Monday, October 3, 2011

Do's & Don'ts /with Explanations


DO give a treat for every click, even if you clicked by mistake, you can sort out the mistaken click later.  Clicker training is very forgiving.  It is far better to give an occasional reward that wasn't earned than to dole out punishment that wasn't deserved.

DON'T give treats during training unless you clicked first because it weakens the training bargain.  Your training bargain is: click equals treat.  In the beginning of training, you want to carefully keep that bargain.  Later on you can be more flexible AFTER your dog knows her "job".

DO understand that click ends the behavior and so it is expected that your dog will stop doing the behavior when you click.  You can work on other criteria (such as duration of stays or heeling) at another time by using a "keep going signal" (KGS).  I use the word "yesssss" (with an extended "s" on the end) to tell the dog she is getting closer to the click and treat.

DON'T try to shape more than one criteria at one time.  Dogs learn better, faster and more consistently if they are presented with just one challenge at a time.  So if you've been teaching sit stay and your dog is able to do it for one minute and now you want to take it on the road to teach it in a distracting environment, do expect to lose your duration.  Start with a much shorter duration and build the time up gradually under this new distracting condition.

Gimme here:  My person taught me to stay by first teaching me to wait for permission for important stuff, like breakfast.  So then the day she taught me stay was very easy.  Then when we took it to the shopping center, it was hard to stay for very long.  We started out with me just staying for two seconds.  Over time we built the time up again.  My person knew that stays are going to be very important in my career, so she really took her time when she taught me and now my stays are amazing.  And that's the truth.

DO use the clicker to reinforce attention from the dog.  You can use the clicker to reinforce any behavior and attention is likely the single most important behavior you want your dog to learn.

DON'T use the clicker as an attention getting device; you could be reinforcing behavior you don't want.  Its not a remote control.  Whatever the dog is doing at the instant you click is the behavior that is rewarded and likely to happen again.  So be sure to use it only when the dog is doing something you want to see more of.

DO teach your dog the meaning of a keep going signal (KGS) and use this in training to tell the dog to continue what they are presently doing and that a click/treat is coming.  The best way to use the KGS is to simply use it appropriately in training.  Your dog is really smart and she will figure out what it means.  So if you have been clicking for two steps of heeling and you want more, then give your keep going signal at two steps and then click on the third or fourth step.  Extend this gradually and your dog will learn what it means.  Also, remember that you can use the keep going signal more than once before you click to help your dog understand to keep on trying the present behavior.

DON'T use the clicker as a KGS because your dog will be confused.  Dogs do not understand "context" in all its nuances like humans do, therefore they can only understand one meaning for any one cue (or marker).  If you use the clicker as a marker for "you are right in what you are doing and here's your treat" some of the time and then some of the time as a keep going signal that means "you are getting close to being right", the dog will become confused and not understand which meaning you intend.

DO reinforce behavior you want to see more of, rather than ignore the dog when it is behaving well.  If puppy has been wild all day and has finally settled down to chew its bone, its our human tendency to not do anything that might get puppy stirred up again.  If you do that you are ignoring good behavior.  If you ignore that calm behavior, you may end up with a dog that only calms down when he's exhausted.

DON'T punish misbehavior, you could be reinforcing the behavior with attention.  Dogs thrive on attention; dogs are very social and need our attention.  Remember the song refrain, "Even a bad love is better than no love at all"?   Your dog is likely to think your attention (even a swat) is better than being ignored.

DO manage your environment and training to set your dog up for success and reward.  It is far better to plan life with puppy so that she doesn't have the opportunity to practice bad behavior.  Practice does not make perfect, only perfect practice makes perfect.  If your dog gets to practice bad things, he'll get better at doing bad things.

DON'T "proof" your dog by setting him up to make a mistake. You don't like to be set up to fail and neither does your dog.  "Correction Proofing" only teaches your dog that she can't trust you to be fair.  You want to use "Challenge Proofing" where you set your dog up to be right in gradually increasing challenges.  This will teach your dog how to succeed and be confident.

DO use a high rate of reinforcement and your dog will be eager to train.  In a study of dog trainers, the biggest thing that separated merely good enough trainers from really good trainers was their generosity at dishing out reinforcement.  That doesn't mean you reward any old behavior; rather you carefully observe your dog to find LOTS of good things to reward.

DON'T worry about spoiling your dog with too much reinforcement.  Positive does not equal permissive.  You can set appropriate limits and train our dog to whatever level of excellence you want and still be a kind reward-based trainer.

DO train off leash whenever possible in a safe environment.  Don't let the leash become a physical cue to your dog that now is the only time my person can enforce the rules.  It is far better to teach your dog to want to give you what you want so she can get what she wants.

DON'T use your leash as a training tool.  There is no such thing as a "motivational leash pop" and calling it motivational does not make a leash correction any more fun for the dog.  That's about like calling a face slap "motivational".  It is still a punishment, plain and simple.

DO have a clicker on your person at all times and treats stashed in many places so you can reinforce good behavior whenever it occurs.   Always be prepared to get in a few minutes of training.  The best training sessions are very short.  Fifteen minutes of training per day is a good goal, but you will get better benefit from three 5-minute or five 3-minute sessions per day.  Plus it will be easier for you to squeeze that into your busy life.

DON'T miss impromptu training opportunities.  Training opportunities are present all the time, so don't miss out on using them.  If you are working on loose leash walking and you come upon something enticing on the sidewalk, let go of LLW and use that real life chance to practice "leave it".

Gimme here:  I was walking with my person and found this really great bag of garbage stuff smashed into the pavement.  It was very interesting and I really thought we should investigate it and probably take it home with us.  My person had a different idea.  So we walked by it several times and even around it several times and she gave me treats to ignore it.  As cool as that smelly garbage was, the cheese and steak and chicken I got for ignoring it was even better.  It pays to be clever and smart like me.  I'm just saying.

DO make a list of all of the things your dog finds reinforcing and rank them in order -- remember this is the list for your dog's preferences, not what you think they should be.  There are so many things your dog is willing to work for, if you just know what they are and control access, you can use them to your advantage.

DON'T limit the kinds of things you use for reinforcers.  If my dog wants to go outside, then its a want that I can use to get in a training moment.  Cue a sit or down and reward that behavior by giving the dog what she wants.  You are going to have to open the door anyway, you might as well make a training moment out of it.  All those moments add up.

DO carry a variety of treats so you can use a special treat for a special effort (such as ignoring distraction) or for sudden leaps into brilliant performances.  A special effort or moment of brilliance that goes unrewarded is an opportunity missed.  It may be a long time before it comes again.  Always be ready... even if that means you say your "super reward" word as a marker and follow that with a dash into the house, with pup hot on your heels, to the refrigerator.  Running with you to get the reward can be part of the reward since it adds an element of FUN.

DON'T become predictable to your dog about what you will treat with.  You can use anything that works for your dog.  My young dog loves garlic bread (cut up into little 1/4 crouton sized pieces).  My old dog will do anything for popcorn.  What is your dog's favorite?  Make a potpourri of dog training treats that range from "okay" to "scrumptious".

DO be a benevolent leader who sets fair rules and limitations and teaches the dog what they are and how to live within those rules.  Dogs don't understand our rules and for them its like living in a foreign country with weird customs.  Take the time to teach them how to be a welcomed addition to the family.

DON'T be a malevolent dictator who changes the rules or behaves unpredictably or unfairly to the dog.  You wouldn't like this in a boss or friend.  Why would anyone think a dog would like it?  We all prefer to be around someone who is fair, even if they have high expectations.

DO split goal behaviors into fine criteria and shape from one to the other in a steady progression.  Learning a new behavior is like a set of stairs... its far easier to get to the top one step at a time than it is to try and do it in one giant leap.

DON'T lump behaviors and wait for complete or flawless performances before you reinforce them.  When you get a new job, you will do better and have a smoother transition if you learn the new skills one at a time, using the skills you learned the day before to build on.  The same is true for your dog.

DO plan your training session and how you will shape each behavior so you can reinforce the progression as it occurs.  If you plan ahead how you want to approach the day's training, you will be more likely to make steady progress.  All big projects go better with a little bit of planning.

DON'T miss out on those sudden intuitive leaps skipping several steps, this is the perfect time for a jackpot.  You don't have to be a slave to the plan though, be prepared to change gears and take advantage of the training opportunities that come your way.

DO concentrate on the behavior you want to see your dog do, instead of the things you don't want to see.  Extinction occurs when a behavior goes unrewarded.  So if you reward what you do want and use jackpots to really emphasize the best efforts, your dog will spend her efforts trying to get the jackpots.

DON'T seek to get rid of an undesirable behavior without having a new behavior to replace it with.  Idle paws abhor a vacuum, so always be prepared with a replacement behavior.

No comments:

Post a Comment