TRAIN YOUR DOG: Sue Ailsby Clicker Training Seminar - Part 2
By Sharon Schaefer
In Part One I covered the mechanical skills of clicker and treat delivery; dogs targeting fingers or pens; and starting hand or counter zen (respecting food not belonging to them). Sue no longer "charges the clicker" because she wants the dog’s first learning to be that they control the click which equals the treat. The click is never independent of the dog’s prior behaviour. Good trainers give information to the dog five times faster than poor ones so practise the skills, become faster, and never start until your treat is ready to deliver.
The click is also NEVER given without the treat. If you click accidentally still give them a treat, no matter what their actual behaviour might have been. They need to know in the core of their being that click IS treat (c/t). You can treat without clicking whether deliberately (eg. a dog who needs some calming) or accidentally (food which falls before they’re reliably progressed on ‘zen’). Don’t panic – cover it with your foot if you can or just let them eat it unless it’s your Sunday roast and you need to rescue it!
Sue starts with target because it gives immediate, frequent, positive feedback to the dog and most traditionally trained dogs and people can still engage with these requests. The dog targets my fingers or my pen and gets a click/treat – if they want another treat most dogs within three attempts will have learned how to control my behaviour so that I will deliver the treat. Hand zen and counter zen are introduced early because this lifeskill is very important to most people living successfully with a dog.
Training is all accomplished by shaping, capturing, or luring and most training is a combination of all three. Like us dogs also process experiences while they sleep and so apparent lack of progress one day or several can be followed by leaps in understanding both for the dog and for the owner!
Luring is the quickest means to an end behaviour in many cases (eg. luring a sit is quicker to the second sit than either shaping or capturing but after a few lured sits let the dog offer them without the lure – capturing the behaviour instead strengthens the learning). The lure needs to be faded quickly so the dog does not rely on it unless the lure mimics a cue/hand signal in which case that cue can continue sometimes (if you always use it the dog will not offer the behaviour unless lured) as training progresses. When you reach the stage of wanting a stand, a nice one can be lured by planning to have the dog back into it – lure from end if nose along mid-jawline below to throat so they are tucking head and backing up.
Shaping a dog to a particular behaviour means c/ting closer and closer approximations in the dog’s behaviour. Begin with baby steps so you both can succeed. If you want the dog to go around a post (or stool or wastebasket or whatever is handy) sit yourself down, toss a treat towards the pole from a foot or so away (a bit of a lure) then c/t while they are moving in that direction, and toss another treat towards or in front of them; c/t several times while they are in area of pole and in motion; then let them come back to you or wherever (no reward) but keep c/ting every time they take any step towards the pole – the direction (clockwise/counterclockwise) doesn’t matter. A 10-20 second session (8-16 kibbles) to begin is best to be really effective. Keeping sessions short will prevent you becoming frustrated! For most dogs who have never been rewarded for taking initiative, shaping requires a constant stream of reinforcement. However, 2-10 minutes after your 10-20 second training session you can do it again; and again; and again. When the dog is offering to go around the pole, click three quarters of the way around and drop the treat at your feet. Let them do this a few times then move your chair back a couple of feet and see if they will still offer the behaviour. After they’ve offered it 100 times you’re ready to add the cue word immediately after the click (if you will be doing herding the cue will probably be different than if you’re into agility). Read Sue’s Levels Training for the means of progressing with distance, direction, and additional posts.
Anything which the dog frequently does in daily life, such as sit and lie down, are easiest (not fastest necessarily) to train by capturing it. You are ready with clicker while you watch TV and the dog happens to lie down so you c/t immediately by tossing the treat so the dog will get up. Then they need to think about what they did that made you c/t; stay still and keep your arms crossed so they can’t target. Be very patient – whether they repeat the down quickly or not they are thinking. Eventually they will get tired and lie down again – immediate c/t. This is one case where you are not rewarding frequently until they start offering the behaviour so don’t do this before a mealtime.
Formal training times should be brief – 2 seconds to 10 minutes is sufficient! It will take you longer to set up the room until you get comfortable with the process. Early on you could have stashes of kibble and cheese around the house and on your body so you can always deliver a treat when needed. The dog will tell you if a reward is good – the behaviour will increase. If a desired behaviour decreases then that is not a reward to this dog at this time or the reward is not frequent enough (about 8 treats in 10 seconds for beginning a new behaviour) or there is something in the environment which is punishing the desired behaviour. Perhaps the dog is not hungry or is too distracted in which case you need to change times or locations. The half hour before the dog’s mealtime is precious training time (treat with their meal kibble) and you can use the bathroom as the first training centre.
The focus is to catch them doing it right – over and over and over again. Food is always the primary motivator/reward because in a training session, if you are going to use a toy as a reward, then the dog needs to be given a minute to actually play. As your partnership progresses, learn what else is rewarding for your particular dog in different circumstances so you can vary the rewards and opportunities to interact. Teach your dog to play with a wide variety of toys both independently and with you. Some dogs might love to tug with you but only in private. Some dogs will retrieve a certain size or feel of ball better than other types. Some dogs will kill every squeaky toy and others will carefully save them to cuddle. Destroying a certain type of toy is a reward to some dogs so plan to give them the chance sometimes!
Lifestyle rewards: going out into the backyard is a reward – ask (by luring) them to target a pen or to sit (not sit/stay yet just one thing at a time) before you open the door. Anything that the dog wants to do can be a training opportunity: jumping up on the couch or going for a walk or getting breakfast are all chances to catch them doing something right then REWARD by allowing the behaviour.
Sue’s oldest dog spends time each evening playing a completely new shaping game as she solves problems for other service dog owners/trainers, her younger dog plays sometimes. These behaviours are never named; they are problem-solving exercises and practise thinking sessions for her and the dogs. When requested to demonstrate she shaped first 10 year old Scuba, then 1 year old Stitch to spin a wheel on a dolly which had been turned on its side and placed across the room – Scuba worked up to solidly spinning the wheel and repeated the behaviour several times in less than two minutes. Sue’s Levels are designed to develop you and your dog into a thinking team that can learn most new behaviours in three minutes. Putting it on cue takes longer.
If anyone wants to compete in any dog sport, including conformation, or simply to develop a better relationship with their partner-dog, I would strongly recommend attending any of Sue’s seminars.
Sue has authored a Training Levels book which is available in three different formats for free download at her website www.dragonflyllama.com.
Subscribe to:
Post Comments (Atom)
No comments:
Post a Comment