User:Lindsay Ridgeway/Reward-based Field Training for Retrievers/Blind Retrieve
- 1 Blind Retrieves
- 1.1 Sight Blinds
- 1.2 Handling
- 1.3 Handling 101: Pinball Drill
- 1.4 Handling 102: The Loopback Drill
- Gives dogs confidence that a bird is out there, useful for multiple marks as well as for blinds
- Lets the dog run with confidence, rather than developing a tentative send-out
- Re-use the same sight blinds at various training locations day after day and help the dog develop memory skills, which may improve his memory for multiple marks as well
- Sight blinds to a pile, to condition both "back" and "over" cues
- The locked-in stare
- The cue: "back"
- Straight away from you
- Used for send-out
- Used from whistle sit
- — Spin either direction
- — Why use it? The dog is already going that direction
- You anticipate that the dog is about to veer around a factor
- Left and right
Angle Back and Angle In
Although it is traditional to train "back" and "over" first, a case can be made for training the angle back and angle in first:
- Let's think of directions as a clock face, with 12 o'clock being "back" and 6 o'clock being recall. Your visual cue is the hour hand on the clock face, and the dog's direction is that clock face lying on the ground. So, for example, you point directly to the right, 3 o'clock, and your dog is to run directly to the right. You point to the right but on an upward angle, about where the hour hand will be for 1:30, and your dog is to run toward 1-2 o'clock. You point to the left on an upward angle, and the dog is to turn away from you the other direction and run toward 10-11 o'clock.
- Besides recall (6 o'clock), by far the most common directions are 10-11 o'clock on the left, and 1-2 o'clock on the right. These are called angle backs.
- The second most commom directions are 9 o'clock and 3 o'clock. These are called overs.
- Somewhat less common are 7-9 o'clock and 4-5 o'clock. These are called angle ins.
- Therefore, you may find it best to train angle backs and angle ins until they are solid. At that point, you may find that the dog automatically knows what an "over" is without having to train it.
- "Give it"
- Whistle "sit"
- Steady sit until released
Training the Behavior
- Place a single dummy at between 1:00 and 2:00 on the clock face, and another between 10:00 and 11:00. Put the dog in the center of the clock face and stand directly in front of him, so that he is facing you and toward 6:00.
- Fold your hands into a prayer position and whistle "sit" (one tweet).
- At the very instant that the dog sits, reinforce his response by rapidly lifting your left or right arm straight out to the side, but at an upward angle. Use an open hand with the palm facing forward. While moving your hand, also move your body in the same direction, as though you were about to run toward the bumper on that side.
- If possible, do not use a verbal cue. Ideally, the dog will respond to the visual cue of your rapidly moving arm. The reason you'd like to skip using a verbal cue is that eventually your dog may be competing at distances where he can't hear you. If necessary, send him with the verbal cue "back", but fade that as soon as possible.
- When the dog picks up the dummy, whistle recall.
- When the dog returns the dummy to hand, reinforce well, with treats, praise, a hey-hey dummy, tug, or whatever maintains and increases the behavior.
- If necessary, put on the dog's lead so that you can throw the dummy back where it was and set up for the next rep.
- Practice both sides.
- Keep the sessions short and fun.
Some dogs will see that two dummies are out there and will want both of them. They may run to the one you send them to, but then, instead of returning to you, they'll head over to the other dummy.
When a dog interacts with a different bird/dummy than the one you sent him to, it's called "switching", and it's important that the dog not develop that habit. If you find that your dog switches during the initial training of angle backs:
- If possible, increase the distance between the dummies. The best situation would be where the dog is aware of, but not distracted by or attracted to, the dummy that is out of play.
- As soon as you release the dog, run behind him so that you are quite close to him by the time he picks up the first dummy.
- Then cue "give it". Because you are so much closer than the other dummy, and because he knows that "give it" will be highly rewarded, he'll return the dummy to you. At that point, you can throw a hey-hey in the opposite direction or reinforce in other ways, but block him from running to the other dummy or catch him if he gets past you. After reinforcement, put the leash back on him to set up for the next rep.
- Another method is to work with a barrier between the two dummies, such as a fence or hedge, starting the dog at a gap in the barrier.
- Over time, gradually reduce the distance you need to run with the dog to prevent switching, until you don't need to run with him at all.
- Then over more time, gradually reduce the distance between the dummies.
- The send-out:
- — Lining up the sit on either side
- — Focusing the eyes
- — Sending out with "back"
- — When to challenge the line:
- – – In training, no sooner than 75% of the way if possible
- – – In competition, as soon as dog is off line
- Holding the line
Training the Behavior
A drill known as the wagon wheel is used to train lining:
- With dog watching and waiting, throw out a number of dummies in various directions.
- Choose one of the dummies.
- Walk with dog at heel toward the dummy.
- Cue "sit".
- Use hand on dogs side to focus dog's eyes on dummy. Train him that your hand tells him when he is looking the correct direction:
- — Once dog is looking at dummy, keep hand in position while dog maintains stare.
- — Remove hand when dog looks away.
- — When dog looks back, replace hand.
- When stare is locked in, send with "back" cue.
- If he gets off line and goes to the wrong dummy:
- — Occasionally, you can call him back to heel without letting him finish the retrieve, but that is highly de-motivating to some dogs.
- — Instead, you can let him finish the retrieve, then use a gentle, neutral tone to cue "sit", take the dummy, and throw it back where it was, without reinforcing the retrieve.
- If he goes to the correct dummy, reinforce in one or more ways (remember, the dog's behavior tells you what he finds reinforcing), such as:
- — Whistle recall as soon as he picks it up.
- — Cheer as he returns, such as with applause and/or "Yay!".
- — Play tug after he returns the dummy to hand.
- — Throw a happy dummy when he returns the dummy to hand.
- — Give him a treat when he returns the dummy to hand.
- Throw the dummy back out to its position.
- Repeat to a different dummy.
Extending the Behavior
- Increase the distances of the dummies.
- Decrease the angles.
- Gradually work with dummies that are different distances. Dogs tend to swerve off line to retrieve the closest dummy, so be patient.
- Retrain in different and increasingly distracting locations.
- Introduce more difficult distractions. For example:
- — Send to a dummy when a dead bird is on either side.
- — Scent an area with a dead bird and then send the dog through that area to a dummy that is not in sight until the dog has gotten through the scented area.
Handling 101: Pinball Drill
Here's a drill Lindsay Ridgeway used to strengthen Lumi's handling skills in the late summer of 2007. Lindsay liked the drill so much that he ended up using it as the initial method of training handling to Laddie, then four months old:
- Prerequisite: Whistle sit.
- Use six surveyor's flags in a zig zag pattern leading away from the "start line". Space the flags according to your results over a series of training sessions. An initial spacing of 10 yards may be about right.
- Spend as much time as needed, possibly a few seconds, possibly a few sessions, to train the dog to run to between flags 5 and 6, and sit on single tweet of the whistle as he reaches the flag. Click and treat as he reaches the flag and responds to the whistle sit. If the dog bites the flag, don't reinforce that rep, and make a point of whistling "sit" well before he reaches the flag on future reps.
- When you start the training, you may find it most effective to run with the dog, then send him the last few steps as he locks onto the flag. Gradually reduce the amount of distance you are running with the dog, until you and the dog can start near flag 5, you can send him to flag 5, and then you can send him to flag 6 with minimum movement yourself.
- Continue to train, moving your starting point closer to flag 4.
- Add flag 4 to the training. Now the dog runs to flag 4 and sits on cue, then flag 5, then flag 6. As he sits at flag 6, sometimes you can run to him to reinforce, other times you can call him running to you and reinforce as he reaches you.
- In the same way, add flags 3, 2, and finally 1 to the training.
- As you practice this drill in session after session, the dog is learning the pattern of an actual cold blind, a narrow zig zag between the start line and the article being retrieved.
- Use as many sessions as you need to develop this invaluable skill. Keep each session short and always end while the dog wants to keep playing.
- Continue editing here
- Over several sessions, make the diamond larger and larger. Soon, the dog should not be able to see the flags when he first goes out on each cast. This is especially true if you use orange flags, which are difficult for a dog to see until the dog is close.
- Add more flags to the course, enabling you to practice all the casts from a variety of distances.
- When the dog is ready, add a dummy or bird at the far end of the course. Place it so that the dog cannot see it until he is on the way from the last cast. Having the dog complete the retrieve takes extra time but builds motivation.
- To make the drill even more challenging, train the dog to ignore distractions and take the cast. Add distractions such as dummies or dead birds, first so far from the dog's line of travel that he barely notices them. Over time, move them closer to the dog's line. If the dog goes off line to one of the distractions, walk out to him, bring him back to the flag, walk back to the start line, and try again. If he again strays, you've moved the distraction too close too soon. If he gets it right, find a way to immediately reward him for his success.
- Similarly, gradually proof for other factors: obstacles such as logs for the dog to jump over, crosswinds, hills, and so forth.
To build the dog's love of the game, you may find it beneficial to occasionally and randomly leave the start line at the instant the dog sits and run out to reinforce, with food, petting, a game of tug, a thrown dummy, or anything the dog loves. Then put him back in his sit, go back to the start line, and give the next cast. You may find it best to reinforce a lot at first, then gradually introduce intermittent reinforcement, keeping the dog guessing as to whether this is the time that you're going to come running out.
Why Surveyor's Flags?
Here's a post to the PDG list describing the reasons behind using surveyor's flags for this drill:
On Aug 25, 2007, at 4:53 PM, Lindsay Ridgeway wrote:
Lumi and I started working on cold blinds about a month ago, and when we tried one with a Field Trial training group Wednesday afternoon, I realized that we needed to train more on our own before trying one with a group again.
As background, I began training Lumi overs and backs, angle backs and angle ins, months ago, and we've got a dozen or more schooled blinds we practice daily, one or two per location we happen to train at. We've also done many sight blinds at various distances up to 300 yards for months.
But when it comes to cold blinds, Lumi exhibits some of the same characteristics I've seen in other, more advanced dogs, all of whom are trained with e-collars:
- "Do I really have to sit and/or change directions when I think I already know where the dummy/bird is?"
- "I saw your visual cue to send me after the whistle sit, and here I am on my way, but am I going the right way? Gosh I hope you don't stop me and throw another cue at me, I'm doing the best I can."
- "First you send me to the right, then you send me to the left. I'm starting to wonder if you even know where the bird is yourself."
- "Whistle sit? Oh, I remember that. I was in the middle of a delicious run toward a dummy/bird in clear sight, but before I could get to it and bring it back to you, you made me screech to a halt, turn to face you, and sit. That was no fun. And not just once, but over and over again, every time we practiced handling. Every time you did it, that whistle sit became less and less appealing. Sure you'd send me again after the sit, but I was already on the way. Why couldn't I just keep going?"
BTW, Lumi has HD and arthritis in both hips. Sitting has never been her favorite thing anyway.
I could see from the beginning that sending Lumi toward a dummy and then stopping her before she got to it, so I could cast her in a different direction as in a baseball diamond, was de-motivating, but at first I figured she'd get over it. In reality, I think the opposite may have been happening. I think stopping her on the way to the visible dummies at "second base" may actually have been cumulatively undermining Lumi's reinforcement history for the correct response.
We also trained on carved out paths. I could send Lumi straight down the path, whistle a sit at the fork, and then cast her to one of several dummies on the various spokes from there. But I never found anyplace to practice that drill with much distance, and even that seemed to be de-motivating. "All those dummies out there, but I have to stop and turn to face you."
Trying to find a way to build motivation, I also tried sitting her in place, throwing dummies in different directions away from her, and then walking away to a "start line" at some distance from where I'd left her sitting. I think that drill improved her understanding of the casts, and it seemed to be fun for her, but it was such a different scenario from a cold blind that it only helped so much.
What I really wanted was a pin-ball set up, where I could send Lumi from target to target, and reinforce every sit in some way, such as throwing a dummy or running out to give her a bite of food.
But the first time I tried it, I realized what a disaster it was if I used dummies as the targets. Either she would pick the dummy up, meaning that she'd either want to bring it to me or she'd have to carry it to the next one, and THEN what would she do? Or I'd have to persuade her not to pick it up. Just great, training a dog to run to a dummy and NOT pick it up. After one session, I could see that using dummies as targets for pin ball was a mistake.
A couple of days ago, I hit upon what may be a solution for all these difficulties. I tried it that morning, with Lumi and even with little Laddie, and was pleased with the results. It trains the remote sit and the casts, and it's FUN. No running toward dummies and not being able to pick them up. Instead, both dogs were excited the whole time. At each sit, Lumi's body language gradually became, "I'm sitting, I'm sitting! What's next, Daddy, what's next?"
I used the same technique a couple more times for Lumi, stretching the pin-ball course out to 100 yards so far, and for the first time I have a dog running from target to target and throwing herself into a sit each time I whistle.
The solution was surprisingly simple. What I did was to set up my pin-ball course with surveyor's flags. Those are thin wire poles with small flags attached. You can buy bundles of them at the hardware store. I use the orange ones because I've heard that dogs can't see orange.
These were the advantages I was looking for:
- The dog has no particular inclination to retrieve the surveyor's flags. Laddie (4 mo) wanted to bite them some, but it was pretty easy to show him that running to the next one was even more fun.
- The further they are apart, the harder they are for the dog to see. At some point as you stretch out the distances, the dog is running the right direction based on your cast, trusting that eventually the target will come into sight, and sure enough, it does.
- With no dummies in the field, you can focus your training exclusively on the cast and the whistle sit, and use any method you like (or rather, that the dog likes) for reinforcing each success. Lumi's favorite version of the game was the session when I had our "start line" next to a creek and sent her into the water as soon as she came racing back from the furthest flag.
- You don't have to make the dog not pick up a dummy.
- The dog has the satisfaction of knowing with confidence that she's given the correct response, because the flags provide a physical representation of the remote sit concept. By analogy, a novelist might describe a character carrying a knife, rather than just saying that the guy looked dangerous.
- Once you've stretched your patterns out far enough, you can add a dummy at the end, out of sight until the dog has been cast from the last flag. Now you've got an actual cold blind, and a precise course in which to train the exact whistle sits and casts you wish to.
- One day, the flags play such a small role in running the course that they become superfluous and are used only for tune-ups after that.
The idea of using surveyor's flags seems so obvious to me now, that I have to believe other people also use them. But I haven't heard or read of anyone else doing it, and when I mentioned the idea to a couple of my new training buddies, they looked at me like I had two heads. Since they use -R and aren't worried about building a +R history for their dogs' behaviors, they might not have much use for the idea.
Well, my goal was to have Lumi fall in love with whistle sits and remote casts. The surveyor's flags seem to be giving that result, so I thought I'd pass the idea along.
Maybe it will help someone else. Maybe someone can suggest an even better way to train remote casts. Or maybe someone can warn me of some shortcoming to the method I've described that I haven't realized yet.
Handling 102: The Loopback Drill
While the angle-back is an excellent initial handling cue, one of the most common situations that requires handling is when the dog is tempted to "cheat", that is, to detour away from the direct path to the bird. He may detour:
- To avoid an obstacle
- To take a land route instead of a water route ("bank running")
- To touch a land point instead of swimming past it
- To avoid a headwind
- To climb upward or downward on a hillside
- To check out scented ground
- To get to a diversion bird
- For other reasons known only to the dog
In those kinds situations, using an angle-back may not be satisfactory, because the dog's score will be reduced, or the dog may be dropped from the competition, if he does not take the required route. Instead, you may need to undo the detour by sending the dog "over", directly to his right or left, often just a few yards, and then using a back or an angle-back to send him toward the bird on the original line. The dog's ability to respond correctly to "over", for each of the factors listed above, is a key handling skill.
One way to train "over" specifically to erase an attempt to cheat is with the loopback drill. Once trained, it looks like this:
- Place the dog in a sit next to an obstacle. For example, the obstacle could be an agility jump bar, a log, the shoreline of a water feature, a stiff crosswind, or halfway up a hillside. Eventually, the dog should practice all of the obstacles listed above.
- Place a dummy or bird behind the dog at 12 o'clock, some distance back, perhaps 20 yards.
- Stand in front of the dog at 6 o'clock.
- Cue "over" both with voice and visually. The dog runs directly to his left or right and overcomes the obstacle.
- When the dog has overcome the obstacle, whistle "sit". The dog sits, facing you.
- Cue an angle-back to the dummy (to train an angle-back, see "Handling 101: The Pinball Drill"). The dog runs to the dummy and picks it up.
- Whistle recall. The dog runs back and delivers the dummy.
- Reward the dog.
Training the Behavior
Although your dog may be able to learn the loopback drill in a single session, he may not be able to do it correctly if you simply set the loopback course up as described and try to direct him through it.
To solve that, take as many sessions as necessary to train each of the individual components:
- "Over" through jump bars or over a log in either direction, perhaps with a dummy on the other side for the dog to retrieve.
- An angle-back to a pole where the dummy will eventually be placed, first halfway from the corner where you'll be whistle sitting him, then 3/4ths of the way, and so forth.
- "Over" through an obstacle to a pole at the corner where you'll whistle sit the dog, followed by an angle-back to the pole where the dummy will eventually be placed.