Charging the Clicker

Here is a clip we took of Arlie’s clicker introduction last January (I’m a little slow getting posts up!).

All we’re doing in this clip is teaching him what the sound of the clicker means. This is called “charging the clicker”.

With no commands involved , you click/treat, click/treat, click/treat. Within about 20 or 30 repetitions most dogs will begin to anticipate that a treat will be coming shortly after the click. At one point late in the video, I click while he’s looking away and he swings his head to me. I know at that point, that he knows what the click means.

If you notice in the video I keep my hand motions to a minimum. Dogs are masters at picking up on our body language. I want him to associate the food with the click, not with my hand motion.

Clicker training

I teach all new behaviors through clicker training (also called marker training). A clicker is nothing more than a little plastic box that has a metal springy thing in it that makes a clicking sound when you press it.

What the clicker does is sets up a clear line of communication with your dog. Clicker training is the closest you will ever get to talking to a dog. Dogs see the world in a series of quick photo like snapshots. In order for a dog to learn, reward or punishment needs to come within a second or two of the behavior that you are trying to teach. Because of this very short time window, many times people reward or punish the dog for a behavior that is coming AFTER what they are trying to teach. While some people do use punishment markers, we’re going to stick with reward markers.

What a clicker does is mark a desired behavior at the exact moment it occurs. Think of the clicker as a photo for the dog and also a paycheck.

Just like in the gunfire introduction where we paired the sound of gunfire with a bird coming (boom-bird, boom-bird, boom-bird), in clicker training you pair the sound of the click with food coming (click-food, click-food, click-food). The click will become a reliable predictor that food is on the way and it will elicit the same response as the food. It’s another good example of classical conditioning as it relates to dog training.

You need to do what is called "charging" the clicker. With no commands involved, you click, give the dog a treat. Click, give the dog a treat. Click, give the dog a treat. Eventually, the dog will learn that every time they hear a click a treat is on the way.

What this does now is gives us the ability to mark the desired behavior with a clicker at the EXACT moment it occurs, buying us time to deliver the reward, so we’re not rewarding for a behavior that occurred after what we wanted. Lets take sit for example. You can lure a dog into a sit and mark the behavior with the clicker at the exact moment their butt hits the ground. The clicker say "yes, right there, that is exactly what I want! Your reward is on the way".

This is very important feedback for the dog. It makes it very clear to them when they’ve done the right behavior. 

There are 3 different ways of getting a behavior. "Food luring" (self explanatory, yet there is an art to doing it effectively), "capturing" which is waiting for a behavior to occur naturally.....in the above example you would wait for the dog to sit on their own, and then mark it with the clicker, and the third is "shaping". Shaping is like playing a game of hot or cold. Here is a good example of shaping. Let’s say you wanted a dog to stand on a place board for whoa training. The dog has no idea you want her to stand on the board. As soon as she gets close to the board, you click and treat (C/T). Then, if she sniffs the board  you  C/T. Then, when she put one paw on the board you C/T. Then, when she puts two paws on the board she gets a C/T. Then, three paws- C/T. Then, all four on the board-C/T and JACKPOT! Lots of treats for four paws on the board! You shape her into getting on the board by marking the behavior every time she gets closer to our final goal. Training like this creates a thinking dog.

Whether you use luring, capturing or shaping to get a behavior while clicker training you should never give the verbal cue/command until the dog is performing the behavior with regularity. Words will do nothing but confuse the dog while they are learning something new. It makes it much easier for them to understand if you teach the behavior first, then put it on cue later. By teaching the behavior first, then naming the command after, the dog says “hey I know what that means, no problem!”. By saying the command before they know what you want, it’s just confusing chatter to them, making it harder for them to learn.

The order is:

1. Get the behavior
2. Mark the behavior
3. Name the behavior

One very important thing. Always follow up every click with a treat! If you start clicking/marking and don’t follow up with something the click or mark will lose it’s value. Even if you click accidentally, you must always follow up on it.

Although it’s not a bird dog, here are a couple video clips I found that explain the clicker method very well.

http://www.youtube.com/watch?v=FAb3KJyxq0o&feature=related

http://www.youtube.com/watch?v=H4DknI1KEsE

http://www.youtube.com/watch?v=Ciy7uUtA6S8

Here is a clip of Chad Hines from Willow Creek Kennels (one of the most innovative bird dog trainers around) explaining clicker training.

http://www.youtube.com/watch?v=nfC8-qjtm9A&feature=plcp

Obedience Training

If someone would have asked me 5 years ago about training birddogs with food I would have laughed at them. I had the old school mentality that a dog should do what it’s told because I said so. I was 100% wrong. Now, the majority of my obedience training consists of food/reward based training.

There are two opposite ends of the spectrum with dog training. At one end you have the old school yank and crank compulsion based trainers. A typical compulsion based trainer puts a choke collar or a pinch collar on the dog and start yanking. A good example would be heeling. The trainers say “heel” and yanks on the choke chain, “heel” yank, “heel” yank, sooner or later the dog figures out that if he stays by your side he avoids being yanked. The dog learns to avoid the yank by staying at your side.

Does it work? It sure does. Avoiding something unpleasant is a very strong motivator. But when you think about it, it makes about as much sense as spanking a child until they figure out what you want them to do. Eventually they learn, but the results usually aren’t very pretty. Since their only motivation is to avoid  an aversive, instead of working to obtain something, they’ll do it, but only because they have to and are afraid of  being corrected, not because they want to. A dog that’s only trained through compulsion  doesn’t have an open mind to learning, and doesn’t enjoy training sessions. You can usually tell a dog trained in this manner. They typically will have a lower body posture when training, tail down, head down. They look like they want to be anywhere but there. It robs the dog of confidence and style. Usually the trainer conditions the dog to tense up when they hear the command because they’re expecting a physical correction every time they hear it.

The upside of  compulsion training is that it’s reliable and  fast.

At the other end of the spectrum is positive reinforcement only trainers. A positive only trainer relies strictly on giving and withholding of  rewards to train. Reward the dog for performing the correct behavior, withhold the reward for unwanted behavior. The beauty of  positive reinforcement training is it creates a dog who thinks for itself to access things that it wants, creating a bold confident dog  that looks forwards to training and has an open mind to learning. It encourages the dog to offer up new behaviors because he enjoys learning. They become an active part of the training process, not reactive. It doesn’t rob the dog of style like compulsion based training, and there is virtually no fallout if done improperly.

The downside of using only positive reinforcement, is reliability amongst distractions. Although there are some trainers of working dogs who have trained to very high levels using positive only methods, positive only training typically doesn’t produce the reliability that compulsion based training does when the distraction level gets high.

So where does that leave me?

Some where in the middle with a balanced common sense approach! Compulsion based trainers feel that positive trainers are only bribing their dogs, while positive trainers think that all physical corrections are a level 10 and should never be used.

When I train with food, I’m no more bribing my dog than my employer’s bribing me by giving me a paycheck. If I don’t work, I don’t get paid. The same is true for my dogs. When it comes to physical corrections with dogs, not all corrections need to be a 10. If I said your name and you weren’t listening and I walked over and tapped you on the shoulder (physical correction) to get your attention, I doubt that would be considered excessive force. The same would be true if I held your hand to physically guide you, which is no different than leash pressure when done properly.

My training philosophy is to train the behavior with 100% positive reinforcement methods, then teach the dog to turn off leash pressure separate from any commands, then apply the pressure to the known command. Don’t let the term pressure fool you. My idea of pressure is simply nothing more than teaching the dog to give into the lead, then eventually an e-collar. I’ll teach him/her to give into the lead with positive reinforcement. Once the dog learns to give into the lead, a correction usually becomes nothing more than gently moving the dog with the lead to guide them to the correct behavior. If  the dog knows the command that has been taught with positive reinforcement and knows how to turn off pressure (this is called negative reinforcement)  without any commands, then you apply very gentle pressure to the known command - the dog will have a complete understanding of how to respond and will keep their enthusiasm the whole way even when a correction is applied. This may all sound very confusing, don’t worry, I’ll be explaining this in great detail as I continue with this blog.

Here is a video clip of one of the best protection dog trainers, Michael Ellis, explaining this concept and the fallout of using compulsion to teach new behaviors and why it’s important to teach the dog to turn off pressure without any commands. I wish I would have seen this video and understood this concept 10 years ago!

http://www.youtube.com/watch?v=2wQdqa2WfzU

Arlie Update

Here are a couple recent clips of Arlie working launchers. He’s doing very well at getting steady without pressure. I started planting multiple launchers to teach him to be careful, “there just might be one more here”. I think it’s working!

Here’s one of him letting Sue get in front.

 

He doesn’t know any stand or whoa command. He’s had zero physical corrections in the presence of birds. This has all been taught by the birds. We’re very proud of him!

Gunshy cure / Gunfire introduction

Once the dog is actively pointing and bird crazy, I like to introduce the dog to the gun.  Unfortunately, many dogs that come into the rescue had bad prior experiences with the sound of gunfire due to poor introduction. Err on the side of caution and treat all dogs as if they are gunshy. Gunfire is one area in dog training where an ounce of prevention is worth a pound of cure. I remain optimistic that most dogs, if they like birds enough, can overcome gunshyness with a patient trainer.

Don’t ever test a dog for gunshyness by shooting over him! Condition a dog to the sound of gunfire and you’ll never have a gunshy dog.

We’re going to use a behavioral term called counter conditioning, combined with desensitization. Counter conditioning is a form of classical conditioning. Classical conditioning pairs something that has no meaning to something of high value to achieve the same response. Do you remember back in school learning about Pavlov’s dogs? Scientist Ivan Pavlov learned that every time he rang a bell prior to feeding dogs they would salivate. Counter conditioning is the same as classical, except instead of pairing two things to achieve the same response, the goal of counter conditioning is to change the response. The gunfire isn’t neutral to the dog anymore. We want to change to dogs fearful response to the sound of the gun, to “alright! I heard a gun! Whoo hoo! Where’s my bird!”

We’ll achieve this by exposing the dog to gradually increasing sound levels of gunfire (desensitization) while pairing the sound with something of high value…..a bird.

 Here are the pre-requisites you and your dog need to get started.

1. Your dog needs to be bird crazy. He needs to be actively pointing birds and aggressively hunting.
2. Your dog needs to like retrieving, not necessarily to hand just likes to get his mouth on a bird.
3. You need the ability to be patient and work at the dog’s pace. The dog will determine how long this takes, not you.

 Here are the items you need.

1. A good supply of live birds. I prefer quail for this because they’re small and won’t spook a skittish dog on a retrieve, but you could also use pigeons. Don’t try to skimp and use a dead/frozen bird. Nothing is more exciting to a pointing dog than a live bird.
2. An open mowed area to work that you can also shoot a shotgun. It’s imperative that the grass isn’t so tall as to hide a bird on the ground. The dog must be able to see a bird on the ground. This is very important.
3. .22 blank starter pistol
4. Bag to hold your birds
5. An assistant
6. A small gauge shotgun, .410 with ½ oz. load (preferred) or a  28 gauge or 20 gauge with a ¾ oz. load max. If you have a break action gun, another option is to buy .410 gauge reducers. Gauge reducers are sleeves that go into the chamber of a larger gauge shotgun, that enable you to shoot a smaller gauge shell in your larger gauge. Fiocchi also makes ¾ oz. trainer loads for a 20 gauge that is relatively quiet compared to a standard 20 gauge load. Our goal with the light loads is to keep the noise level down. We don’t want to be using a super loud 1 1/8 oz. trap load.
7. Bright surveyors tape
8. Children’s cap gun
9. 20 ft. check cord

Let’s get started.

If a dog is really, really gunshy, they may panic at just the sight of a gun. If this is the case, I would take a gun out just prior to feeding each night. Let the dog see the gun and then feed him. Do this every night until he makes the connection that every time he sees the gun, a good thing happens. He gets fed.

Once the dog is ok with the sight of the gun, it’s time to work on the noise fear. First thing you need to do is get set up in the mowed area and have 2 quail ready with 4 or 5 primary flight feathers removed from each wing to limit its flight. We only want the bird to go about 15 to 20 yards then flutter down. Have your assistant step off about 40 yards with the toy cap gun. Remove a bird from your vest and tease the dog with it. The dog should be focused completely on the bird in your hand. With the dog focused, throw the bird and let him chase. With the dog in hot pursuit and about 2 seconds before he gets his mouth on the bird your helper should fire the toy cap gun. It’s very important to have this timing right. You don’t want to fire as he’s retrieving or after he has the bird. Nor do we want it 10 seconds before he has the bird. For the best association, it should be within a couple seconds prior to the bird in his mouth (remember Pavlov’s experiment). This is why we’re using a mowed area. If the grass is too high it will take the dog too long to find the bird and the association won’t be as strong. If all goes right, he will have a bird in his mouth and be happy about it. DON’T rush and take the bird away. Let him hold his prize as long as he wants. If he wants to chew it, let him. We don’t want any bad association at all. Only take the bird when he’s dropped it on his own accord. If at anytime during the chase the dog took notice of the noise, (even just a look back) the shooter must move at least 20 or 30 steps back before you do another bird and start over. Even if he retrieved the bird with enthusiasm. This is rarely a problem with the cap gun but it can happen. If he doesn’t take notice, then have the shooter move up 20 steps and repeat the same drill. If he doesn’t take notice from there then move up another 20 steps until the shooter can fire right next to you and the dog will take no notice to the cap gun while in hot pursuit.

I typically don’t like to do more than 2 birds per day. The reason being is I always want to end with the dog wanting more. I don’t want him to get bored of the retrieve. He’ll come back the next day and want the birds twice as much.

Ok, we made it thru the cap gun with the dog’s tail wagging. Now it’s time to step it up to the .22 starter pistol. This time we’re going to have the shooter step out 100 yards. We’re going to do the same drill. Pull out a bird, tease the dog with it, throw it and have the shooter fire 1 shot about 2 seconds before the dog gets his mouth on the bird. (Note:  at 100 yards you’re going to have to do some orchestrating with the shooter. You will have to tell him when to fire because he probably won’t be able to see well enough to time it right.) Just like before, if the dog shows even the slightest sign of noticing the shot you must move back farther. We’re building on success here, not failure and once again, don’t take the bird from the dog. Let him do what he wants with it. This all must be a good experience. Only take the bird when he drops it on his own.

If all goes well and the dog doesn’t take any notice, move forward 10 steps and repeat the drill. From now on, because of the increased noise level of the .22, we’re going to move forward in 10 step increments only, unless the dog takes notice. Then you have to move 20 to 30 steps back and start over. Don’t ever move forward if the dog takes notice. Even if it’s just a pause or a look back. No more than 2 birds in a day.

Repeat this drill with the .22 until the shooter can fire right next to you without the dog taking notice.

Once you can fire right next to the dog with the .22, you can now move up to the .410 gauge. Yeah, you guessed it, right back to 100 yards and repeat exactly what you did with the .22 until you can fire the shotgun right next to the dog without him noticing. Through all this, his whole focus should only be on the bird.

Now we can move to the field and actually shoot a bird for him. It’s not uncommon for a dog to get a little unsteady on his birds after this. After all, we did give him a lot of free retrieves. Don’t sweat it. You can fix that later.

When it comes time to shoot one over him I still like to make sure he gets a quick retrieve. I’ll plant a bird in a launcher with a couple wing feathers removed to make sure the bird comes down just in case you miss.  Tie a piece of surveyor tape to the bird’s leg so it is easier to spot in taller grass/field. Once again (I know I sound like a broken record) but don’t take the bird away from him until he drops it. I firmly believe a lot of dogs are turned off to retrieving by people taking the bird from the dog too fast, making the dog think he did something wrong, especially softer dogs.

Congratulations! You’re now shooting over your previously gunshy dog!

What we’ve done here is desensitized the dog to gunfire by gradually bringing the noise level up slowly, while making a positive association with the gunfire by the dog getting a bird in his mouth a split second after hearing the shot every time.

A few things to keep in mind. Even though you’re shooting over him, you must still be careful of a few things when you take him into the field.

1. Don’t shoot near him unless a bird is involved, for example: keep him away from target shooting etc.
2. Don’t shoot non-pointed birds. Only shoot when he’s pointing a bird. Shooting a non-pointed bird could catch him off guard and startle him.
3. Don’t have multiple people shoot at once over him.  Too many shots at once sounding like World War III can still spook him. I have a 2 shot limit per bird.

I also use this same system for introducing a dog to gunfire that has never been exposed to it before. The only difference is if the dog has never been exposed to gunfire before, I can move thru the program much faster. I will typically move ahead in 20 step increments instead of 10. And if all is going well I might do up to 6 birds a session versus 2 for a gunshy dog. I can typically do the whole .22 segment in 1 session, and the shotgun segment in another session with a dog that has had no previous exposure to gunfire.

Getting there

Here's a couple videos of Sassy and Gunner getting closer to our final goal. Both dogs get a little antsy and move when we walk in front of them. I'll have a few tips in an upcoming post on what to do about this.

Uncertain points

Many times the dog will get slight wind of a bird but not be quite sure of its location usually caused by a wind shift. They’ll establish point but then start looking around and or start wagging their tail (this is called flagging).

Here are a couple clips of Gunner and Sassy. In these videos both dogs stop at first scent but are uncertain. They both break point, but not to pounce on the bird. They break point because they’re doubting themselves. You can tell by the way they put their heads down and move in another direction that they’re not trying to move in on the bird. I treat this the same as creeping, as soon as they move I launch the bird.