1. Effective training requires effective reinforcers. Establish a large selection of effective reinforcers, preferably while your dog is still a puppy/adolescent. Take care not to ruin your dog’s appetite, and emphasize teaching your dog to play with many different objects. So-called high-drive dogs and the classic food hounds are the easiest dogs to (clicker) train. It’s also great if you can condition an arsenal of reward substitutes (anything from clapping your hands to ruffling the dog’s fur, and so on, can be effective reward substitutes if you associate them repeatedly with food or play). And not least – teach you dog to work for things he wants in the environment. Then you will even be able to use things others see as distractions as effective reinforcers. By working systematically on developing many different rewards and reward substitutes you won’t have to get stuck with only using click + treat.
2. Socialize your dog well, teach him how to be around other dogs, and make him used to a bit of rough, uncomfortable handling by both you and other people (he should learn to like this!). It doesn’t matter how well you clicker train your dog if he spends a lot of his attention on trying to control the environment.
3. When you want to improve your training or solve a problem you should first examine 1) timing, 2) criteria, 3) rate of reinforcement and 4) quality of reinforcement before you look for more advanced solutions.
4. Keep your training clean. Avoid sloppy training (luring, nagging, repeated cues, hands in your pocket before the click, treats in your hands, smacking your lips, bending over the dog, visible treat bag, reward hand held in front of your belly/chest, low rate of reinforcement, unplanned criteria and so on and so forth). Practice your mechanical skills. Video tape your training if possible, attend classes with a skilled clicker trainer – or join a Chicken Camp!
5. Start early with building a large file of offered foundation skills. You don’t necessarily have to train them perfectly right away. You can do that later, once you have quickly touched on a large number of behaviors so that your dog has become creative. THEN you can start establishing good fluency with the behaviors.
6. To develop the behaviors to fluency is crucial to effective clicker training. Clicker trained behaviors without fluency are usually worthless in the real world (and in competition). Start training new behaviors in a distraction-free environment. Establish fluency for the behavior before you take the behavior out into the world and really generalize it.
7. Always use a high rate of reinforcementin early training stages. Ideally the dog should be earning a click at least every three to five seconds (depending on which behavior you’re working on). It is not good clicker training to stand around waiting for a long time before the dog is finally able to offer the correct behavior.
8. If the dog performs correctly for at least 80% of the repetitions you can increase the degree of difficulty. If the dog performs correctly for less than 80% of the repetitions you should consider temporarily lowering criteria. This yields a quicker progression and a more reliable behavior in the end (you avoid mistakes attaching to the behavior’s record) But remember that the 80%-rule is only a guideline, not a law of nature!
9. Use a higher quality reinforcer when you’re training new behaviors, difficult behaviors, strenuous behaviors, or when there are a lot of distractions around.
10. The most important thing is to click accurately, but remember that the dog keeps learning after the click as well. With some behaviors you can achieve faster results by being conscious of how you deliver the reward (not only of when you click). This is called strategic delivery of reinforcement or reward placement. This can be particularly useful at the beginning of training. But don’t get dependent on a particular reward placement forever. On the contrary, once you’ve got the behavior, you should start to proof it by using e.g. reverse luring, distance rewards and so on so that the dog is able to perform the behavior correctly no matter where the rewards are coming from.
11. My ideal for training new behaviors is 1) as little help as possible 2) as few errors as possible 3) as quick a progression as possible 4) as little correction of mistakes afterwards as possible. Choose a training technique which will help you to achieve this. Use free-shaping/capturing as your plan A. Adapt the environment if necessary so that the shaping process will be easier. Consider using targeting if shaping is difficult and leads to a lot of errors or little progress. Luring is only to be used as a last measure, and should then be removed as quickly as possible.
12. Pavlov is always looking over your shoulder. Remember that e.g. the intensity level is conditioned along with the behavior whether you want it to be or not. It can therefore be a good idea to train typical high-intensity-behaviors (such as recalls, jumping and so on) when the dog is fresh and full of energy, while you train low-intensity-behaviors (long down, crawling and so on) towards the end of the session or after a long walk, when the dog is calm and relaxed.
13. As soon as the behavior is fluent you can put in on stimulus control (if you need it for that behavior). But don’t lose the offered behavior completely. It is always nice to be able to go back if you need to brush up the behavior / cue later.
14. Overtrain foundation skills and other behaviors that are important to you (this means establishing extreme fluency and training with more difficult distractions / variations than you really need. Then you will always have a margin of error in case you need it.
15. If you want to progress as quickly as possible, train frequently, every day if you can. If you think it’s more important to get the maximum benefit out of every hour you train, you can instead let a few days pass between each time you train a particular behavior. Dogs seem to learn during breaks, as well... And not least, you often get much more out of training when the dog is well rested. If you want to train a lot your dog should be allowed to get used to this gradually – and you should have a breed/individual which is suited for large amounts of training.
16. If you want to clicker train effectively, don’t mix punishment and luring with your training. Punishment and luring work if you use them correctly, but they always reduce the effect of clicker training if you mix them with the same dog.
17. Train a new behavior you haven’t trained before as often as possible (or train an old behavior in a new way, or with some other animal!) That’s the best way to keep evolving as a trainer. It’s not so important whether it’s a trick or a useful behavior, but that you progress as a trainer.
18. Clicker training is a young training method in the dog world, and continually developing. Knowledge gets old quickly, so keep up-to-date. There are good resources to be found both in books and on the Internet. Stay in touch with other clicker trainers. If you don’t, it’s easy to stagnate, because the development is rapid.
19. Clicker training can be made very advanced for those who wish it. Clicker training is based on science which takes years of full-time study to learn. But don’t let that scare you! The fact is that if you only master the essential main principles well, you can teach your dog practically anything! (leave the deep-down theory to the clicker training nerds).
20. Don’t just clicker train your dog. Bring the principles of learning into your daily life as well. Use positive reinforcement with your children, your spouse, friends, colleagues, waiters and parking guards. It could change your life. And if you work a lot with teaching people – learn Tag Teach!
21. If you discover any new training method which is more effective than clicker training (and still ethically sound) – switch to that method the very same day, and never look back!
Vir: Morten Egtvedt, Bob Bailey, Karen Pryor, Ken Ramirez, Kathy Sdao, Morgan Spector in Cecilie Køste.