed under "Biting and Barking During Training" in the chapter "Principles", reinforcing a behavior while the dog is barking may reinforce the barking as well, so it's important not to do that.
Barking can also occur in other situations. Some of the problems, and suggested solutions, follow.
Barking in Crate
For dogs who bark in their crates, it's sometimes helpful to immediately cover the crate as soon as the dog begins to bark.
Once some trainers find that the dog won't bark with the cover in place, they begin to cover the crate as soon as they put the dog in.
Others prefer to wait until the moment the dog actually begins to bark. The advantage is that you can learn whether the dog actually prefers to have the crate covered:
- If over time the dog stops barking while the crate is uncovered, that indicates that he doesn't like having the crate covered and learned to stop the behavior that was causing it to happen.
- If over time the dog continues to bark when the crate is uncovered until you cover it, that indicates he found the having the crate covered pleasant and learned that barking was a way to get the cover in place.
Barking During Activities
Here's a method Lindsay Ridgeway used with Lumi and Laddie, his two Goldens, when each was a puppy and occasionally barked at other dogs in pet stores or to demand something in the kitchen. Douglas St. Clair, noting that this technique is too mild to be called a "time out", dubbed it a "time in":
- React every single time the dog barks.
- React at the very instant that the dog barks.
- If necessary, gently snap on the dog's lead. You can use a target hand or a verbal cue to bring him and keep him with you if that works.
- Walk some distance away to the most boring available location. Use a neutral or even cheerful demeanor. All that is happening is that the dog has made the choice of barking, and he is learning that his behavior predicts a particular consequence. That consequence is being taken out of the fun location. Harshness adds nothing useful to the lesson, and has the potential for diminising your relationship.
- Cue or gently press your dog into a sit. The reason why you might press your dog, rather than cue the sit, is that you're not training "sit" now. If he won't respond reflexively to a verbal or visual cue, you're distracting him from the learning the lesson at hand.
- Wait 10-30 seconds with him. You can stand or kneel beside him, you can pet him if you like. It is unnecessary to be unpleasant, but of course it should not be so much fun that the dog would rather do this than whatever he was doing before.
- After the boring wait, run back together to what you were doing.
Barking in Vehicle
Some dogs bark in the vehicle when waiting their turn for training with a group. They can use a lot of energy doing that, affecting their performance. In addition, it may annoy the other dogs and trainers, perhaps even affecting their performance.
It's possible that if you move the vehicle further from the activities, the dog will stop barking and rest, awaiting his turn. Over time, learning to rest between turns may become a habit and you won't need to continue moving the vehicle as often.
More about Barking
Perhaps it would be best to see if the following post has any useful information in it, and extract that information into a compact form for addition to the material above.
Here's a DogTrek post from Lindsay Ridgeway, responding to a question from a woman named Christina about her Golden puppy barking in his crate:
On Aug 27, 2007, at 9:01 AM, Lindsay Ridgeway wrote:
On Aug 27, 2007, at 12:14 AM, Christina wrote:
> He realized yesterday that he has a voice adn barks A LOT when he is > in his pen. I know that sometimes he is barking because he has to pee, > but often it's because he wants to be out with the rest of us. How > long should he be quiet before I let him out? I know with my other > dog, barking is a self-rewarding behavior... is ignoring him enough? > what about when I know he has to pee? Is it ok if he barks when he has > to pee? > > He is 13.5 weeks and very, very smart!
Hi, Christina. Congratulations on your Golden puppy. He sounds like a great dog.
I'm not an expert, but we have three Goldens in the house (among other creatures), including Laddie, my 4 month old. Laddie is the only one who spends any appreciable time in the crate, so I'll tell you about Laddie and crates:
- We have three crates: one in my office (I often work at home), one
in the car (unlike the other dogs, Laddie only travels in the crate), and one in the bedroom.
- Peeing has rarely been an issue. I always let Laddie out to pee
immediately before he goes into the crate and as soon as he comes out, and except when he used to sleep in the crate, he was rarely in for more than 2-3 hours at time, usually less. I've read that a two- month old can go three hours, a three-month old can go four hours, and so forth. I found with Laddie that after about the age of 9 weeks, he could go longer than that over night (six hours or more), but as I mentioned, I can't remember him ever having to go more than about three hours during the day. At about 3-1/2 months, I became confident enough in his ability to hold it that I let him take a nap with me in bed a couple of times, and when that went OK, my wife and I let Laddie start sleeping with us and the other two Goldens. He'll probably do that from now on, so he's not in his bedroom crate much any more, and doesn't bark when he is.
- If your pup needs to pee more than the guidelines I mentioned
above, you might want to check with your vet about a possible medical issue. Laddie will pee pretty much every time he gets on the grass, but he doesn't have to go any more often than those guidelines.
- For the office crate, Laddie generally barks when no one is in the
same room with him, but activities are in progress in an adjoining room. About two weeks ago, I began the practice of putting a blanket over three sides of his crate immediately at the very first bark. I believe that "immediately" and "very first" were both important. Two things happened. First, he'd stop barking within a minute or two. Second, after a few days of that happening, he stopped barking altogether at most of the times when he would have before, making it unnecessary to cover his crate for several days in row now. I read that experience to mean that the blanket took away whatever stimuli were triggering the barking, enabling him to relax, and also that Laddie gradually concluded that he'd rather have those stimuli than not have them.
- I think the bark-suppression would have worked just as well if I'd
covered his crate as soon as I put him in for situations where I expected barking. But then he wouldn't have learned how to stop me from putting on the blanket, and I wouldn't have learned whether he prefers the blanket on or off.
- For the vehicle crate, Laddie generally barks if he can see Lumi
(my big dog) and me training nearby while he awaits his turn. If I have the time, I move the vehicle further from where we're training, and Laddie quiets after a minute or two. I want him to learn to relax when it's not his turn, so that's my preferred approach, though sometimes I just have to let him keep barking. My hope is that in time, by my moving the van so that he can relax, he'll gradually develop the habit of relaxing between turns and I won't have to continue moving the vehicle.
I'd also like to mention some lifestyle issues that may or may not be relevant to the barking:
- Laddie gets a lot of time with me. Besides just spending time
together, we train at least twice a day, usually more often. When we train, it's always highly active, generally outside, and involves a good bit of running and/or swimming. My intent is that Laddie have a distinct on/off switch, with the ability to play at full speed for short bursts, and then to thoroughly relax during the rest of his life. I've tried to create a schedule for him that promotes that rhythm.
- Also on the idea of an on/off switch, Laddie and I have a daily
bodywork session. This is when I trim his nails if needed, clean his ears, go over every inch of his body massaging him and checking for ticks, give him his stretches, brush his coat, and brush his teeth. Of course, when I started this when he was seven weeks old, he was unable to relax during those activities. But he quieted gradually day after day, and now is almost as quiet as Lumi during those sessions. I can't prove that this carries over to a more general ability to relax, but maybe it does.
- Laddie often gets food in his crate with him. This includes his
daily afternoon snack of a frozen chicken wing, the occasional treat- filled Kong, and me often walking over to the crate to give him a small piece of chicken jerky, half a cashew, or some other treat through the bars.
I've often read the rule that you should not pay any attention to the dog when he barks -- no talking (gentle or harsh), no letting him out, not even looking at him. Before I discovered the blanket trick, I followed that rule and asked everyone in the house to also honor it. I don't know if it made a difference, but it's possible it did. For example, maybe the blanket trick wouldn't have worked a month ago, but if we'd reacted to Laddie when he barked at that time, the blanket trick also wouldn't work now. I just don't know about that.
Good luck with the barking and everything else about your new pup. I hope you'll keep us up to date with your life together.
Dropping the Article
Your dog will need to be able to go into a sit while holding an article. This can be surprisingly difficult for a dog, who may have a strong inclination to drop the article before sitting.
In addition, your dog will need an instant release, not when you grasp the article, but when you cue "out". The reason for not releasing when you grasp the article is that the dog can't judge whether you have a good grip and may drop the article on the ground, thinking you're ready for him to release it.
Shaping "Take It"
- Start by using your finger as the "article".
- In the first session, click and treat for the motion of grabbing your finger.
- Be sure to click the bite, not the release.
- When you can anticipate the moment of the bite, you can add a verbal cue such as "take it". You won't need this cue in the field, but it will help with training the hold.
Taking an Article
Once the dog can hold and out with your finger, train the entire behavior again from the beginning with a dummy.
Training the Hold
Here's a DogTrek post from Jody Baker describing how to train the whole process:
On Oct 13, 2007, at 2:01 PM, Jody Baker wrote:
It's sometimes hard on us people, when we start out wanting nothing more then to play at something, only to change our minds and suddenly want the dogs to stop all the stuff they've previously done and now do something not nearly as much fun for them.
Several very experienced people now recommend NOT using the clicker for duration with the retrieve. They use it just a bit at the beginning and then not at all after that. Other very experienced people use it the entire time. I'm one who doesn't use it for this particular behavior, I do not want the dog spitting/dropping the object when I click. I prefer to hold it and have the dog back away (move the head back). In other words I don't "take" it away from the dog, instead I put my hands on it and the dog removes his head when I say "give" or "out" or word of your choice. To me this is easier for the dog to understand.
An old retriever way of teaching "hold" is to use your finger. If it's a dog who mouths (which probably won't make any difference for a water dog) you can wear a leather glove. Basically it's putting your finger (I use my forefinger) behind the canine teeth, using the other fingers and thumb to hold the mouth closed, but not tightly. You can use your other hand on the collar to help keep the dog steady (most dogs throw their heads around, fling their bodies etc). You can tell the instance she slightly accepts holding your finger and release. The release is the reinforcement for her "hold" no matter how short that hold is at first.
You can shape her taking your finger if you want. I don't do that, too impatient, besides my dogs are use to me sticking my finger/hand in their mouth from puppy hood, it may be different for you. So I just open their mouth, put my finger in place and wait to see what happens.
The treat I use for doing this is canned cat food. You have the twin kitties (maybe more) so this shouldn't be a problem. I give the food on a spoon as I want the dogs to be OK with metal.
This takes how ever long it takes, some dogs quickly - just a day or two - other dogs maybe a couple of three weeks..
- Step 1 is as described
- Step 2 the dog opens her mouth for you to place your finger
- Step 3 the dog reaches forward for your finger when it's presented
- Goal - dog takes and holds finger as long as you want with no mouthing.
Now you're ready to use something else. I would start with a dowel about the diameter of your finger - maybe 6 inches long.
You'll want her accepting, reaching forward and holding various objects in the long run. A dumbbell is very good as it helps the dog understand how to pick up something, hold it in the center for balance. After that it will be easy to introduce the other things - ropes, etc that will be needed.
As you and the dog work together, he will soon learn that every time he sits down with the article, the next thing that happens is that you grasp the article and take it. He will begin to anticipate that chain of events, and may begin to release the article prematurely, allowing it to drop to the ground.
To prevent his anticipating the "out" cue, you can train him a set of special rules using a dummy. The rule is that when you take the rope, or slide your hand from the dummy to the rope, it means that you are about to initiate a game of tug by trying to pull the dummy away from him, and he is allowed go into a game of tug with you. In this game, you will growl, and pull, and try but fail to take the dummy away from him. Besides enhancing your relationship with a cooperative game, tug is great fun for the dog, so he has a strong incentive to hold on tight, because if he doesn't, he'll "lose" the game.
Make both this game and the "out" cue highly reinforcing, so that the dog is equally happy whichever one occurs, and is waiting to be surprised by whatever is coming next.
Not Returning with the Article
To be edited
On Aug 13, 2007, at 9:17 PM, email@example.com wrote:
> My 7-month-old Lab is very enthusiastic about running out to get the bumper/duck, but not so good about bringing it all the way back to me at the line. > Any tips for nipping this in the bud?
Hi, Kathy. Good question. Since no one else has replied yet, here are my thoughts.
First, the enthusiasm you mentioned is great, and IMO the most important thing. Whatever you do, you do not want to do anything to diminish your pup's desire.
Second, I've often heard that it's not unusual for dogs to behave as you're describing, and that later, some of those same dogs go on to become accomplished retrievers. Your dog loves not only the game of retrieving, but also the object, and that's a good thing. Someday the retrieves will require him to run straight through cover, to climb up and down steep banks, and to swim long distances, and he'll need a high level of desire to overcome the challenges.
Third, at 7 months, your dog may still be teething. If so, carrying the object may be causing him discomfort. Some trainers stop asking their dogs to retrieve until the adult teeth are in. I'm not clear on the timing of that, so maybe someone else can answer:
- How do you know when to stop asking for retrieves?
- How do you know when to start again?
Fourth, many trainers seems to separate out the game your dog is playing, and the Taught Retrieve. I think the idea is that, if the dog has not yet been through the Taught Retrieve training, you wouldn't necessarily expect a reliable return. I think your dog is almost to the age, maybe a month or two short, where you might want to carry out that training.
That said, I'd also like to add some specific thoughts on the return.
As a reward-based trainer who does not plan to use force/coercion/pressure to train your dog, you have several other tools available to you to deal with the issue: management, luring, and reinforcement. I've broken those out separately.
Management means controlling the dog's access to undesirable reinforcers during the period while the dog is learning the desired behavior.
One kind of management for retrievers is to attach a long lead (maybe a 20' nylon rope) to the puppy's collar whenever training. That allows you to catch the dog and reel him in if needed.
Perhaps more important is choosing the right environment to train in. The basic idea is that if you can't get the dog to come to you in the environment you're currently working in, you shouldn't be working in that environment.
A hallway of the house with all the doors closed and minimal distractions leaves the dog little choice but to return. Once he's confidently returning in that context, you can open doors, gradually introduce distractions, and little by little move into more challenging environments.
Luring means showing the dog something to return to. When you use luring, you're not asking the dog to come to you because he's been trained to, you're asking him to come to you because you're showing him something he wants.
The risk of luring is that the dog may only learn to come to you while that lure is present. But used judiciously, luring can be a useful tool for getting the dog in the habit of coming back to you.
The choice of a lure depends not on you but on your dog. What will he come back to you for? Many dogs will chase you if you run away from them, so that's a great lure. Many dogs are attracted by high pitched voices. Another possibility is waving a toy. I found with my puppy that he'd come running if I dropped down into a play bow position. You can also use food as a lure, though there are arguments against that that I won't get into here.
In the end, the most important tool for training the return is making it worth the dog's while. That's operant conditioning: the dog learns that something wonderful happens when he makes a particular choice. The next time he's presented with the same options, he'll be more likely to make the same choice again, and again something wonderful happens. After a substantial history of such reinforcement, that choice becomes habit, virtually reflex.
As with a lure, the choice of a great reinforcer is the dog's, not yours. You might WANT the dog to be highly reinforced by petting, for example, and some dogs are. But some aren't. The rule of thumb is, if the behavior increases (or a well-established behavior continues), it was reinforced. If it doesn't, it wasn't.
The reinforcement might come from the behavior itself. That's why your dog has that enthusiastic run-out. He loves doing it, possibly because it simulates the hunt of his ancestors. That's intrinsic reinforcement, and the behavior is sometimes called a self-reinforcing behavior.
Some behaviors are not immediately self-reinforcing, but later become self-reinforcing. A dog may not love swimming at first, for example, but after a few easy-going sessions gradually discovers what fun it is.
For a retriever, the most valuable reinforcement you can offer is typically another opportunity to retrieve. Again, there's that prey instinct.
Another way to tap into prey instinct is with the game of tug. My older dog's enthusiasm for the retrieve soared to a whole new level the day someone suggested I start playing tug with her when she got back with the article. The only thing I'll caution you about this, though, is that some trainers do not believe in playing tug with a retriever. I'm not sure if it's because they think it will develop a hard mouth, or because they think it will break the dog's "out", that is, willingness to give up the article when cued. I haven't found either to be the case with my dog.
Any of the ways I mentioned as luring might also work as reinforcement. The range is endless: besides throwing a dummy and offering a game of tug, other possibilities are praise, petting, a game of chase, rough-housing, and treats (again, note that some trainers do not use food for training retrievers). The key concept is that you don't decide what the dog finds reinforcing, the dog does. If you try to use praise as a reinforcer, for example, and the dog doesn't find praise reinforcing, then you are not going to get good results. Knowing this, you have the incentive to be a keen observer of how your dog responds to particular stimuli in particular environments.
I'd like to highlight the difference between luring and reinforcement:
- Let's say the dog is on his way back to you. If you start swinging a dummy and calling hey-hey, that might make him come running to you, which makes you happy, but he hasn't learned from that experience to come to you when you're NOT swinging a dummy and calling hey-hey.
- On the other hand, let's say that somehow the dog comes back to you (perhaps because of management or luring), and then at that very instant you begin swinging the dummy and calling hey-hey and then you throw it for him to chase. Now the dog has learned that something great happened when he got to you. After a few of those experiences, you won't need to entice him to come to you, because he'll be interested in the consequences of doing so, not the stimulus you used to get him to you.
A few other notes on reinforcement. First, just because a dog likes something doesn't mean it will reinforce behavior. The measure is whether the desired behavior increases. There's a good chance that offering the dog something he likes will increase the behavior he just completed, but it's not guaranteed.
Secondly, reinforcers tend to be context-specific. The dog may find a particular consequence reinforcing for one behavior but not another, reinforcing in one environment but not another, reinforcing on one day but not another.
Thirdly, the amplitude (effectiveness) of a reinforcer varies with Establishing Operations (EOs). Whether the dog is hungry will affect his enthusiasm for food. Whether he's tired will affect his enthusiasm for play. I'd say you want the dog to be fairly hungry and well-rested whenever training.
Sorry if I've gone past your intent when you asked for "tips", but without knowing you, I thought I'd try to cover the main bases.
Best of luck with your puppy. Please keep us up to date with your progress.
For a refresher course to build your dog's reinforcement history for the retrieve, see Practicing Fetch in the chapter "Marked Retrieve".
From private correspondence from Alice Woodyard to Lindsay Ridgeway on October 29, 2007:
"It sounds like you are being observant of their bird handling but remember that a bunch of repositioning of the bird, especially AFTER it has been picked up once, is a form of prey-possessiveness that is a close cousin to the not returning, eating, etc. It is a way to procrastinate the delivery and savor the prize. It is not "innocent" if you are seeing quite a bit of it. Not in a dog who is experienced with handling live game. Uh uh. The "N" on this one should go down with practice also. Also a big loop behind you to get into heel position, bigger than the dog normally needs to make, that's a form of parading with the prize, a form of "keep away" from you. As is any slight turning the head away right before the delivery or lowering the head to "get a better hold" when at heel and just before delivery. And of course any mouthing. These are all more little signs that the dog's head is not in the game the way you need it to be, not your game anyway."
To be edited
On Aug 25, 2007, at 8:41 AM, firstname.lastname@example.org wrote:
> Does anyone have some favorite drills/methods for getting your dog > back to the line? I've had him on a long line, but I'm thinking I > need to backtrack and do some drills separate from the actual > retrieving. He runs/swims out to the mark with lots of energy, but > slows down and stops before returning to me with the bumper/bird.
Hi, Kathy. The question I'd ask is, why SHOULD he rush back to the line?
Lumi had the same issue. So for several sessions, we practiced retrieves with our "start line" next to a creek. If I could, I'd get people to throw for me, otherwise I'd do poor-man's marks (Lumi in a sit, I go out and throw bumper, come back to her and send her). Either way, I reinforced every return with a rousing send-out into the water. Needless to say, after a rep or two, she was racing back for that chance to swim.
Then when we started training with a group, I took a chance on looking like a fool by doing the same thing with the group. Lumi would take her turn at the line, but as soon as she brought the bird back, instead of sending her out for another single or casually meandering back to our vehicle like every one else, I'd take her delivery of the bird just as calm as could be, and then, at the moment I had the bird, I'd spin on my heels and race off with Lumi to the nearest water, which was sometimes as much as 200 yards away.
After I'd done that a couple of sessions, I'd have her retrieve both birds, and THEN we'd go racing to the water.
You know what? A time came when she didn't seem to need those send outs to the water any more. She just seemed to enjoy running both ways. I call that "discovery training". The dog discovers that a behavior is fun, even though she didn't find it to be self- reinforcing at first. After that, the behavior maintains itself.
As for looking like a fool, that didn't bother me too much. No dog has a more enthusiastic return than Lumi, and some of the advanced ones I train with are like molasses.
Cheating is the field trainer's name for a dog running around on land instead of swimming straight to the article, or back to the handler, on a water retrieve. Of course, the dog isn't doing anything dishonest, he's just trying to get to the article or handler as fast as possible, and he can run faster than he can swim. Other factors may also trigger cheating, and the term is also broader than water retrieves, and includes for example running around high cover on a land retrieve.
For some of the less advanced retriever titles, such as Working Certificate (WC) and Junior Hunter (JH), cheating doesn't disqualify the dog. But at the more advanced levels, the dog is trained to take a straight line. That training is called de-cheating.
A number of training methods exist for de-cheating on water retrieves:
- One reward-based training method for de-cheating is called the swim-by method.
- Another reward-based training method for de-cheating is called pole de-cheating.
- Most field trainers also rely on e-collars for de-cheating these days. Use of an e-collar does not fall within the scope of this book (see Preface).
The swim-by method require a technical pond with some fairly specific traits not necessarily available to all field trainers.
(details to be added)
Here's a description of the pole de-cheating method, invented by Lindsay Ridgeway in July 2007 for de-cheating Lumi, his 3-1/2 year old female Golden:
- A prerequisite is that the dog has some experience with "over", a hand signal that sends the dog left or right. It's not necessary that the dog's response be perfectly reliable. In fact, if it were, you'd have no problem with cheating, you'd just cue "over" to keep the dog in the water.
- The set up is a straight or preferably convex shoreline, open water, and a pole of some kind in the water. For example, you could use a length of PVC pipe and wade out about 10 feet to push the pole into the floor of the pond.
- Stand at the neutral point on the shoreline, closest to the pole. Bring the dog to heel, and throw the dummy to the opposite side of the pole, then send the dog.
- When the dog turns back with the dummy, you are already the shortest swim back, so the dog has no way of cheating and will come straight back. But use an "over" cue to send the dog around one side of the pole or the other.
- Practice that a few times to be certain that the dog is reliably responding to "over" and swimming on the side of the pole you cue.
- Now move a few yards to the left or right for the next retrieve, and again throw the dummy over the pole. When the dog is swimming back, the shore may act as slight suction, but cue "over" to send her to the far side of the pole, thus keeping her away from the shore. If that doesn't work, you may have stepped too far down the shoreline, or you may need to work more on "over" around the pole by going back to practing at the neutral point.
- Since it's preferable that the dog not require handling, that is, not require you to cue "over", use extrinsic reinforcement (praise, tug, a hey-hey-dummy, food) to reward the dog when she takes the far side of the pole without having to be cued.
- When the dog consistently takes the far side of the pole, you can move a few yards further to reduce the angle between the shoreline, and the dog's swim path, even more.
Extending the Behavior
The pole de-cheating method can be gradually extended to an advanced version of the behavior:
- Gradually move the starting line further down the shoreline from the pole, reducing the angle until the dog is swimming parallel to the shore.
- Increase the distance the dog swims to get to the dummy. You can do this by leaving the dog in a sit while you go to place the dummy, or by taking the dog with you and then walking back to the starting line, or by having someone else place the dummy for you.
- Place the dummy on land so that the dog swims in a straight line, climbs onto shore to retrieve the dummy, then re-enters the water to swim back.
- Place the dummy on water on the other side of land, so that the dog swims to the far shore, crosses the strip of land, re-enters the water to retrieve the dummy, and then returns the same way in a straight line.
- Switch to a smaller and smaller pole so that the dog becomes less and less dependent on that visual cue.