Identifying the right rewards
When starting training for obedience or a behavior modification program, we must first identify very valuable rewards for the pet. For most dogs this will be delectable food treats. Food treats should be tiny (less than the size of a pea) and readily consumable. Some options include small pieces of deli meat, small cubes of cheese, small strips/cubes of fresh meat, or semi-soft liver jerky cut into tiny pieces. Consideration should be given to any medical dietary restrictions (e.g. low fat cheese). Generally, for most dogs, dry dog food and even dry Shmacko strips are not delicious enough to be a worthy training reward that the dog will really strive to win. The reward MUST be craved by the dog to be effective in the reinforcement of behaviours!
Some dogs do not value food treats at all, which can make training more complicated. If your dog is not food-motivated, you must find an alternative valuable reward, such as a short game of tug or fetch, or, in the affection-craving dog, a soft tummy rub or rump scratch. Praise should always be part of the reinforcement package in addition to food treats, affection, games etc.
The training of toy/game motivated dogs will require slight adaptions to the reward system, which is traditionally designed around the handler instantly providing food morsels in the hand. The use of a conditioned reinforcer, however (which will be explained below), negates this problem. While games need to be exciting enough to be rewarding, stimulation can cause excitable dogs to become distracted, so games need to be as low key and short in duration as possible, while still providing the required level of reinforcement.
Always remember that the greatest reward can be giving the dog his freedom to go off and play, even if only momentarily, and this should be incorporated into all training sessions at every opportunity. In other words, when the dog excels at something, even if it is only a couple of minutes your planned fifteen-minute training session, end the lesson and reward the dog with his freedom, followed by his favourite game, or his dinner if the dog is food motivated.
The Four Key Principles of Successful Training
1) Repetition and progression in small steps
Dogs do not have the powers of cognitive reasoning that we humans possess, so they need lots of repetitious training to learn what it is that we want of them. We must train over and over again at the same thing, progressing in small steps only when the previous level has been achieved at a near perfect success rate (success = nine times out of ten). We must starting training each individual skill in very quiet environments with no distractions (such as the hallway, lounge room or quiet back yard), building up very gradually to complex environments with loads of distractions (such as the off-leash dog park or beach) only when all the other levels have been successfully achieved. Jumping ahead to the latter is a recipe for disaster and the dog is sure to fail. This is the most common mistake dog owners make. The dog will sit for his dinner in the kitchen, so they assume he will sit at the beach when dogs and kids are running everywhere! This is unrealistic, but many owners do not understand the power of distractions in turning an obedient house dog into a confused, distracted beach boy!
2) The Conditioned Reinforcer – “Yes!”
The fastest way to train a dog is to ‘mark’ the desired behaviour the instant it is offered, whether unsolicitored, or in response to a command.
In the Clicker Training method, a click of the clicker becomes the ‘mark’, given the split second the correct action is offered. The mechanical nature of the clicker is excellent because it offers the same sound every time it is pushed. This is important because dog training must be consistent and form a repetitious pattern for the dog to recognise and respond to, in order to win a reward. The need to carry a clicker everywhere we go, however, can be a problem and for that reason, I tend not to use clickers, but prefer to use my voice, which is always readily available, to ‘mark’ a desired behavior.
Instead of a clicker, the ‘mark’ most widely used in international training circles is simply the word “Yes!” It should be said in a happy, congratulatory way, as if giving the dog a verbal high five. To be effective, the ‘Yes!” must sound the same every time, so establish how you are going to say your “Yes!” and stick to it. Don’t let it be prone to your emotions, such as frustration, excitement, boredom, anger or relief etc.
Having earlier established what reward/s each individual dog most wants, we need to train the dog that the “Yes!” means: “That’s correct, and now your reward is coming”. We need to condition the dog that the sound of the word “Yes!” means good things, by immediately following each and every “Yes!” with a treat of whatever they most want – a delicious morsel or a game of tug/fetch. To do this, set the dog up in a quiet place, say the word “Yes!” and immediately hold a treat in front of the dog’s nose for him to take. Wait a few seconds, say “Yes!” and give another treat. Do this over and over again for several minutes, with pauses of varying short lengths, in batches of between four and eight ‘yesses’ in a row. Perform this cycle of conditioning of the “Yes!” at least four times a day for the first week.
As the dog progresses, occasionally say “Yes!” when he close but is not looking at you, and be ready to present the treat to his nose to ensure you immediately follow every “Yes!” with a treat. After a few more days, when you think he’s getting really good, try it when he is even more distracted by something, but be sure to follow up with the treat, even if you have to ‘chase’ his nose! After a few days, your dog should begin to look at you as soon as he hears the “Yes!”. Once you are at this stage, introduce a longer period between saying the word “Yes!” and following up with the treat – extend the gap to three seconds, then five, then seven and up to nine seconds, but no longer than that.
The aim of this conditioning exercise it to have the dog whip his head around to look at you in an excited, expectant manner, waiting for the treat when you utter the word “Yes!” in that particular way, and wait expectantly for between five and ten seconds to receive his treat reward. When you have achieved this, you have succeeded in conditioning the word “Yes!” and making him feel as good when he hears the word “Yes!” and during the anticipation period of waiting to receive the treat, as he does when he actually receives the treat. This means you have turned a word (a sound) into a reinforcer – a conditioned reinforcer! Congratulations! You can now train your dog to do anything that is physically possible, through use of this conditioned reinforcer, even if it takes you up to ten seconds to follow-up with a treat.
Another important training element is that the timing of our ‘mark’ must be extremely accurate. If we are too early or too late with our mark, the dog may associate it with something other than the desired action. For example, having commanded a sit, if we say “Yes!” just as the dog begins to bend its hind legs, but before its bottom actually touches the ground, it could jump back up without sitting at all, and we have marked the wrong action. If we are too late, and say “Yes!” after the dog has put its bottom on the ground and had time to start having a scratch, or, worse, jumps up just as we get around to saying it, or if the dog spies next door’s cat run by just as we belatedly say “Yes!”, the dog associates the mark with something completely wrong, such as chasing a cat. For the dog to understand the reinforcement process, our timing must be absolutely split-second accurate – every time. If you get it right, he will quickly understand that “Yes!” is the best thing he can hear because it always means something great will follow, and he begins to associate it with the action he just performed. When he gets this, and wants to repeat it to win that reward, your training is going to go ahead in leaps and bounds.
4) The Release
We have established the process whereby the dog responds to our command by offering us the desired behaviour, and we mark the action with our conditioned reinforcer – the “Yes!” followed by the dog being given his primary reinforcer reward (delicious treat/game etc). Incorporating the release into this process becomes the next important step. When applicable, such as when the dog has been asked to sit, drop, heel, come, stand, or take up any other static position, the dog should be released from that position, using the (internationally-preferred) release word “Free!”. As with your ‘Yes!” word, your expression of “Free!” should be the same every time, preferably uttered in a high pitched, happy tone, because this release becomes a reward in itself and it is a happy occasion. Immediately follow the release with lots of other reinforcement. Always remember that the concept of freedom, whether from a static position or from the entire training session, is very much a reward that can be utilised to great effect, especially in very active dogs. The release, therefore, becomes a primary reinforcer, in conjunction with the dog’s other favourite food treats or game rewards.
Putting It All Together
To summarise this process, let’s use the example of The Recall – commanding your dog to come. This is one of the most common problematic areas of obedience!
When commanding the recall, the sequence should be as follows:
1. If necessary, get the dog’s attention by saying its name, “Fido…”
2. Once he looks at you, in response to his name, run backwards and tap your leg to make him run towards you
3. As he rushes towards you, give the command: “Come!” ONCE only.
4. This can be followed (if necessary, and if there’s time) with further encouragement: “That’s it, good boy, keep it up, this way!” etc as you run backwards, slapping your leg. Be excited and be attractive – make yourself a ‘party’ to run towards! You need to be much more fun and interesting than other temptations.
5. The instant Fido reaches you and you take hold of him/his collar, and mark his recall response with: “Yes!”
6. Have your treat ready, give the release word “Free!”, and let go of his collar, then immediately offer him a delectable treat, followed by praise, affection and a short romp – or whatever else the dog most wants – which is usually his freedom.
7. Repeat the process a number of times with your “Yes” perfectly times to match his arrival, but gradually pause before giving the release, using lighter and lighter hand holds to keep him there, until you don’t have to hold him at all before your say “Free!” and allow him to break from his position. Correspondingly, allow a longer period of playtime, or give a jackpot reward (three consecutive treats).
8. NOTE: Initially, it doesn’t matter whether he sits, drops, stands, or does a back flip when he comes in – only ever expect him to simply come!! Don’t expect the dog to immediately follow every recall with an automatic sit. This is a common error, because initially he should be immediately be marked with the “Yes” and receive lavish praise and treats to reinforce his recall, not a sit. We must not complicate the entire process or confuse him by waiting for a sit, thereby ruining the effect of quickly marking and reinforcing the actual response to the initial “Come!” command! It is a recall exercise, not a sit exercise!
9. Never say “Yes!” until you have hold of him or his collar, because dogs that learn to grab the treat and bounce back away again before being officially released have not completed the recall and should not be reinforced! Never allow a dog to develop this behaviour! Always have hold of his collar or his leash before you let him have the treat.
10. If you repeat with this process over and over when next training the recall, you will have the best recall response in the land! For dogs whose recall is not good (or non-existent) when off leash, start by perfecting it on a leash over a very short distance, then move to a longer and longer leash. Better still, start in the closed-off hallway at home, where there are no distractions, then progress to the back yard. Only when each stage is almost perfect (the recall is prompt and responsive nine times out of ten), should you move onto the next level. If you try to perfect recall in an off leash at the park, don’t be surprised (and certainly not angry or frustrated) when he runs in the opposite direction. It must be perfected in small stages in quiet environments under the absolute control of a leash/long-line first.
11. Always finish on a particularly good response – many short lessons are more effective than one or two long ones, so let your dog know when he hits the jackpot with a fast, accurate recall, by giving him the jackpot of reinforcement – an end-of-lesson release, five treats and a big game of fetch or his dinner!
THOSE LITTLE EXTRAS!
The Target Fist
When an animal is trained to attend to a target, they will follow that target, allowing the handler to easily lure them into certain positions (e.g. sit or drop) and to redirect their attention away from competing attractions. Using the closed fist as the target makes great sense, since it is always with us. It also is a natural place to hold a treat/toy. To train a pet to target the fist, simply put the treat in your hand and close the hand into a fist. Allow the pet to smell the closed fist, then release the treat. After many repetitions, the pet readily focuses on the closed fist in anticipation of a tasty morsel. Then the fist can be manipulated in different directions. Where the closed fist goes, the head follows, and then the body follows. If the target fist is brought from the pet’s nose up and back over the head in a gentle arc, the pet will sit; if the target fist is brought up toward the trainers’ eyes, the pet will make eye contact, etc. As the pet successfully completes these tasks, it is rewarded by release of the treat from the target fist. Once the pet has established great compliance with following the target fist, the rewards can become intermittent from the fist.
Many people yell commands repeatedly at their dogs in order to achieve compliance. In all pets, but especially those with behavioural problems, yelling/loud voices can increase arousal levels and/or aggravate anxiety. Both of these consequences are counter-productive when you are trying to teach a pet to respond and behave in a tranquil manner. Besides which, dogs have excellent hearing, so keep your voice firm but calm when giving a command.
If the pet is not paying attention, before giving the command, gain the pet’s attention by saying their name (once), then the command should be given – ONCE – in a quiet, calm, firm voice, and there should be a pause to allow the pet to respond. Good responses must be immediately marked with your conditioned reinforcer (“YES!”), and rewarded with the primary reinforce treat/game. Non-responses or undesirable behaviour is not rewarded. You may be able to gain compliance by luring the dog into position with the treat in the fist, or with some gentle pressure to guide the dog into the desired position (e.g. sit). If this is not possible, the situation needs to be changed so the pet can be compliant.
It is very important to remember to release the pet from the commanded position, not let it break the position of his own volition or wander off, because ultimately, the pet should learn that he must maintain the commanded position until told otherwise. The word/command “stay” is therefore obsolete in the well-trained dog under modern training psychology!
An unresponsive pet may need constant direction when exposed to a provocative stimulus. For example, if the dog is jumping up at a visitor, you will need to become more interesting than the visitor if you expect to get the dog’s attention and compliance. Similarly with an anxious pet, you will need to become the pet’s rock of assurance when faced with a fearful stimulus, such as the approach of a large dog with an unknown handler.
The pet should stay engaged with the handler via a constant dialogue, so the handler could say “Sophie, sit. Yes! (treat). Watch me. Yes! (treat), good dog (treat). Watch me. Yes! (treat) Good dog. Yes, Sophie! Watch me. Yes! (treat), good dog”. Success is unlikely if the pet is given a single verbal command such as “sit” or “stay”, and is then expected to hold that commanded position for a prolonged period without the distraction of the dialogue, particularly if fearful or over-excited. In particularly difficult situations, the command, followed by assisted compliance (guided by hand contact or by using a leash and head collar for control) may be required, then the conditioned reinforcer can be given to mark the compliance: “Yes!”, and finally, reinforcement through treats, praise – “Good Dog!” – and release if appropriate or games/affection, should be forthcoming.