I read an article recently by veterinarian and animal behaviourist Dr. Sophia Yin, in which she was talking about a book that she had read. The book was titled, “The Compass of Pleasure: How Our Brains Make Fatty Foods, Orgasm, Exercise, Marijuana, Generosity, Vodka, Learning and Gambling Feel So Good,” by neurologist David Linden. I have to say that is quite the title, and I did wonder how he decided on the order of the list of things! Let’s take the example of the person who loves to eat chocolate. If the chocolate lover eats an entire bar at once, he gets a big hit of pleasure, but can’t do it very oft en. Meanwhile, the chocolate lover who breaks the bar into many small squares and enjoys them one at a time, throughout the day, receives many small hits of pleasure in his brain. Linden states that the brain gets more pleasure sensor stimulation overall from the second method of eating chocolate.
We can apply this understanding to what’s going on in our horses’ brains during clicker training. In feeding two sets of 20 treats, one at a time, as rewards for desired behaviour in our horse, we have provided 40 hits of pleasure to our horse’s brain. This quantity is very easily reached; in practice, a full day’s training can deliver close to 100 treats! The many small rewards given in a training session repeatedly stimulate the brain’s pleasure centre, leading to a strong desire by the animal to repeat the behaviour that led to the reward and the stimulation of the pleasure centre. What a great thing.
Given this scientific explanation of why high rates of reinforcement can help with training, we should practice and reward the behaviours we desire in our horses frequently, keeping the reinforcement rate high enough for them to want to repeat the behaviour. (Note: A horse or any animal trained with negative reinforcement will work only at the level necessary to avoid the negative stimulus.) Let’s apply this idea of Linden’s to the standing on a mat exercise. When we fi rst begin this exercise, we want to make standing on a mat a really good and pleasurable experience so that the horse wants to do it again. To achieve this, we need to start out with a high rate of reinforcement to stimulate that pleasure centre. This high rate of is high rate of reward is also called “sequential rewards” (Dr. called “sequential rewards” (D Yin). Th e treats are delivered about eve e treats are delivered about every 3 seconds to start with, and that is just about as fast as you can possibly click and treat!
(Note: To avoid weight gain from extra calories, take the “treats” you use for clicker training out of your horse’s normal rations.)
We will assume that your horse is not afraid of the mat and is willing to walk up to it and put his feet on it. (For a detailed explanation of this foundation exercise, please check back in the archives of Saddle Up.) As soon as your horse’s two front feet land on the mat, start your high rate of reinforcement. Remain aware of where you are feeding him and be prompt with the treat, but without a rushed feel to it. Click and treat several times, then ask him Click and treat several times, then ask him to move off move off the mat. If you continue too the mat. If you continue too long on the mat you will get your horse on the mat you will get your horse “stuck” on the mat. We want him to seek on the mat. We want him the mat but not be stuck to it if we ask him not be stuck to it and to move off it. All things must be kept in balance, just like any good training.
Now ask him to return to the mat, and once again repeat the high rate of reward sequence. Keep doing this loop of “on mat” and then “off mat” until he has the idea that standing on a mat is a great thing to do. Dr. Yin suggests that when the behaviour is reliable we can move to a slightly longer time interval between treats - extend it to 5 seconds at first, and then increase to 7, then 9, 12, 15 seconds, and so on, gradually building longer duration into the behaviour. A point to remember: do not to wait too long to click - you risk causing frustration and a breakdown in the behaviour. Observe your horse carefully and use your best judgment.
Sequential rewards using several small treats will stimulate the pleasure centre multiple times compared to a single stimulation using the same number of treats but given all at once. Dr. Yin concluded her article with, “I think the take-home message is clear. Some people may think that rewards are a somewhat weak way to train... however, when the right rewards are used consistently and predictably at a high reinforcement rate, they can become extremely strong.” In other words, the method of giving small, frequent rewards during training is much more likely to reinforce desired i behaviours than providing a large, single b treat at the end. Perhaps the next time someone comments about you using treats to train, you can use the chocolate example to help explain why clicker training works so well.
For fun, take a look at the above video. Tell me if you think this horse knows what a click means! I think you will recognize the famous pair in the video, too (Steffen Peters and Weltino's Magic). If you have a training issue that you would like to see covered in a future article, please email me at firstname.lastname@example.org. Until next time, keep it positive.